Unify DataServer: RHLI Reference
Transcription
Unify DataServer: RHLI Reference
Unify DataServer: RHLI Reference R E 1996, 1997, 2001, 2005 by Unify Corporation, Sacramento, California, USA All rights reserved. Printed in the United States of America. No part of this document may be reproduced, transmitted, transcribed, stored in a retrieval system, or translated into any language or computer language, in any form or by any means, electronic, mechanical, magnetic, optical, chemical, manual or otherwise without the prior written consent of Unify Corporation. Unify Corporation makes no representations or warranties with respect to the contents of this document and specifically disclaims any implied warranties of merchantability or fitness for any particular purpose. Further, Unify Corporation reserves the right to revise this document and to make changes from time to time in its content without being obligated to notify any person of such revisions or changes. The Software described in this document is furnished under a Software License Agreement. The Software may be used or copied only in accordance with the terms of the license agreement. It is against the law to copy the Software on tape, disk, or any other medium for any purpose other than that described in the license agreement. Unify Corporation values and appreciates any comments you may have concerning our products or this document. Please address comments to: Product Manager Unify Corporation 2101 Arena Boulevard Ste. 100 Sacramento, CA 95834-1922 (800) 248-6439 (916) 928-6400 FAX (916) 928-6406 UNIFY, ACCELL, VISION, and the Unify Logo are registered trademarks of Unify Corporation. Unify DataServer is a trademark of Unify Corporation. UNIX is a registered trademark of the Open Group in the United States and other countries. The X Window System is a product of the Massachusetts Institute of Technology. Motif, OSF, and OSF/Motif are trademarks of Open Software Foundation, Inc. SYBASE is a registered trademark, and SQL Server, DBĆLibrary, and Open Server are trademarks of Sybase, Inc. INFORMIX is a registered trademark of Informix Software, Inc., a subsidiary of IBM. INGRES is a trademark of Computer Associates International, Inc. ORACLE is a registered trademark of Oracle Corporation. Sun is a registered trademark, and SunView, SunĆ3, SunĆ4, X11/NeWS, SunOS, PCĆNFS, and Open Windows are trademarks of Sun Microsystems. All SPARC trademarks are trademarks or registered trademarks of SPARC International, Inc. SPARCstation is licensed exclusively to Sun Microsystems, Inc. Novell is a registered trademark of Novell, Inc. Macintosh is a trademark of Apple Computer, Inc. Microsoft, MS, MSĆDOS, and Windows are registered trademarks of Microsoft Corporation. All other products or services mentioned herein may be registered trademarks, trademarks, or service marks of their respective manufacturers, companies, or organizations. Part Number: 7806-02 2 Contents About This Manual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Related Documents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Syntax Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Icons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Function Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 11 12 12 12 Function Names and Categories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Function Naming Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Function Categories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Message Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transaction Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Locking Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Security Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Data Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Name Binding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ID Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scanning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Data Manipulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Journal Access Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ordered Access Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XML Processing Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 19 19 20 20 21 22 23 23 24 25 25 25 Returned Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 RHLI Application Basics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Include Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Data Types and Null Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Converting Internal/External Data Formats . . . . . . . . . . . . . . . . . . . Converting Internal to External Format . . . . . . . . . . . . . . . . . . . Converting External to Internal Format . . . . . . . . . . . . . . . . . . . Handling String Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Handling U_AMT data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 35 35 35 36 36 Referencing Database Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Compile-Time Name Binding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Runtime Name Binding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 38 41 ID Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 3 4 Automatic ID Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Partial ID Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Manual ID Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Initializing Mapping Structures . . . . . . . . . . . . . . . . . . . . . . . . . Using Compiler Directives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Using a Custom ID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 45 46 47 48 48 Error Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Status Values of Function Operations . . . . . . . . . . . . . . . . . . . . . . . The Status and Statusl Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . Testing for Partial Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Testing for Complete Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . Retrieving a Status Value Message . . . . . . . . . . . . . . . . . . . . . . . Most Common RHLI Function Status Values . . . . . . . . . . . . . . . . . System Error Log File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Error Log File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 51 51 52 53 54 54 61 61 Transaction Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Type 2 Transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Type 1 Transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Type 0 Transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 62 63 63 Scanning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . An example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 66 Compiling and Loading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Compiling With ucc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Loading Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 70 72 The RHLI Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 About the Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Structure Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 77 The DBAUTH Structure: Schema Information . . . . . . . . . . . . . . . . . . 80 The DBBTREE Structure: B-tree Index Information . . . . . . . . . . . . . . 81 The DBCGRP Structure: Column Group Information . . . . . . . . . . . . . 83 The DBCOLNM Structure: Column Name Information . . . . . . . . . . . 85 The DBHASH Structure: Hash Table Index Information . . . . . . . . . . . 86 The DBINFO Structure: Database Information . . . . . . . . . . . . . . . . . . 88 The DBLINK Structure: Link Information . . . . . . . . . . . . . . . . . . . . . . 90 The DBTABLE Structure: Table Information . . . . . . . . . . . . . . . . . . . . 92 The DBTBLNM Structure: Table Name Information . . . . . . . . . . . . . . 99 The DBUSER Structure: User and Default Schema Information . . . . . 100 The DBVOLM Structure: Volume Information . . . . . . . . . . . . . . . . . . 101 Contents Contents The UACSINF Structure: Access Method Information . . . . . . . . . . . . 105 The UATHINF Structure: Schema Information . . . . . . . . . . . . . . . . . . 107 The UATHMAP Structure: Authorization ID Mapping . . . . . . . . . . . . 108 The UBTINF Structure: B-tree Information . . . . . . . . . . . . . . . . . . . . . 109 The UBTMAP Structure: B-tree ID Mapping . . . . . . . . . . . . . . . . . . . 111 The UCGPINF Structure: Column Group Information . . . . . . . . . . . . 112 The UCOLINF Structure: Column Information . . . . . . . . . . . . . . . . . . 113 The UCOLMAP Structure: Column ID Mapping . . . . . . . . . . . . . . . . . 116 The UDBINF Structure: Database Information . . . . . . . . . . . . . . . . . . 118 The UGRTDS Structure: Grant Privilege Descriptor . . . . . . . . . . . . . . 120 The UHSHMAP Structure: Hash Table ID Mapping . . . . . . . . . . . . . . 124 The UHTINF Structure: Hash Table Information . . . . . . . . . . . . . . . . . 125 The UJRNINF Structure: Journal Information . . . . . . . . . . . . . . . . . . . 127 The ULNKINF Structure: Link Information . . . . . . . . . . . . . . . . . . . . 129 The ULNKMAP Structure: Link ID Mapping . . . . . . . . . . . . . . . . . . . 130 The UNIPCOLL Structure: Column Access Information . . . . . . . . . . . 131 The UNIPEXT Structure: Column Conversion Information . . . . . . . . 140 The UOACOLS Structure: Ordered Access Row Order . . . . . . . . . . . . 141 The URVKDS Structure: Revoke Privilege Descriptor . . . . . . . . . . . . 142 The UTBLINF Structure: Table Information . . . . . . . . . . . . . . . . . . . . 145 The UTBLMAP Structure: Table ID Mapping . . . . . . . . . . . . . . . . . . . 148 The UTXINF Structure: Transaction Information . . . . . . . . . . . . . . . . 150 The UTXOPT Structure: Begin Transaction Options . . . . . . . . . . . . . . 152 The UUSRINF Structure: User Information . . . . . . . . . . . . . . . . . . . . . 154 The UVOLINF Structure: Volume Information . . . . . . . . . . . . . . . . . . 155 The UVOLMAP Structure: Volume ID Mapping . . . . . . . . . . . . . . . . . 158 Function Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 uabttx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 uaddath . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 uaddbt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 uaddcgp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 uaddcnm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 uadddb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 5 6 uaddhsh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 uaddlnk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 uaddprf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178 uaddprm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 uaddprv . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 uaddtbl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 uaddtnm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188 uaddusr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190 uaddvol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192 ualcscn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194 ualctbl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196 uallath . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 uallbt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201 uallcgp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 uallcol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206 uallhsh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208 ualllnk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210 uallpid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212 ualltbl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214 ualltx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216 uallusr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218 uallvol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220 ubegord . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222 ubegscn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225 ubegtx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227 ubndath . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 ubndbt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231 ubndcol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233 ubnddb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 ubndhsh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237 ubndlnk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239 ubndtbl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241 ubndusr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 Contents Contents ubndvol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245 ucbgtx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247 uclritr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249 uclsdb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251 uclsjrn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253 ucmttx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255 ucrypwd, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257 udelrow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259 udrpath . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261 udrpbt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263 udrpcgp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265 udrpcnm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267 udrphsh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269 udrplnk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271 udrpprf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273 udrpprm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275 udrpprv . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277 udrptbl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280 udrptnm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282 udrpusr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284 udrpvol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286 uendord . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288 uendscn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289 uexit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291 uextint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292 ufchath . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295 ufchcnm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297 ufchjrn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300 ufchlnk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302 ufchmsg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305 ufchord . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307 ufchrow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310 ufchscn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312 7 8 ufchtbl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314 ufchtnm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317 ufrescn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319 uinfacs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321 uinfath . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323 uinfbt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326 uinfcgp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329 uinfcol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332 uinfdb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335 uinfhsh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338 uinflck . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341 uinflnk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344 uinfsch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347 uinfscn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349 uinftbl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351 uinftx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355 uinfusr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357 uinfvol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360 uinimsg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363 uinsand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365 uinsbuf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367 uinsor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372 uinsprj . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 uinsrow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377 uinssrt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379 uinstbl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381 uintext . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382 ulckdb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385 ulckrow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387 ulcksch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390 ulcktbl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392 ulogmsg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394 umapath . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396 Contents Contents umapbt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398 umapcol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400 umaphsh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402 umaplnk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404 umaptbl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406 umapvol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408 umodath . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410 umodbt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412 umoddb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414 umodhsh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416 umodlnk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418 umodusr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420 umodvol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422 unumath . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424 unumcgp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426 unumchd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428 unumcol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430 unumprn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432 unumrow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434 unumtbl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436 unumusr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438 unumvol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440 uopndb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442 uopnjrn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447 upkfrow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449 uposord . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451 usetitr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453 uskrord . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456 utrnctbl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458 uulkdb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460 uulkrow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461 uulksch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463 uulktbl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465 9 10 uupdrow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLAddAttribute() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLCreateElement() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLDeleteAttribute() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLDeleteCurrentElement() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLDeleteCurrentElementValue() . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLDestroy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLGetAttributeCount() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLGetAttributeNames() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLGetAttributeValue() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLGetAttributeValues() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLGetCurrentElementName() . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLGetErrorInfo() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLGetValueOfCurrentElement() . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLModifyElementValue() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLMoveToFirstChild() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLMoveToLastChild() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLMoveToNextSibling() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLMoveToParentElement() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLMoveToPreviousSibling() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLMoveToRootElement() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLParse() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . uXMLSetAttributeValue() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467 469 471 473 475 477 479 481 483 485 487 489 491 493 495 497 499 501 503 505 507 509 511 Appendixes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513 Appendix A: Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transaction Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scanning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . An Ordered Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Name Binding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ID Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transaction Journaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DDL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Text Conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XML example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515 516 520 524 528 532 534 536 539 542 Contents About This Manual This document is written for C programmers who want to access a Unify DataServer database through the Relational Host Language Interface (RHLI). The RHLI is a library of more than 100 functions. To access the functions, you must include RHLI header files and use Unify DataServer tools to compile and load the C program that contains the functions. You use the functions to create and change date in a Unify DataServer database. For instance, you can use RHLI functions to perform the following tasks: create a database define database structures add data to a database retrieve data from a database modify data stored in a database delete data stored in a database retrieve information about database objects The first chapter of this manual describes the categories of functions and gives a brief description of each function. The chapter also describes the function naming conventions. The second chapter describes the C structures used with many of the functions. The third chapter describes all of the RHLI functions in alphabetical order. Related Documents This manual describes the functions in reference-style format. For a description of RHLI concepts, see Unify DataServer: Developing a Database. 11 Syntax Conventions Icons This section describes the syntax and naming conventions used to describe the functions described in this manual. bold Indicates a keyword, structure name, or data type that must be entered as shown. italic Indicates an argument or parameter name that you supply. The manual uses the following icons: Tip A Tip contains helpful information. Warning A Warning cautions against actions that could cause data loss or damage to the database. Additional Help Additional Help tells you where to find more information about described topics. Function Format Each function description is divided into several parts: Header Name of the function. Summary statement (following the header bar) A brief statement of the function’s purpose. 12 Syntax Syntax for the function. Arguments Required and optional arguments that are used when calling the function. Description Function usage and any special conditions and notes. Status Values Indicates the status of the function’s operation.statusl is a list of one or more values indicating the status of each object operation corresponding to the object list for the function. Each statusl value can be checked to determine which objects incurred the error. Security Permissions required to execute the function if other than a regular user. Example Example of function usage or a reference to the Examples appendix. See Also Cross references to related information. 13 14 Function Names and Categories 15 Chapter Focus This chapter describes the naming conventions, categories and returned values of the RHLI functions. 16 Function Names and Categories Function Naming Conventions Each RHLI function has a seven-character name. The name is created by combining the operation that the function performs and the object the function operates on. For instance, the ufchrow function fetches a row. The following diagram shows how the function name consists of an operation and an object abbreviation. RHLI Function Naming Conventions PREFIX OPERATION OBJECT for all RHLI the object the ( prefix ) ( indicates ) ( database ) functions RHLI operation operation acts on The following tables list the operations and object abbreviations that compose the RHLI function names. Not every combination of the operations and object abbreviations create a valid RHLI function. Function Names and Categories 17 Operation Abbreviations Used in Function Names Op Description Op abt add alc all beg bnd cbg cir clr cls cmt del drp end exit ext abort add allocate get all begin/initiate name bind commit and begin clear a condition clear a condition close commit delete drop/delete end/terminate program termination convert from external format fetch/reference free/deallocate grt inf ins int fch fre Description grant get information insert convert from internal format lck lock log make a record of map map a compile-time ID to a runtime ID mod modify num retrieve number of opn open pkf primary key fetch pos position rvk revoke set set a condition skr seek to a row ulk unlock upd update Object Abbreviations Used in Function Names Obj Description Obj Description and ath bt cgp chd cnm col db ext scan “and” groups authorizations/schemas B-trees column group child of link relationship column names columns databases convert to external format hash tables convert to internal format interrupt journal links messages or ord prf prj prm prn scan “or” groups ordered access volume preferences scan projection list permissions parent or link relationship privileges database row scan scan sort database table table names transaction users volumes hsh int itr jrn lnk msg 18 prv row scn srt tbl tnm tx usr vol Function Names and Categories Function Categories Functions are categorized into the following eleven categories: message handling transaction management locking management security management data definition name binding ID mapping scanning data manipulation journal access ordered access Message Handling Message functions access the message gathering facilities of the database. ufchmsg uinimsg ulogmsg Transaction Management Transaction management functions begin, commit, and, abort transactions. uabttx ubegtx ucbgtx ucmttx uinftx Function Names and Categories Retrieve the message corresponding to a status value Initialize the messageĆlogging facilities Append a message to the error log Abort a transaction Start a transaction Commit a transaction and start a new transaction Commit a transaction Retrieve transaction information 19 Locking Management Locking management functions manipulate locks on and retrieve lock information for databases, tables, and rows. uinflck uinfsch ulckdb ulckrow ulcksch ulcktbl uulkdb uulkrow uulksch uulktbl Security Management Security management functions grant and revoke permissions and privileges to and from users and schemas. uaddprm uaddath uaddprv uaddusr ucrypwd udrpath udrpprm udrpprv udrpusr ufchath uinfath uinftx uinfusr umodath umodusr 20 Retrieve data lock information Retrieve table definition lock information Lock a database Lock a row Lock a table definition Lock a table Unlock a database Unlock a row Unlock a table definition Unlock a table Grant schema access Add a schema to the database Grant user or schema privileges Add a user to the database and/or set a user's default schema Encrypt a password Drop a schema from the database Revoke schema access Revoke user or schema privileges Drop a user from the database Set a current schema for a user Retrieve schema information Retrieve transaction information Retrieve user information Modify a schema name Modify a user name and default schema Function Names and Categories Data Definition Data definition functions add, drop, and modify databases, volumes, tables, columns, column groups, users, schemas, B-trees, links, and hash tables. uaddath uaddbt uaddcgp uaddcnm uadddb uaddhsh uaddlnk uaddprf uaddtbl uaddtnm uaddvol udrpath udrpbt udrpcgp udrpcnm udrphsh udrplnk udrpprf udrptbl udrptnm udrpvol ufchcnm ufchtnm uinfacs uinfath uinfbt uinfcgp uinfcol uinfdb uinfhsh uinflnk uinftbl uinfvol Add a schema Define a BĆtree Add a column group to a table Add a column name (or synonym) Create a new, empty database Define a hash table Define a link Add a volume preference for a table Add a table Add a table name (or synonym) Add a volume Drop a schema Drop a BĆtree Drop a column group from a table Drop a column name (or synonym) Drop a hash table Drop a link Drop a volume preference for a table Drop a table Drop a table name (or synonym) Drop a volume Fetch column names (or synonyms) Fetch table names (or synonyms) Retrieve access information Retrieve schema information Retrieve BĆtree information Retrieve column group information Retrieve column information Retrieve database information Retrieve hash table information Retrieve link information Retrieve table information Retrieve volume information continued Function Names and Categories 21 Data Definition Functions (continued): umodath Modify a schema name umodbt Modify a BĆtree name umoddb Modify a database umodhsh Modify a hash table name umodlnk Modify a link name umodvol Modify a volume name unumath Retrieve number of schemas unumchd Retrieve number of rows in a table unumcol Retrieve number of columns in a table unumprn Retrieve number of rows in a table unumtbl Retrieve number of tables Name Binding Name binding functions obtain object IDs for database objects. uallath uallbt uallcgp uallcol uallhsh ualllnk uallpid ualltbl ualltx uallusr uallvol ubndath ubndbt ubndcol ubnddb ubndhsh ubndlnk ubndtbl ubndusr ubndvol 22 Retrieve a list of authorization IDs Retrieve a list of BĆtree IDs Retrieve a list of column group IDs Retrieve a list of column IDs Retrieve a list of hash table IDs Retrieve a list of link IDs Retrieve a list of process IDs Retrieve a list of table IDs Retrieve a list of transaction IDs Retrieve a list of user IDs Retrieve a list of volume IDs Bind a schema name to an authorization ID Bind a BĆtree name to a BĆtree ID Bind column name to a column ID Bind a database name to a database ID Bind a hash table name to a hash table ID Bind a link name to a link ID Bind table name to a table ID Bind a user name to a user ID Bind a volume name to a volume ID Function Names and Categories ID Mapping ID mapping functions map compile-time object IDs to runtime object IDs. umapath umapbt umapcol umaphsh umaplnk umaptbl umapvol Scanning Map a compileĆtime BĆtree ID to its runtime ID Map a compileĆtime column ID to its runtime ID Map a compileĆtime hash table ID to its runtime ID Map a compileĆtime link ID to its runtime ID Map a compileĆtime table ID to its runtime ID Map a compileĆtime volume ID to its runtime ID Scanning functions set up and perform a scan on a table. A scan searches for rows that meet the specified criteria, retrieves the desired columns from each row and sorts the rows according to user-defined sort criteria. ualcscn ubegscn uendscn ufchscn ufrescn uinfscn uinsand uinsbuf uinsor uinsprj uinssrt uinstbl Function Names and Categories Map a compileĆtime authorization ID to its runtime ID Allocate a scan information structure Begin a scan End a scan Retrieve the next row from a scan Free or clear a scan information structure Create a list used for scan data retrieval Insert an AND group Allocate a remote scan buffer Insert an expression into an OR group Add a column to the projection list of a scan Insert a column to the sort criteria Specify the target table for a scan 23 Data Manipulation Data manipulation functions are used to access data from the database and perform operations such as adding, updating, deleting or getting information about a database row. The scanning functions are listed on page 23. ualctbl uclritr uclsdb udelrow uexit uextint ufchlnk ufchrow ufchtbl uinsrow uintext unumath unumcgp unumchd unumcol unumprn unumrow unumtbl unumusr unumvol uopndb upkfrow usetitr uupdrow 24 Dynamically allocate a UNIPCOLL structure Clear the process interrupt Close the database Delete a row Terminate a process Convert column values from external format to internal format Fetch the parent row through a link access method Fetch a row Fetch a row sequentially Insert a row into a table Convert column values from internal format to external format Retrieve the number of schemas in the database Retrieve the number of columns in a column group Retrieve the number of child rows for a link Retrieve the number of columns in a table Retrieve the number of parent rows for a link Retrieve the number of rows in a table Retrieve the number of tables in the database Retrieve the number of users in the database Retrieve the number of volumes in the database Open the database for access Fetch a row through its primary key Set the process interrupt Update a row Function Names and Categories Journal Access Functions Journal access functions read the contents of the transaction journal. uclsjrn ufchjrn uopnjrn Close the transaction journal Fetch the next update operation from the transaction journal Open a transaction journal Additional Help For more information about the transaction journal, see Unify DataServer: Managing a Database. Ordered Access Functions Ordered access functions set up and perform an ordered access on a table. ubegord uendord ufchord uposord uskrord XML Processing Functions Begin an ordered access End an ordered access Retrieve a row from an ordered access Position by values to a row within an ordered access Position by row ID to a row within an ordered access The XML processing functions perform operations on XML data.. uXMLAddAttribute Adds a new attribute to an element uXMLCreateElement Creates a new element uXMLDeleteAttribute Deletes an attribute uXMLDeleteCurrentElement Deletes the current element uXMLDeleteCurrentElementValue Deletes the value of an element uXMLDestroy Destroys an XML object uXMLGetAttributeCount Returns the number of attributes uXMLGetAttributeNames Returns the attribute names of an element Function Names and Categories 25 uXMLGetAttributeValue Returns an attribute value uXMLGetAttributeValues Returns the attribute values of an element uXMLGetCurrentElementName Returns the tag name of the current element uXMLGetErrorInfo Returns an error description uXMLGetValueOfCurrentElement Returns the value of the current element uXMLModifyElementValue Modifies the value of an element uXMLMoveToFirstChild Moves the current element pointer to the first child uXMLMoveToLastChild Moves the current element pointer to the last child uXMLMoveToNextSibling Moves the current element pointer to next sibling uXMLMoveToParentElement Moves the current element pointer to the parent uXMLMoveToPreviousSibling Moves the current element pointer to the previous sibling uXMLMoveToRootElement Moves the current element pointer to the root element uXMLParse Parses an XML file uXMLSetAttributeValue Sets an attribute value 26 Function Names and Categories Returned Values Most of the functions return an integer value upon completion. The returned values are defined as: USUCCESS UFAILURE The operation was successful. A failure occurred during the operation; check status value for the error. If the function did not execute correctly, test the value of the status parameter or statusl list to determine which error occurred. All status and statusl status values are of type USTATUS (or long) and are defined in the rhlierr.h header file. When an RHLI function fails (it returns UFAILURE), but the status status value is UENORM, this typically indicates that the function parameters were not correctly defined and that the RHLI function could not set the status status value. Refer to the function description to verify the parameters. Additional Help See the Unify DataServer: Error Messages. Function Names and Categories 27 28 Function Names and Categories RHLI Application Basics 29 Chapter Focus This chapter describes the features and information that help you write an RHLI application. To write an RHLI application, you must have a good understanding of C language concepts, including structure manipulation and memory allocation and deallocation. This chapter contains the following sections: Include files Data types and null values Referencing database objects ID mapping Error handling Transaction kinds: type 0, 1, or 2 Scanning Compiling and loading 30 RHLI Application Basics Include Files Unify DataServer provides two include files that contain definitions for RHLI database options. Include files are sometimes called header files. The include files are: rhli.h Contains definitions for database options, column data types and lengths, lock and transaction types, security privileges and capabilities, return values, and other database specifications. rhlierr.h Contains define statements for all error status values returned by RHLI functions. When an RHLI function returns a value of UFAILURE (or 0), it also returns at least one status value. The status values can be tested to determine why the function failed, and if multiple arguments were passed to the function, which particular argument failed. In addition, the rhlierr.h file contains brief comments describing the meaning of each error status value. This may also be helpful in determining the cause of an error. uXML.h Contains definitons used by the XML processing functions. You include both the rhli.h and rhlierr.h at the top of an RHLI program, by using the #include statement. For example: #include <include/rhli.h> #include <include/rhlierr.h> The include files are contained in a directory named include. If the include directory is not located in the current directory, you have two ways of specifying to the compiler (and preprocessor) where the directory is located: specify the directory path in the –I option on the ucc command specify the relative path in the include statement itself; use the angle brackets (<>) to delimit the include directory and file name RHLI Application Basics 31 Example To specify the relative path name in the include statement, add the full relative path and file name in the include statement. Delimit the path and file name with double quotes (” ”). #include “../../include/rhli.h” #include “../../include/rhlierr.h” The example instructs the preprocessor to search for the include files beginning in the current directory and apply the ../../ relative path name. Additional Help About 32 See –I option on ucc Unify DataServer: Configuration Variable and Utility Reference Status values Unify DataServer: Error Messages RHLI Application Basics Data Types and Null Values When you define a column, you must specify a data type for the column. The column data type specifies how that column’s value is stored in the database and how the column value is displayed. To define a column, use the uaddtbl function. A null value is a data value that is unknown or not applicable. While a null value is a valid value, a column containing a null value is essentially empty. The following table compares Interactive SQL/A data types with the equivalent RHLI representations. SQL/A and RHLI Column Data Types Interactive SQL/A RHLI SMALLINT NUMERIC U_INT INTEGER DECIMAL NUMERIC(5–9) U_HINT HUGE INTEGER U_GINT FLOAT U_FLT AMOUNT U_AMT HUGE AMOUNT U_HAMT CURRENCY U_GAMT DATE U_DATE HUGE DATE U_HDATE TIME U_TIME CHARACTER U_STR table continued on next page RHLI Application Basics 33 SQL/A and RHLI Column Data Types Interactive SQL/A RHLI TEXT U_VTXT BINARY U_VBIN BYTE U_BYT REAL U_REAL DOUBLE PRECISION U_DBL All RHLI data type declarations are found in the rhli.h header file. For complete descriptions of the data types, including their precision and scale, see the DBCOL structure on page 94. Additional Help About 34 See Creating a table and defining column data types The uaddtbl function What the data types are used for The “Working With the Data Types and Null Values” in Unify DataServer: Writing Interactive SQL/A Queries UNIPCOLL and UTP_VDDS The chapter “RHLI Structures” Null values The chapter “Working With the Data Types and Null Values” in Unify DataServer: Writing Interactive SQL/A Queries. RHLI Application Basics Converting Internal/External Data Formats Before you can display values for columns declared as data types U_DATE, U_HDATE and U_TIME, you must first convert the column values from internal to external format: this is because U_DATE, U_TIME and U_HDATE column values are stored in internal format and cannot be displayed as such. The same is true when inserting or updating U_DATE, U_HDATE, and U_TIME column values into the database: you must convert the external (display) value into an internal (database storage) format. The uintext function allows you to convert date and time column values from internal to external format, and the uextint function converts external format date and time column values into internal values. Column values for all data types other than U_DATE, U_HDATE, and U_TIME do not need to be converted; if the values are converted by using the uintext or uextint functions, the values remain unchanged. Converting Internal to External Format If you want to display date or time data that you have retrieved or fetched with an RHLI function it is necessary to convert the data from internal to external format. To convert data, specify the uintext function. To convert data from internal to external format with the uintext function, you must first describe the columns you want to retrieve data from by initializing the internal structure UNIPCOLL. Converting External to Internal Format The uextint function converts column values from external data format to internal format and is used before inserting or updating the column values into the database with the uinsrow or uupdrow functions. Before you can convert the date or time data with the uextint function, you must first describe the external format of the column data you are updating or inserting by initializing the internal structure UNIPEXT. RHLI Application Basics 35 Handling String Data When storing or retrieving string values from a Unify DataServer database with RHLI functions, the RHLI function handles string (or data type U_STR) column values as C language strings. A C language string is defined as a series of ASCII characters followed by a zero-byte terminator. When storing a string value into a column of type U_STR, you must terminate the end of the string with a zero-byte. A zero-byte at the end of a string indicates the end of the string value. However, only the actual string value is stored in the database. Blank characters at the end of the string, as well as the zero-byte itself, are not considered part of the string value and are stripped off before the string value is stored. When a string column value is retrieved from the database, the RHLI row retrieval function adds a zero-byte to terminate the string. A string column of length N requires a maximum of N+1 bytes to return its value. The extra byte is for the terminating zero-byte. Handling U_AMT data Values that are of data type U_AMT are normally values that represent a value with a scale of 2. U_AMT values are stored with a default scale of 2. However, when a U_AMT value is stored into a C variable of type long integer, the scale is lost. If a U_AMT value is stored into a database column, the scale is retained. To print U_AMT values that are stored in C integer values, divide the value by 100 to restore the scale of 2. 36 RHLI Application Basics Referencing Database Objects All database objects in Unify DataServer can be referenced with RHLI functions by either of two methods: by using a unique numeric identifier that identifies the object; called the object ID by using a unique name The following database objects can be assigned names or IDs: volumes databases tables columns column groups hash tables B-trees links schemas users transactions (IDs only) Object names other than schema names must be between 1 and 44 characters. All object names must begin with a letter, and cannot include spaces or any of the following reserved characters: ():.$@#*!’”~‘%^&–+={}[]\|;,<>? Schema names, however, are limited to a maximum of 18 characters in length. Because many database operations performed with RHLI functions require the object ID, you must ensure that the correct object ID is specified. Performance To improve the performance of name binding, use the name cache. The name cache keeps a list of current object IDs in local process memory, where they can be accessed more quickly. The name cache is controlled by the NAMECACHE configuration variable. To retrieve an object ID, you bind the object’s name to its current ID. There are two kinds of name binding: compile-time name binding runtime name binding RHLI Application Basics 37 Compile-Time Name Binding To perform compile-time name binding, the object name is embedded inside the RHLI function call that requires the ID in your RHLI program. At compile-time, the Unify DataServer C preprocessor upp maps the object name to the unique object ID and substitutes in the current ID for that object. The preprocessor generates code used to detect changes made to the database object(s) between the time the program was compiled and eventually run. If any changes have been made in the given schema, an error code is returned and the affected modules must be recompiled. Compile-time name binding can reduce program runtime overhead, but may require additional compilation and loading steps. RHLI also features a “remapping” capability that allows you to remap all compile-time object IDs in the program to their current runtime IDs. To compile-time name binding, you delimit an object name with special characters and keywords. When the preprocessor encounters these special notations in the source code, the object name is replaced with the appropriate object ID. The format of compile-time name binding directives is: $[( class [ : attribute ] )] object_name $ The class indicator specifies the class of object to be addressed. The attribute indicator specifies the object attribute to be substituted into the source code. The attribute indicator is optional; if it is not specified, the object’s ID is substituted into the source code. Each object class has a unique set of attributes. The object name is the name of the object. 38 RHLI Application Basics The following tables lists the class and associated attribute keywords: Compile-Time Name Binding Keywords With this class: You can specify this attribute: UCOL Object is a column column’s ID (default attribute) TID column’s table ID AID column’s authorization (schema) ID DEFID column’s definition ID MAPID column’s ID-mapping ID COLTYP column’s type COLLEN column’s internal length (precision) COLSCL column’s internal length (scale) DSPLEN column’s display length (precision) DSPSCL column’s display length (scale) COLDESC column’s description DSPPICT column’s display picture clause QNAME column’s fully-qualified name UTBL Object is a table table’s ID (default attribute) table’s authorization (schema) ID DEFID table’s definition ID MAPID table’s ID-mapping ID SEGSZ table’s segment size EXPNUM table’s expected number of rows TBLBASE table’s base value TBLDESC table’s description QNAME table’s fully-qualified name UATH Object is an AID CID TID AID authorization (schema) QNAME authorization’s ID (default attribute) authorization’s fully-qualified name continued RHLI Application Basics 39 Compile-Time Name Binding Keywords (continued) With this class: UHSH Object is a hash You can specify this attribute: HID index QNAME Object is a B-tree index UBT BTID QNAME hash index’s ID (default attribute) hash index’s fully-qualified name B-tree index’s ID (default attribute) B-tree index’s fully-qualified name ULNK Object is a link index LID QNAME UVOL Object is a volume VID QNAME UUSR Object is a user UID link index’s ID (default attribute) link index’s fully-qualified name volume’s ID (default attribute) volume’s fully-qualified name user’s ID (default attribute) For example, to retrieve a column’s Table ID, use the following compiler directive: $(UCOL:TID)column_name$ When using compiler directives, the class indicator specifies the type of object. This means that the auxiliary name binding conventions described in the description for ucc are not needed to identify the object. 40 RHLI Application Basics Runtime Name Binding To perform name binding, specify the object name with one of the several RHLI name binding functions. The function returns the appropriate object ID. To perform name binding on a given object, you must meet one of the following criteria: have DBA authority (see the uaddprv function) be located in the same schema in which the database object resides (see the ufchath function) be located in a schema that has been granted schema privileges that allow you to access the desired object To perform runtime name binding, you call one of the RHLI object binding functions in your RHLI program, provide the object name, and it returns the object ID. Then you pass the ID into the RHLI function that requires the object ID. Additional Help For a list of the name binding functions, see the chapter “Function Names and Categories”. RHLI Application Basics 41 ID Mapping ID mapping provides the option to remap object IDs that are used in an application to the correct internal set of object IDs. Each time you refer to a database object in your application programs, Unify DataServer cross-references its own internal ID table and substitutes in the correct object ID. ID mapping alleviates the worry of running your applications on different database environments. The schema or tables you were able to access in a previous application are still accessed by the new application, whether or not the internal structure of the database has been changed. ID mapping is implemented in one of three ways: automatic ID mapping happens implicitly when a database is opened partial ID mapping manual ID mapping happens explicitly during execution of the application Each time a database is opened, you can specify that all the object IDs you use in your application are to be remapped before any database operations begin. A significant benefit of remapping object IDs when you open a database is that the object remapping process is all done at once, saving you processing time later. This feature is particularly helpful for applications that refer to and access many database objects. However, manual ID mapping allows you the flexibility to selectively remapped those objects you feel are critical to the application (or that are subject to change). The processing of ID mapping is then controlled locally, only where it is needed. Manual ID mapping also provides you with the option of choosing a “custom” object ID that never changes and can be used in your application programs without having to worry about what the internal object ID really is. The custom ID is remapped by Unify DataServer to the correct internal ID each time your custom ID is used to reference an object. 42 RHLI Application Basics In either case, once an object ID has been remapping, it stays set until the database is closed, and is only available to the application program that initiated the remapping (and opened the database). Similar to name binding, once an object ID has been remapped, Unify DataServer assumes you will be accessing the object soon and consequently places a definition lock on the object. Object definition locks prevent design changes to the locked objects until the calling program closes the database and unlocks the objects. This helps make ID mapping consistent and stable for the duration of the database session. Additional Help For more information about name binding functions, see the “Data Access and Manipulation” chapter. Automatic ID Mapping Automatic ID mapping remaps all the object ID references you use in your application programs to runtime IDs when the application opens a database. To implement automatic ID mapping, you must: set the DB_REMAP option each time you open a database in your application The DB_REMAP option specifies that all object IDs referenced in your application programs are to be remapped to their corresponding runtime object IDs before any processing is allowed on the database. use compiler directives in place of each object ID reference in your application programs compile and load your application programs by using ucc and uld Example The following example shows how to set the DB_REMAP option with the uopndb function. /* define uopndb() parameters */ UDBID dbid; char * dbname = ”/usr/db/parts.db”; USTATUS status; RHLI Application Basics 43 if ( uopndb(dbname, DB_REMAP, &dbid, &status) != USUCCESS ) { printf(”Unable to open the database %s\n”, dbname); uexit(1); } A compiler directive is special syntax that you embed into your application program(s) wherever a reference to an object ID is made. Compiler directives are processed by the Unify DataServer compiler, ucc. When you compile your application programs by using ucc, all compiler directives embedded in the program code are substituted with a compile-time object ID. When your application programs are executed, the compile-time object IDs that were provided by ucc are then substituted (or remapped) to the correct runtime object IDs for that particular database. Compiler directives for automatic ID mapping must always be delimited with dollar signs ($). Additional Help For more information about compiling programs with ucc, see the “Compiling and Loading” chapter. The following table shows the compiler directive syntax for specifying a compile-time object ID. When specifying an 44 ID for a The compiler directive syntax for the compile-time ID is Schema $ [ (UATH:AID) ] schema_name $ B-Tree $ [ (UBT:BTID) ] B-Tree_name $ Column $ [ (UCOL:CID) ] column_name $ Hash table $ [ (UHSH:HID) ] hash_table_name $ Link $ [ (ULNK:LID) ] link_name $ Table $ [ (UTBL:TID) ] table_name $ Volume $ [ (UBOL:VID) ] volume_name $ RHLI Application Basics Example To refer to a schema named admin, the compiler directive for the compile-time object ID would be: $(UATH:AID) admin$ Or, to refer to the table misc contained in schema admin, the compiler directive for the compile-time object ID would be: $(UTBL:TID) admin.misc$ This example shows the use of a compiler directive that specifies the table commissions contained in schema sales. /* Access table “commissions” using a compiler directive */ if ( ufchrow(tx, $(UTBL:TID)sales.commissions$, USLCK, rid, &unipcol, statusl, &status) != USUCCESS ) { printf(”Cannot not access commissions (Table ID=%ld)\n”, $(UTBL:TID)sales.commissions$); uexit(99); } Partial ID Mapping Unify DataServer provides for automatic mapping of compile-time database object IDs referenced in a program to current runtime database object IDs. The implementation of partial ID mapping addresses the problem of assignments of object IDs to a schema change over time or that differ between application sites. Automatic ID mapping is accomplished by using the DB_REMAP option when opening the database with the RHLI function uopndb(). Upon opening the database, all database objects that are referenced by a program are remapped. Automatic ID mapping requires that all of the database objects referenced by the program be present in the database. Otherwise, the remapping, and therefore the database open, will fail. Partial ID mapping eliminates this restriction and defers an error condition until an unmapped ID is actually used in a later RHLI function. Partial ID mapping enables delivery of a schema that supports only a selected set of subsystems of an application. At runtime, the application is responsible for disabling access to subsystems that will not be supported. RHLI Application Basics 45 To implement partial ID mapping: set the DB_PARTMAP option OR’d with the DB_REMAP option when using the uopndb() call use special syntax compiler directives in place of each object ID referenced in the application program compile and load the application program using ucc and uld Manual ID Mapping Manual ID mapping allows you to selectively remapped compile-time object IDs to their correct internal IDs. Unlike automatic ID mapping, manual ID mapping is performed when you reference the database object, not when the database is first opened. To implement manual ID mapping, you must: set the DB_NOLOCK option each time you open a database in your application The DB_NOLOCK option specifies that no locking is to be performed when the database is first opened. This option is typically used when you want to control which database objects to lock. initialize a mapping structure and call the associated mapping function for each object your application references compile and load your application programs by using ucc and uld Remember that when you remapping an object ID, a definition lock is implicitly placed on that object. Therefore, by remapping object IDs, you are in effect controlling which objects are protected from other transaction operations. By mapping and locking IDs selectively in this manner a minimum of database objects are tied up at any one time. This allows for greater concurrency since only the objects your application program uses are locked. To use manual ID mapping, you must use the mapping functions. Each function has an associated structure. The structure must be initialized before you call a mapping function. 46 RHLI Application Basics Initializing Mapping Structures Mapping structures are initialized manually by using a mapping macro. A mapping macro is a special command that is interpreted by the Unify DataServer compiler, ucc, and initializes the specified mapping structure information. The following table lists the syntax for mapping macros. Mapping Macro Syntax for Manual ID Mapping To initialize this structure: The mapping macro syntax is: UATHMAP INIATHMAP( athnm, aid, –>UATHMAP); UBTMAP INIBTMAP( btn, bid, –>UBTMAP); UCOLMAP INICOLMAP( colnm, cid, –>UCOLMAP) UHSHMAP INIHSHMAP( hshnm, hid, –>UHSHMAP) ULNKMAP INILNKMAP(lnknm, tid, –>ULNKMAP); UTBLMAP INITBLMAP( tblnm, tid, –>UTBLMAP); UVOLMAP INIVOLMAP( volnm, vid, –>UVOLMAP); where: athnm btcnm colnm hshnm lnknm tblnm volnm = schema name = B-tree name = column name = hash table name = link name = table name = volume name aid bid cid hid lid tid vid = Authorization ID = B-tree ID = Column ID = Hash Table ID = Link ID = Table ID = Volume ID The first argument in a mapping macro is the object name: the name of object is entered (without quotes). The second argument in a mapping macro is the object ID: you can use a compiler directive or a custom ID. The third argument in a mapping macro is a pointer to the structure that describes the object to be initialized. RHLI Application Basics 47 Using Compiler Directives The following example shows the manual remapping of two B-tree IDs, commissions and revenue, by using compiler directives. #define NBTMAP 2 /* define umapbt() parameters */ UDBID dbid; USTATUS statusl[NBTMAP]; USTATUS status; UBTMAP btmapl[NBTMAP]; /* Initialize UBTMAP structures */ INIBTMAP(sales.commissions, $(UBT:BTID)sales.commissions$, &btmapl[0]); INIBTMAP(sales.revenue, $(UBT:BTID)sales.revenue$, &btmapl[1]); /* Re-map B-trees */ if ( umapbt(dbid, NBTMAP, btmapl, statusl, &status) != USUCCESS ) { printf(”Unable to map B–tree ID info: status = %ld\n”, status); uexit(101); } When you compile this program (using ucc), the compiler directives contained in the mapping macros are substituted with B-tree IDs that are valid at compile-time. The compile-time B-tree ID and the B-tree name are then used to initialize the UBTMAP structure. When you execute the program, the function umapbt function uses the values stored in the structure UBTMAP to find the current, runtime B-tree IDs. Using a Custom ID A custom ID is a unique integer number that you choose to refer to a database object. Instead of using a compiler directive, you simply use this “customized” ID number each time you reference an object in the database. For example, to initialize the schema admin with a custom ID of 10001, you simply substitute the aid argument in the mapping macro INIATHMAP with the value 10001: INIATHMAP(admin, 10001, &athmapl); 48 RHLI Application Basics After the umapath function is executed, you can then refer to the schema admin (with RHLI functions) by using the Authorization ID value 10001. And the same example, using a compiler directive: INIATHMAP(admin, $(UATH:AID)admin$, &athmapl); Example The following examples shows the remapping of the table manf (found in the PUBLIC schema), using a custom Table ID. #define COLNO 18 /* define ’umaptbl()’ arguments */ UDBID dbid; UTBLMAP utblmap; USTATUS statusl[COLNO]; USTATUS status; /* Initialize UTBLMAP structure */ INITBLMAP(PUBLIC.manf, 1124, &utblmap); /* Re-map table manf */ if ( umaptbl(dbid, 1, &utblmap, statusl, &status) != USUCCESS) { printf(”Mapping of manf failed: status=%ld\n”, status); uexit(1); } /* Access table manf using custom Table ID */ if (ufchrow(tx, 1124, USLCK, rid, &unipcol, statusl, &status)!= USUCCESS ) { printf(”Data retrieval failed on Table ID 1124\n”); uexit(99); } Notice the initialization of the compile-time Table ID with the value 1124. This value is then used to reference the table in the ufchrow function. Consider these two things if you initialize a compile-time object ID by using custom initialization: 1. Make sure that the object ID you pick is unique. That is, no two “custom” object IDs should be the same. 2. Pick a non-zero, positive integer that falls within the acceptable range for the type of the object ID. Hopefully, that should be enough to ensure uniqueness Refer to your system manual to verify the range for a long integer on your machine. RHLI Application Basics 49 Additional Help For more examples of how to use manual ID mapping, see the description for each of the mapping functions. 50 RHLI Application Basics Error Handling Errors that occur during execution of RHLI application programs are located in three areas: the function itself returns a status value indicating whether or not the function executed successfully after each RHLI function executes, a status value is returned; some functions return a list of status values, each status value corresponds to an object the function operates on each system error that occurs is logged to the system file errlog Each of these is described in this chapter followed by a description of signal handling. Status Values of Function Operations All RHLI functions return a status value that indicates whether the function operation was a success or a failure. The return values are: The status of a function operation can be verified by checking the function return value. USUCCESS The operation was successful. UFAILURE A failure occurred during the operation. If the return value is equal to USUCCESS (or 1), the function operation was totally successful. However, a return value of UFAILURE (or 0) indicates that either the operation failed partially or completely. See Status and Statusl Values on page 51. The Status and Statusl Parameters If an RHLI function does not execute correctly, the error that occurred is returned in the status or statusl parameter. If the function you are executing allows you to specify a list of one or more objects to be operated on, one or more of the operations involving these objects may fail: this is considered a partial failure. RHLI Application Basics 51 Testing for Partial Failure A partial failure is indicated when the RHLI function returns the value UEPART into the status parameter. One (or more) of the objects listed in the object list was operated on and an error occurred. After each function operation, the statusl list contains a corresponding status value for each object operation listed in the object list. If a partial failure occurs, simply check the status value for each element of statusl to diagnose which object operation failed. For example, if the first object operation in the object list failed, the first element in statusl contains the appropriate error status value; if the third object operation failed, the third element value of statusl contains the corresponding error status value (and so on). If an object operation was successful, however, the corresponding statusl entry contains the status value UENORM. All status and statusl status values are of type USTATUS (or long). Example The following is a sample function (named reterr) that can be used if a partial failure occurs: the reterr function checks each entry in statusl and returns a status value for the first error it finds. #include <include/rhli.h> #include <include/rhlierr.h> USTATUS reterr(status, nstat, statusl) USTATUS status; /* status value */ int nstat; /* # of entries in statusl */ USTATUS statusl[]; /* status list */ { int indx; /* temp index variable */ /* If the status value is set to UEPART, check each statusl */ /* entry for an error status value. */ if ( status == UEPART ) { for ( indx = 0; indx < nstat; ++indx ) { if ( ((status = *(statusl + indx)) != UENORM) && (status != UENOWRK) ) { break; } } } return(status); /* return the first error status value */ } 52 RHLI Application Basics After verifying that status is equal to UEPART, the reterr function checks for each element in statusl for an error status: UENORM indicates success, and UENOWRK indicates that no work was done on this object due to an error elsewhere in the operation. Once you know the reason for the partial failure, you have the option of redoing the entire operation (after the current transaction has been aborted or rolled back), or, to just redo the operation(s) for the object(s) that failed after you correct any programming errors in the application that caused the error. Testing for Complete Failure If a function operation fails completely, a non-UEPART status value is returned to the function parameter status. For example, if the umodath function fails, the status parameter is set to one of the following values: UEILTX Invalid Transaction ID UEDBCLS Database is closed UENOAID Invalid Authorization ID UENLONG Name too long UEBDNM Illegal characters in the user name If status is set to UEPART, then a partial failure occurred. Additional Help About RHLI Application Basics See All status and statusl status values Unify DataServer: Error Messages, or the rhlierr.h header file Partial failures “Testing for Partial Failure” on page 52 53 Retrieving a Status Value Message Once you have a status value, you can check what the error code means by either: checking the Unify DataServer: Error Messages manual checking the rhlierr.h header file retrieving (and displaying) the error message The first and second options assume your application programs print (or store) the status values so that you can look them up. The third option, however, allows you to retrieve (and if you wish, to display) the error message text for the status value. This is accomplished by using the ufchmsg function. Additional Help For more information about how to retrieve text for status value error messages, see the ufchmsg function. Most Common RHLI Function Status Values This section describes the most commonly occurring status values returned from each RHLI function. You can use this section to determine the conditions that you can test for when a function is used in your application. The following status codes are the most commonly occurring codes for all of the functions: 54 UENORM The function completed successfully; no error condition is returned. UEPART Errors are returned through the status list. UENOMEM More (dynamically allocated) memory is needed. UELMOUT A locking conflict exists. UEINTRPT An interrupt was received. UESGTERM A SIGTERM interrupt was received. RHLI Application Basics UEDBDOWN The database is being shut down, gracefully close database. UEILTX An illegal transaction id was encountered. UENOPRV Current user does not have privilege required for operation. UENOPRIV Same as UENOPRV. UEEXCPMP The database is opened with conditional ID remapping (DB_REMAP|DB_CONDMAP). It indicates that a compile-time ID used in the function failed to remap at the time uopndb() is called. (Note that conditional mapping defers reporting the error until the ID is used.) For more information about the status codes, see Unify DataServer: Error Messages. The following table lists status codes that are associated with specific functions. These status codes occur most frequently with the associated function, in addition to the list of commonly occurring functions listed above. Function Most Frequent Status Codes uabttx UENOABRT uaddath UENOXID uaddbt UEBDNAM UEDUPNM UEMXBKL UENMLONG UECOLTP uaddcgp and uaddcnm UEBDNAM UEDUPNM UENMLONG UENMREQ uadddb UEBDNAM UEDUPNM UEDLONG UENMLONG UERSVNM uaddhsh UEBDNAM UEDUPNM UENMLONG uaddlnk UEBDNAM UEDUPNM UENMLONG table continued on next page RHLI Application Basics 55 Function Most Frequent Status Codes uaddprf UEDUPPRF uaddprm UEDUPK uaddprv UENOGRT uaddtbl UEBDNAM UEDUPNM UENOCOL UENMLONG UENOVID UEDLONG uaddtnm UEBDNAM UEDUPNM UENMLONG UENMREQ uaddusr UEBDNAM UEDUPNM UEDUPUSR UENMLONG UENMREQ uaddvol UEBDNAM UEDUPNM UEEXIST UENMLONG UENMREQ ubegord UENOSAM UENOMEM UEMXSCN ubegtx UEUMXTX ubndath UENMPRS UENONAME ubndbt UENMPRS UEAMBIG UENONAME ubndcol UENMPRS UEAMBIG UENONAME ubnddb UENONAME ubndhsh UENMPRS UEAMBIG UENONAME ubndlnk UENMPRS UEAMBIG UENONAME ubndtbl UENMPRS UENONAME ubndusr UENONAME ubndvol UENONAME table continued on next page 56 RHLI Application Basics Function Most Frequent Status Codes uclsdb UETXACTV udelrow UEREF UEBIGLG udrpath UENOAID UENONAME udrpbt UENOBT UENONAME udrpcgp UENONAME UEDRCGP UENOUNQ udrpcnm UENOTID UENONAME udrphsh UENOHID UENONAME UENOCID UENOHSH udrplnk UENOLID UENONAME UEREF UENOCID UENOLNK udrpprf UENOVID UENOPRF udrpprm UENOUID UENOPRM UENOAID udrpprv UEUKNOBJ UENOOBJ UEUNPRV udrptbl UENOTID UENONAME UEVIEW UENMPRS udrptnm UENOTID UEBDNAM UEVIEW UENONAME udrpusr UENOUID UEDBOWN udrpvol UENOVID UEVLACT UEBDVOL UENONAME uextint UESTRLNG UECONV ufchath UENOAID UENOPRM ufchcnm UENOCID UENONAME ufchlnk UENOCID UENOREF UENONAME UENOLID table continued on next page RHLI Application Basics 57 Function Most Frequent Status Codes ufchord UEBDATYP UENOROW UENOTID UENOCOL ufchrow UENOTID UENOCOL UENOROW ufchscn UEBDSZR UEVDBUF UEBDBNX UENOWRK UEBDFNX UEBDFTP UEEOSCN UECONV ufchtbl UECONV UEVDBUF UENOWRK UEEOSCN UEBDFTP uinfath UENOWRK UENOAID uinfbt UENOBID UENOWRK uinfcgp UENOWRK uinfcol UENOWRK uinfdb UENOWRK uinflck UENOLCK uinflnk UEPART uinfsch UENOLCK uinftbl UEPART uinftx UEPART uinfusr UENOWRK umapath UEISMAP UENONAME UENOMAP UENMPRS UEMNMAP umapbt UEISMAP UENONAME UEMNMAP UENMPRS UEAMBIG UENOMAP UEPART table continued on next page 58 RHLI Application Basics Function Most Frequent Status Codes umapcol UEISMAP UENONAME UEMNMAP UENOMAP UENMPRS UEAMBIG UEBDMAP umaphsh UEISMAP UENONAME UEMNMAP UENMPRS UEAMBIG UENOMAP umaplnk UEISMAP UENONAME UEMNMAP UENMPRS UEAMBIG UENOMAP umaptbl UEISMAP UENONAME UEBDMAP UENMPRS UEMNMAP UENOMAP umapvol UEISMAP UEMNMAP UENONAME UENOMAP umodath UENMLONG UEDUPNM UEBDNAM umodbt UENMREQ UEBDNAM UENMLONG UEDUPNM umodhsh UENMREQ UEBDNAM UENMLONG UEDUPNM umodlnk UENMREQ UEBDNAM UENMLONG UEDUPNM umodusr UENMREQ UEBDNAM UENMLONG UEDUPNM umodvol UENMREQ UEBDNAM UENMLONG UEDUPNM unumcgp UENOCID unumchd UENOTID UENOVID unumcol UENOTID UENOROW table continued on next page RHLI Application Basics 59 60 Function Most Frequent Status Codes unumprn UENOTID UENOVID unumrow UENOTID unumtbl UENOAID uopndb UEDBOPN UENDENT UEINVAL uopnjrn UENOENT UEINBHDR upkfrow UENOTID UENOCID UENOROW UENOKEY UENOCGP uposord UEEOSCN usetitr UEINVAL uskrord UEBDATYP UEEOSCN uulkdb UENOLCK uulkrow UENOLCK UENORID UENOTID uulksch UENOTID UENOLCK uulktbl UENOTID UENOLCK uupdrow UENOTID UENORID uXML* UEDOMEX UEGENEX UENTELEM UENOBIN UEFILNF UENONDVL UENOPELM UENDNTFD UEERRXML UENOCHLD UENONCOL UEBADUN UEXMLEX UENOSIB XMNOATTR UENOROW UENORID RHLI Application Basics System Error Log File Located in every application database directory, the system error log file, dbname.err, is used by Unify DataServer to log system status and error messages. Status and error messages are appended to dbname.err each time a major status event or system error occurs and can be checked by using any standard Unix system editor, such as vi. A dbname.err file is implicitly created each time a new database is created with the uadddb function. While you can delete the contents of the errlog file, do not remove the file. Error Log File All errors and status conditions that occur during the execution of an RHLI application program that accesses a Unify DataServer database are written to the system error log file, errlog. Errors and status conditions are logged to errlog in a format similar to this example: Error Log File **************************Wed Sep 7 09:04:56 1988 Program: creatdb Calling function: bterrlg.c Offending function: bterrlg.c Status: 0 Errno: 0 Notes: errlog created*********Wed Sep 7 09:05:05 1988 Program: creatdb Calling function: bterrlg.c Offending function: bterrlg.c Status: 223 Errno: 4096 Notes: writing normal sync record*********Wed Sep 7 09:13:50 1988 Program: fmdmn Calling function: wrnsnlg.c Offending function: wrnsnlg.c Status: 223 Errno: 8192 Notes: writing normal sync record An errlog file is created each time a new database is created and is located in your local database directory. Every database directory must contain an errlog file. To create a database, use the uadddb function. RHLI Application Basics 61 Transaction Types RHLI features three types of transactions, each of which enforces a different level of concurrency and database integrity: Type 2 Type 1 Type 0 Type 2 transactions are the most restrictive, while Type 0 transactions are the most lenient in terms of concurrency and preserving maximum database integrity. This kind of flexibility allows you to tailor your applications according to the sensitivity of certain database operations. You specify the transaction type when you begin a transaction with the ubegtx or ucbgtx functions. Type 2 Transactions A Type 2 transaction strictly enforces any shared or exclusive locks that the transaction acquires while performing its operations. All locks remain in place until the transaction commits or aborts. This type of transaction limits database concurrency but improves data integrity and provides the highest degree of read reproducability. Type 2 transactions are especially useful for those operations where performance is less important but data accuracy and consistency is critical. For example, a banking transaction dealing with monetary amounts would use a Type 2 transaction to perform database updates. 62 RHLI Application Basics Type 1 Transactions A Type 1 transaction enforces all locks acquired by the transaction, except that shared locks can be released successfully at any time during the transaction. This improves database concurrency but reduces read reproducability of data. Type 1 transactions are useful in cases where good performance is required but data accuracy is less important. For example, a Type 1 transaction would be useful in an inventory system, where updates to the inventory tables must be made quickly and checks for inventory figures can be approximate. Type 0 Transactions A Type 0 transaction is able to read uncommitted data locked by other processes or transactions. This type of transaction cannot be used to perform any update operations and should be used with caution. Type 0 transactions are normally used when high performance is required but the accuracy of the data is not as important. For example, a Type 0 transaction might be used by a reporting application that generates statistical reports used to portray data at a given moment in time (sometimes called “snapshot” reporting). RHLI Application Basics 63 Scanning Scanning is a method for retrieving one or more rows from a table. You determine which rows are retrieved by specifying selection criteria. Unify DataServer examines the existing access methods for the table and determines the best possible access method for retrieving the rows. To perform a scan, follow these general steps. 1. Allocate a scan information structure by calling the ualcscn function. This structure eventually will contain information about the table column values that are compared by using the specified selection criteria. The ualcscn function returns a pointer to the Scan Information Structure. This structure is referenced during the remaining scan operations. 2. Indicate which table the scan is to be performed on by calling the uinstbl function. 3. Indicate which columns are to be projected by calling the uinsprj function. The function uinsprj function must be called once for each column that is to be included in the projection. The sequence of columns that are added to the projection list determines the order that the columns appear in the rows returned from the scan. 4. Allocate a criteria structure by calling the uinsand function. 5. Specify individual selection criteria by calling the uinsor function. Selection criteria can include column value matches or comparisons. Multiple criteria for a single column are grouped together by using “or”, to create one or more “or-groups”. If there is more than one column being matched or compared, the resulting “or-groups” are grouped together with “and”. 6. If you want the retrieved rows to be ordered, specify the sort keys and sort order by calling the uinssrt function. 64 RHLI Application Basics 7. If you want to get DBVAL column structures, call the uinfscn function. 8. Begin the scan by calling the ubegscn function. The ubegscn function uses the scan information in the Scan Information Structure to determine the best method for retrieving the rows: it selects the rows that meet the criteria specified and returns a Scan ID that can be used to identify the row(s) selected by the scan. When calling the ubegscn function, you can specify whether it should lock the rows returned with the scan by using a shared or an exclusive lock, or if it should not lock them. 9. Fetch rows from the scan by calling the ufchscn function. Before rows can be retrieved from the scan, you must allocate space for the column values that are returned with ufchscn. This can be done in one of two ways: by calling the uinfscn function to generate a database value list based on the projection list and implicitly allocate space by manually constructing a database value list that matches the type and number of columns in the projection list, and allocating space in your program The uinfscn function uses the scan projection list information stored in the Scan Information Structure to generate a database value list. The database value list contains the column values and any column options (such as whether the database value is null, or whether it should be ignored when converting from internal to external format and vice versa). The uinfscn function uses the information in the database value list to allocate space to receive the column values when retrieving rows with the ufchscn function. Rows are then locked and retrieved from the scan by using the ufchscn function. 10. End the scan by calling the uendscn function. 11. Deallocate the scan information structure by using the ufrescn function. In addition, you can specify that the structure only be emptied, so that you can respecify all new scan criteria without having to reallocate space by using the ualcscn function. RHLI Application Basics 65 The steps are indicated in the following example. Additional Help About See The syntax of the functions An example Function reference for each function The example in this section performs a scan on a table named table1. The following diagram indicates the table, column names, data types, and values used in the scan. The table1 Table A B C 1 0 2 2 2 2 NORTH NORTH WEST EAST NORTH NORTH 3.00000 3.00000 3.00000 3.00000 4.00000 8.00000 CHAR(20) NUMERIC(9) FLOAT The equivalent SQL/A query that the scan performs is: SELECT A, B, C FROM table1 WHERE A BETWEEN 1 and 100 AND (B = ”NORTH” or B = ”WEST”) AND C < 7.824; This query has four conditions; each condition requires a separate call to the uinsor function. These are the conditions: Column A is between the values 1 and 100. Column B is equal to NORTH. Column B is equal to WEST. Column C is less than 7.824. 66 RHLI Application Basics #include <stdio.h> #include <include/rhli.h> #include <include/rhlierr.h> #define N_TABLE1_COL 3 UDBID dbid; UTXID txid; USTATUS status[1]; USTATUS statusl[N_TABLE1_COL]; UTXOPT txopt; char *scaninf; int scanid; int prevlck; URID rid; UDBVAL search_dbv[5]; UDBVAL *prjvall; UDBVAL *nulldbv; UTID tid = $table1.$; UCID cidl[] = {$table1.A$,$table2.B$,$table3.C$}; long a1,a2; char b1[20],b2[20]; double c1; /* ** INIT_DBVALL ** ** Since these buffers probably only need to be set up once, ** no matter how often you do the query, it is best to ** initialize them in some function once, and if possible ** call it only once. This will save runtime CPU time. */ init_dbvall() { search_dbv[0].unip.hintp = &a1; search_dbv[1].unip.hintp = &a2; search_dbv[2].unip.strp = b1; search_dbv[3].unip.strp = b2; search_dbv[4].unip.fltp = &c1; Step 1 Step 2 Step 3 nulldbv = (UDBVAL *) 0L; } main () { init_dbvall(); txopt.txknd = UTXTYPE2; txopt.waitflg = UWAIT; if (uopndb (””,DB_NORMAL,&dbid,status) != USUCCESS) fatal (”uopndb”); if (ubegtx (dbid,UROOTTXID,&txid,&txopt,status) != USUCCESS) fatal (”ubegtx”); if (ualcscn (txid,0,0,0,&scaninf,status) != USUCCESS) fatal (”ualcscn”); if (uinstbl (txid,scaninf,tid,status) != USUCCESS) RHLI Application Basics 67 fatal (”uinstbl”); if (uinsprj (txid,scaninf,3,cidl,statusl,status) != USUCCESS) fatal (”uinsprj”); init_criterea(); Step 4 Step 5 if (uinsand (txid,scaninf,4,status) != USUCCESS) fatal (”uinsand”); if (uinsor (txid,scaninf,$table1.A$, URNGCC,&search_dbv[0],&search_dbv[1],status) != USUCCESS) fatal (”uinsor.1”); if (uinsor (txid,scaninf,$B$, ULIKE,&search_dbv[2],nulldbv,status) !=USUCCESS) fatal (”uinsor.1”); if (uinsor (txid,scaninf,$table2.B$, ULIKE,&search_dbv[3],nulldbv,status) != USUCCESS) fatal (”uinsor.2”); if (uinsor (txid,scaninf,$table3.C$, ULETST,&search_dbv[4],nulldbv,status) != USUCCESS) fatal (”uinsor.3”); Step 7 Step 8 Step 9 if (uinfscn (txid,scaninf,&prjvall,status) != USUCCESS) fatal (”uinfscn”); if (ubegscn (txid,USLCK,scaninf,&scanid,status) != USUCCESS) fatal (”ubegscn”); while (ufchscn(scanid,&rid,prjvall,&prevlck,statusl,status) == USUCCESS) { fprintf (stdout,”%d | %s | %f\n”, *(prjvall[0].unip.hintp), prjvall[1].unip.strp, *(prjvall[2].unip.fltp)); } if (*status != UEEOSCN) fatal (”ufchscn”); Step 10 Step 11 if (uendscn (scanid,status) != USUCCESS) fatal (”uendscn”); if (ufrescn (txid,&scaninf,DB_RESET,status) != USUCCESS) fatal (”ufrescn”); if (ucmttx (txid,status) != USUCCESS) fatal (”ucmttx”); if (uclsdb (dbid,DB_NORMAL,status) != USUCCESS) fatal (”uclsdb”); uexit (0); 68 /* should bail out here */ RHLI Application Basics fatal (”uexit”); /* otherwise bail out from fatal */ } RHLI Application Basics 69 Compiling and Loading RHLI programs are compiled and loaded in two steps: Step 1 Compile the program source files by using the Unify DataServer C compiler ucc. Step 2 Load the program object file(s) by using the Unify DataServer C program loader uld. Before compiling and loading your RHLI program, make sure that you have: included the proper header file(s) in your program used the correct C data types to maintain compatibility with the Unify DataServer database types named your program by using a file name of 11 characters or less, beginning with a character Compiling With ucc The Unify DataServer C compiler, ucc, compiles an RHLI source file and creates an object file. The ucc compiler uses the preprocessor upp and the system C compiler cc. The upp preprocessor identifies any database object names in the source file that need ID mapping or name binding performed. The preprocessor substitutes appropriate object IDs for objects referenced with name binding or ID mapping constructs. The upp preprocessor produces two intermediate files A file that contains the substituted object IDs for all object references. The file contains the .u suffix. A preprocessed file used as input to the cc compiler. The file contains the .p suffix. The .u file identifies the objects referenced in the source file and contains the corresponding definition ID assigned by the database to each table or column. 70 RHLI Application Basics The table and column definition IDs are used to identify the given table or column at that point in time. If any updates are later made to the table or column in the database, a new definition ID is generated and stored in the database. The definition IDs stored in the .u files are compared to current IDs during runtime to determine whether changes have occurred to the database that affect the accuracy of the name binding that was performed earlier. After the preprocessing phase, ucc renames the .p file by using the format Ufile.c, where the original source file name is prefixed with “U” and concluded with “.c”. ucc then passes the Ufile.c files, along with any other arguments entered in the ucc command line, to the UNIX or host operating system C language compiler, cc. The cc compiler is the UNIX or host operating system C language compiler that compiles the preprocessed source file and generates an object file. The object file contains the suffix .o. RHLI Application Basics 71 The Compilation Process Source File file.c" Preprocessor upp Intermediate Data File file.p" C Compiler (This file is created only if the program references database tables and columns. It is passed ontoăthe loader, uld.) Intermediate Data File file.u" (optional) 1 C Compiler cc ucc Intermediate Source File Ufile.c" 1 Object File Ufile.o" C Compiler ucc Object File file.o" The Loading Process The Unify DataServer program loader, uld, verifies that changes have not occurred to the referenced database objects since the object files were compiled, then loads the files by using the UNIX or host operating system link loader, ld. ld combines the object file(s) and any archive files specified, resolves external references, searches libraries specified, and generates the program executable. 72 RHLI Application Basics RHLI program object files are loaded by running uld from the operating system shell, passing the RHLI program object files (.o) as arguments, as well as any archive files and their corresponding .u files. During loading, intermediate data files are created to aid in detecting database changes between compile-time, load-time, and runtime. The uld utility loads an object file created by the ucc utility. The object files (.o) and any archive files (.a) are passed as arguments to uld. If the specified archive files depended on compile-time name binding, you must specify the corresponding .u file name along with the archive file name in the uld command line. If any of the object files performed name-binding at compile-time, there must be a corresponding .u file existing in the current directory. If one does not exist, then the source file(s) must be recompiled by using ucc (or with upp and cc). For each .o file passed in the command line, uld searches for the corresponding .u file. For those object files or archive files with a corresponding .u file, uld interacts with the database to retrieve information about the tables and columns referenced. If there are any discrepancies between the information in the .u files and the database objects in the database, then the objects referenced have been updated since compilation. In this event, uld fails and returns an error message stating which object files contained the invalid references. The invalid object files must then be recreated by recompiling the corresponding source files. If no changes have been made to the referenced tables and columns since compilation, uld generates an intermediate source code file named Ufile_h.c which contains information that is used when the executable is run and the database is opened. The intermediate source file Ufile_h.c is compiled by uld and the resulting object file is then passed on to the UNIX or host operating system C loader ld. ld combines the object files, resolves external references, searches libraries and produces an executable file. For more information about ld, see the manual for your UNIX or host operating system C loader. RHLI Application Basics 73 The following diagrams shows the loading process. Events That Take Place During Loading Object File file.o" Archive File file.a" Intermediate Data File file.u" (optional) DataServer C PROGRAM LOADER uld Intermediate Data File Ufile_h.c" UNIX C PROGRAM LOADER ld RHLI Program Executable file" 74 RHLI Application Basics The RHLI Structures 75 Chapter Focus This chapter describes the structures used with many of the RHLI functions. This chapter contains the following sections. About the Structures Structure Index Structure Descriptions 76 The RHLI Structures About the Structures The structures used by the functions contain information specific to the operation that the function performs. You must initialize the correct structure with the appropriate data values before you call an RHLI function. The table on page 77 lists the RHLI functions and the structures associated with each function. Some functions do not require the use of a structure. Structures are passed to the RHLI functions as pointers to the structures rather than as the structures themselves. All RHLI function structures are declared and defined in the rhli.h header file. You must specify this file in an INCLUDE statement in your C program that references RHLI functions Structure Index The following table lists the RHLI functions and the associated structures. Use this table to determine which structure to initialize before calling the associated function. Function Structure uabttx uaddath uaddbt uaddcgp uaddcnm uadddb uaddhsh uaddlnk uaddprf DBAUTH DBBTREE DBCGRP DBCOLNM DBINFO DBHASH DBLINK 1UDBVAL None None Function uaddprm uaddprv uaddtbl uaddtnm uaddusr uaddvol ualcscn ualctbl uallath Structure None UGRTDS DBTABLE DBTBLNM DBUSER DBVOLM None UNIPCOLL None is a member of the UNIPCOLL structure. continued The RHLI Structures 77 Function Structure uallbt uallcgp uallcol uallhsh ualllnk uallpid ualltbl ualltx uallusr None None None None None None None None None uallvol ubegord ubegscn ubegtx ubndath ubndbt ubndcol ubnddb ubndhsh ubndlnk ubndtbl ubndusr ubndvol ucbgtx uclritr uclsdb uclsjrn ucmttx None None None None None None None None ucrypwd udelrow udrpath udrpbt udrpcgp udrpcnm udrphsh udrplnk udrpprf None uextint UNIPCOLL, UNIPEXT ufchath ufchcnm ufchjrn ufchlnk ufchmsg ufchord ufchrow None None udrpprm udrpprv udrptbl udrptnm udrpusr udrpvol uendord uendscn uexit ufchscn ufchtbl ufchtnm ufrescn uinfacs uinfath uinfbt uinfcgp uinfcol 1UDBVAL UTXOPT Function URVKDS None DBTBLNM None None None None None UDBVAL1 UNIPCOLL None None UACSINF UATHINF UBTINF UCGPINF UCOLINF uinfdb uinfhsh uinflck uinflnk uinfsch uinfscn uinftbl uinftx uinfusr Structure None UOACOLS None UTXOPT None None None None None None None None None DBCGRP DBCOLNM None None None UJRNINF UNIPCOLL None UNIPCOLL UNIPCOLL UDBINF UHTINF None UNLKINF None UDBVAL1 UTBLINF UTXINF UUSRINF is a member of the UNIPCOLL structure. continued 78 The RHLI Structures Function Structure Function uinfvol uinimsg uinsand uinsbuf uinsor uinsprj uinsrow uinssrt UVOLINF uinstbl uintext None UNIPCOLL UNIPEXT None UNIPCOLL None ulckdb ulckrow ulcksch ulcktbl ulogmsg None None None None None umapath umapbt umapcol umaphsh umaplnk umaptbl umapvol umodath umodbt UATHMAP UBTMAP UCOLMAP UHSHMAP UNLMAP UTBLMAP UVOLMAP DBAUTH DBBTREE umoddb umodhsh umodlnk umodusr umodvol unumath unumcgp unumchd unumcol DBINFO DBHASH DBLINK DBUSER DBVOLM unumprn unumrow unumtbl unumusr unumvol uopndb uopnjrn upkfrow None None None None None None None uposord usetitr uskrord uulkdb uulkrow uulksch uulktbl uupdrow UDBVAL1 1UDBVAL The RHLI Structures None None None Structure UDBVAL1 UNIPCOLL None None None None None None None None None None UNIPCOLL is a member of the UNIPCOLL structure. 79 The DBAUTH Structure: Schema Information The DBAUTH structure has the following declaration. typedef struct { UAID aid; UXID xid; char * authnm; } DBAUTH; /* Authorization ID (returned)*/ /* reserved for future use */ /* pointer to schema name */ The DBAUTH structure is used by the uaddath and umodath functions to describe a schema. The uaddath function adds a new schema. The umodath function modifies, or changes, a schema name. The aid member contains the authorization ID of the schema. If the schema is being added with the uaddath function, Unify DataServer places the authorization ID in aid. Otherwise, you must supply an existing authorization ID. The xid member is reserved for future use. Set xid to UNULLXID. The authnm member contains a pointer to a buffer containing the name of the schema. The name can be specified when adding a new schema or when modifying an existing schema name. If a name is not desired, set authnm to (char *)0 or to an empty string. 80 The RHLI Structures The DBBTREE Structure: B-tree Index Information The DBBTREE structure has the following declaration. typedef struct { char * btname; UBTID btid; UOPTS btopts; UTID tid; UVID vid; int ncol; UBTCOL * btcoll; USTATUS * statusl; } DBBTREE; /* /* /* /* /* /* /* /* name of B-tree B-tree ID B-tree options table B-tree indexes volume on which to place B-tree number of columns list of columns for B-tree key list of corresponding statuses */ */ */ */ */ */ */ */ The DBBTREE structure is used by the uaddbt and umodbt functions to specify B-tree information. The uaddbt function adds a new B-tree index. The umodbt function modifies an existing B-tree index name. The btname member contains a pointer to a buffer containing the name of the B-tree. The name must contain the B-tree name only; it cannot contain the corresponding schema name or table name prefixes. If a name is not desired, set btname to (char *)0 or an empty string. The btid member contains the B-tree ID. If the B-tree is being added, the system returns the B-tree ID in btid. Otherwise, you must supply an existing B-tree ID. The umodbt function can modify a B-tree name only; therefore, only the btname and btid members need to be initialized when calling the umodbt function. The btopts member contains the B-tree creation options: The RHLI Structures DB_INSCATTER data may be scattered across volumes DB_IXUSEVOL specifies preferred volume to use DB_DUPNOK no duplicate keys DB_DUPSOK allow duplicate keys 81 DB_NORMAL DB_DUPNOK is the default value used The tid member contains the table ID to which all of the columns comprising the B-tree index belong. The vid member is reserved for future use. The ncol member contains the count of the number of entries in the B-tree index column list (btcoll). The btcoll member contains a pointer to a list of B-tree index column structures (UBTCOL) which describe the columns comprising the btree index. The list contains ncol entries. The rank (or position) of each column in the B-tree index is implied by its location in this list; for example, the column described by btcoll[0] is the most significant column and is of rank 1. The statusl member contains a pointer to a list of B-tree index column status buffers. The list contains ncol entries. Upon completion of the B-tree operation, each entry in the status list contains an error code which indicates the result of the operation for the corresponding column entry in the btcoll list. The UBTCOL Structure The UBTCOL structure specifies information about a column that makes up the B-tree. The UBTCOL structure is declared as follows: typedef struct { UCID cid; UOPTS btcopts; } UBTCOL; /* Column ID /* DB_ASCEND || DB_DESCND */ */ The cid member contains the column ID of a column which is to comprise the B-tree index. The btcopts member contains the B-tree column processing options as they relate to the corresponding column: DB_ASCEND ascending key value DB_DESCND descending key value DB_NORMAL DB_ASCEND is the default value used 82 The RHLI Structures The DBCGRP Structure: Column Group Information The DBCGRP structure has the following declaration. typedef struct { UCID grpcid; DBCMBR * mbrlst; USTATUS * statusl; int ncol; UOPTS colopts; } DBCGRP; /* /* /* /* /* /* Column Group ID (returned) pointer to column group member column list array of column status codes # of columns in ’mbrlst’ group column processing options */ */ */ */ */ */ The DBCGRP structure is used by the uaddcgp and udrpcg functions to specify information about a column group. The grpcid member contains the column group ID. If the column group is being added, Unify DataServer places the column group ID in grpcid. Otherwise, you must supply an existing column group ID. The mbrlst member contains a pointer to a list of DBCMBR column group member structures which describes the columns which comprise the column group. The list contains ncol entries. The DBCMBR structure is described on page 84. If DB_ORDERD is specified for colopts, the rank (or position) of each column in the mbrlst member list is implied by its location in this list. For example, the column described by mbrlst[0] is the most significant column and is of rank 1. If DB_ORDERD is not specified, Unify DataServer assigns the rank by ordering the column IDs. The statusl member contains a pointer to a list of member column status buffers. The list contains ncol entries. Upon completion of the column group operation, each entry in the status list contains an error code which indicates the result of the operation for the corresponding column entry in the mbrlst list. The ncol member contains the count of the number of entries in the column group member column list mbrlst. The RHLI Structures 83 The colopts member contains the column group options (attributes): DB_ORDNRY ordinary column DB_COLKEY column is primary key for table DB_UNIQUE column must be unique DB_ORDERD column is ordered DB_NORMAL DB_ORDNRY is the default value used You can specify more than one option by separating the options with the bitwise logical OR operator (|). If DB_COLKEY or DB_UNIQUE is specified, each of the members of the column group must have been created with the DB_NONULL option. This option is specified in the DBTABLE structure described on page 98. If DB_ORDERD is specified, you cannot specify DB_COLKEY or DB_UNIQUE. The DBCMBR Structure The DBCMBR structure specifies information about a column member in the column group. The DBCMBR structure has the following declaration. typedef struct { UCID mbrcid; UOPTS mbropts; } DBCMBR; /* column group member Column ID /* column group member options */ */ The mbrcid member contains the column ID of a column that is a member of a column group. The mbropts member contains the member column processing options as they relate to the corresponding column: DB_ASCEND ascending key value DB_DESCND descending key value DB_NORMAL DB_ASCEND is the default value used 84 The RHLI Structures The DBCOLNM Structure: Column Name Information The DBCOLNM structure has the following declaration. typedef struct { UCID cid; char * colnm; } DBCOLNM; /* Column ID to bind name to /* pointer to column name */ */ The DBCOLNM structure is used by the uaddcnm and udrpcnm functions to specify column names. The cid member contains the column ID which this structure describes. The colnm member contains the column name string, zero terminated. The RHLI Structures 85 The DBHASH Structure: Hash Table Index Information The DBHASH structure has the following declaration. typedef struct { char * htname; UCID * colidl; UHID hid; UVID vid; long htthrsh; long htovfsz; int htbchsz; UOPTS htopts; int ncol; } DBHASH; /* /* /* /* /* /* /* /* /* name of hash table Column ID list Hash Table ID volume on which to place hash tbl split threshold value hash table overflow bucket size # of buckets in hash table cache hash table options number of columns */ */ */ */ */ */ */ */ */ The DBHASH structure is used by the uaddhsh and umodhsh functions to specify information about hash table indexes. The htname member contains a pointer to a buffer containing the name of the hash table. The name must contain the hash table name only; it cannot contain the corresponding schema name or table name prefixes. If a name is not desired, set htname to (char *)0 or an empty string. The colidl member contains a pointer to a list of hash table index column IDs that describes the columns that comprise the hash table index. The list contains ncol entries. There is no rank implied in a hash table index. Columns of U_VTXT or U_VBIN data types cannot be used in a hash table key. The hid member contains the hash table ID. If the hash table is being added, the system places the hash table ID in hid. Otherwise, you must supply an existing hash table ID. The umodhsh function can modify a hash table name only; therefore, only the htname and hid members need to be initialized when calling the umodhsh function. The vid member contains the volume ID on which to place the hash table index. If the hash table can be placed on any volume (no volume preference exists), the vid member should contain the value UNULLVID. 86 The RHLI Structures The htthrsh member contains a threshold for splitting hash table buckets. If this number is set to a low value, hash key retrieval time is kept low at the cost of disk space. If this number is set to a high value, space in the hash table index is more fully used at the cost of hash key retrieval time. A htthrsh value of 20L will keep a hash table index moderately full. The htovfsz member contains the number of overflow buckets that should be allocated at one time. A value of 1L means that overflow buckets should be allocated as they are needed. This value incurs the cost of having to manage overflow buckets individually. A value of 16L means that overflow buckets are allocated 16 at a time. This value means that the unused buckets take up space until they are needed, but it is easier for the system to manage the buckets. The htbchsz member is reserved for future use and should be set to 0. The htopts member contains the hash table processing options: DB_DUPNOK no duplicate keys DB_INSCATTER data may be scattered across volumes DB_IXUSEVOL specifies preferred volume to use DB_TYPE00 key folding algorithm #0 DB_TYPE01 (deprecated) DB_TYPE02 (deprecated) DB_TYPE03 (deprecated) DB_TYPE04 (deprecated) DB_TYPE05 (deprecated) DB_TYPE06 key folding algorithm #6 DB_TYPE07 (deprecated) DB_TYPE08 key folding algorithm #8 DB_TYPE09 key folding algorithm #9 DB_NORMAL DB_DUPNOK is the default value used You can select one of the DB_TYPE options only. If you omit a DB_TYPE option, DB_TYPE08 is used. The ncol member contains the count of the number of entries in the hash table index column list colidl. The RHLI Structures 87 The DBINFO Structure: Database Information The DBINFO structure has the following declaration. typedef struct { DBVOLM * rootvol; UOPTS dbopts; /* Operating System int dbmode; int dbusrid; int dbgrpid; char * descrpt; char * dbname; } DBINFO; /* pointer to root volume info. /* database creation options */ */ Specific Values (See: chmod, chown) /* DB file access modes (u/g/o) /* DB file Owner ID (or –1) /* DB file Group ID (or –1) /* pointer to DB description /* pointer to database name */ */ */ */ */ */ The DBINFO structure is used by the uadddb and umoddb functions to specify information about a database. The rootvol member contains a pointer to a DBVOLM structure that describes the root volume information. For more information about defining volumes, see the DBVOLM structure description on page 101 The dbopts member contains the database creation options: DB_PUBLIC Normal mode (do not overwrite any existing databases, create a public database) DB_OVERWR overwrite existing database DB_PRIVAT create private database DB_NORMAL DB_PUBLIC is the default value used You can specify more than one option by separating the options with the bitwise logical OR operator (|). The dbmode member contains the Unix file access modes that all database files are created with; these modes can be changed by the user if so desired. For more information, see the chmod(2) description in your Unix system manual. 88 The RHLI Structures The dbusrid member is the Unix file owner ID that all database files are created with; the owner ID can be changed by the user if so desired. If you specify –1, the current effective user ID is used as the owner ID. You can also use the VOLUSER configuration variable to set the owner ID. If you set the dbusrid member to the value of a user other than the current effective user, you must have the necessary underlying UNIX privileges. The dbgrpid member is the Unix file group ID that all database files are created with; the group ID can be changed by the user if so desired. If you specify –1, the current group ID is used for the group ID. You can also use the VOLGROUP configuration variable to set the group ID. The descrpt member contains a pointer to a buffer containing a description of the database. If a description is not desired, set descrpt to (char *)0 or an empty string. The dbname member contains a pointer to a buffer containing the name of the database. If a name is not desired, set dbname to (char *)0 or an empty string. The umoddb function cannot modify the database root volume (rootvol) or any database creation options (dbopts); therefore, only the dbmode, dbusrid, dbgrpid, descprt and dbname members can be initialized when calling the umoddb function. For more information about user and group IDs, see the chown(2) and chgrp(2) descriptions in your Unix system manual . For more information about configuration variables, see Unify DataServer: Managing a Database The RHLI Structures 89 The DBLINK Structure: Link Information The DBLINK structure has the following declaration. typedef struct { char * lnkname; UCID * pcolidl; UCID * ccolidl; UTID ptblid; UTID ctblid; ULID lid; int ncol; } DBLINK; /* /* /* /* /* /* /* name of link parent column ID list child column ID list parent table ID child table ID Link ID number of columns */ */ */ */ */ */ */ The DBLINK structure is used by the uaddlnk and umodlnk functions to specify information about links. The lnkname member contains a pointer to a buffer containing the name of the link. The name must contain the link name only ; it cannot contain the corresponding schema name or table name prefixes. If a name is not desired, set lnkname to (char *)0 or an empty string. The pcolidl member contains a pointer to a list of link index parent column IDs which describes the columns which comprise the link index. The list contains ncol entries. Note that there is no rank implied in a link index. For the parent of a particular child to be identified, the parent column or set of columns must be unique. See the uaddtb function for information on how to specify a uniqueness constraint on a column. The ccolidl member contains a pointer to a list of link index child column IDs which describes the columns which comprise the link index. The list contains ncol entries. Note that there is no rank implied in a link index. If the link is on more than one column, then the order of column IDs in the lists indicates which columns are to be linked between the two tables. Each column in the child list must have the same type as the corresponding column in the parent list. A column or set of columns can be the child of no more than one link. 90 The RHLI Structures The ptblid member contains the table ID of the parent table to which the columns described by the pcolidl list belong. The ctblid member contains the table ID of the child table to which the columns described by the ccolidl list belong. The child table cannot have the same table ID as the parent table. The lid member contains the link ID. If the link index is being added, Unify DataServer places the link ID in lid. Otherwise, the user is required to supply an existing link ID. Since the umodlnk function can change a link name only, initialize the lnkname and lid members only of the DBLINK structure when calling the umodlnk function. The ncol member contains the count of the number of entries in the link index column list pcolidl and ccolidl. The RHLI Structures 91 The DBTABLE Structure: Table Information The DBTABLE structure has the following declaration. typedef struct { UTID tid; DBCOL * dbcol; USTATUS * statusl; UVID vid; long segsz; long expnum; long tblbase; UOPTS tblopts; int ncol; char * tblnm; char * tbldesc; } DBTABLE; /* /* /* /* /* /* /* /* /* /* /* Table ID (returned) array of column information array of column status codes initial preference volume desired segment size (bytes) expected number of rows direct table base value table processing options # of entries in dbcol list pointer to table name pointer to table description */ */ */ */ */ */ */ */ */ */ */ The DBTABLE structure is used by the uaddtbl function to specify information about a table. The tid member contains the table ID of the table. If the table is being added, Unify DataServer places the table ID in tid. Otherwise, you must supply an existing table ID. The dbcol member contains a pointer to a list of DBCOL column information structures, which provide information about the table columns. The dbcol list contains ncol entries. The statusl member contains a pointer to a list of status buffers, each entry of which indicates the success or failure of the corresponding entry in the column information list. The list contains ncol entries. The vid member contains the volume ID on which to place the table. If the table can be placed on any volume (no volume preference exists), set the vid member to UNULLVID. The segsz member contains the table’s desired segment size, in bytes. If you want Unify DataServer to choose an optimal segment size for you, set segsz to 0L. 92 The RHLI Structures A segment is a contiguous disk buffer where table rows reside; typically, a segment’s size corresponds to the table’s row size. This helps minimize the amount of fragmentation left over in the segment. The expnum member contains the number of rows, if any, designated when the table was created. If the DB_FIXSIZE option is specified, the expnum value specifies a fixed number of rows. The expnum member’s value is advisory only and does not place any restrictions on the number of records that can be inserted into the table. The tblbase member contains the table’s base value, if the table is a direct table. Otherwise, the value of tblbase is zero (0). The tblopts member options are: DB_ORDNRY create ordinary table DB_DIRECT create directĆkeyed table DB_FIXSIZE reserved for future use DB_KEYED all directĆkey column values are userĆsupplied DB_RDONLY reserved for future use DB_SCATTR scatter data across volumes DB_USEVOL must use specified preference volumes DB_NORMAL DB_ORDNRY is the default value used You can specify more than one option by separating the options with the bitwise logical OR operator (|). If you specify the DB_KEYED option, you must also specify the DB_DIRECT option. The DB_USEVOL option indicates that the rows inserted into the table must be stored on the volume specified by the vid member. If DB_USEVOL is not specified, then vid is taken as the desired volume, but other volumes defined for the database are not excluded. The DB_SCATTR option indicates that rows inserted into the table are to be scattered evenly over all volumes in the volume preference list (as space permits). If DB_SCATTR is not specified, then rows are placed in the first volume in the list until it becomes full, then the second volume in the list, and so forth. For more information, see the uaddprf function. The RHLI Structures 93 If no volume is specified in vid, then all volumes defined for the database are used for scattering. The ncol member contains the number of columns in the table column information list dbcol and its corresponding status list statusl. If the value of ncol is zero (0), the dbcol and statusl members are not allocated and should not be referenced. The tblnm member contains a pointer to a buffer containing the name of the table. If a name is not desired, set tblnm to (char *)0 or an empty string. The tbldesc member contains a pointer to a buffer containing the textual description of the table. If a description is not desired, set tbldesc to (char *)0 or an empty string. Performance When the direct access column values are scattered, the system overhead for generating the values can be substantial. If your direct access column values are scattered and you do not want Unify DataServer to generate the values, specify the DB_KEY option. The DBCOL Structure The DBCOL structure specifies column information. The DBCOL structure has the following declaration. typedef struct { UCID cid; int coltyp; int collen; int colscl; int dsplen; int dspscl; UOPTS colopts; char * colnm; char * coldesc; char * dsppict; } DBCOL; /* /* /* /* /* /* /* /* /* /* Column ID (returned) column type column storage precision (XXX.XXX) column storage scale (.YYY) column display precision column display scale column processing options pointer to column name pointer to column description pointer to column display picture */ */ */ */ */ */ */ */ */ */ The cid member contains the column ID of the created table column. If the column is being added, the system places the column ID in cid. Otherwise, the user is required to supply an existing column ID. 94 The RHLI Structures The coltyp member contains the data type of the column. Valid column types are: U_INT for SMALLINT (16-bit ) U_HINT for INTEGER (32-bit) U_GINT for HUGE INTEGER (64-bit) U_FLT for FLOAT (double precision floating point) U_DBL for DOUBLE PRECISION (double precision floating point) U_REAL for REAL (single precision floating point) U_AMT for AMOUNT U_HAMT for HUGE AMOUNT U_GAMT for CURRENCY U_DATE for DATE U_HDATE for HUGE DATE U_TIME for TIME U_STR for CHARACTER U_BYT for BYTE U_VTXT for TEXT U_VBIN for BINARY For more information about column data type declaration, see the rhli.h header file. The RHLI Structures 95 The collen member contains the internal storage precision, or length, of the column. The colscl member contains the internal storage scale of the column, if applicable. The following table shows values allowed for collen and colscl. For columns of this data type: Valid collen values are: Maximum collen value defined as: U_INT 1-4 UIL_INT 0 U_HINT 1-9 UIL_HINT 0 U_GINT 1-19 UIL_GINT 0 U_REAL 32 UIL_REAL 0 U_FLT 64 UIL_FLT 0 U_DBL 64 UIL_DBL 0 U_AMT 3-9 UIL_AMT 2 U_HAMT 3-15 UIL_HAMT 2 U_GAMT 1-19 UIL_GAMT 0–8 U_DATE 0 UIL_DATE 0 U_HDATE 0 UIL_HDATE 0 U_TIME 0 UIL_TIME 0 U_STR 1-256 UIL_STR 0 U_BYT 1-1024 UIL_BYT 0 U_VTXT 0 UIL_VTXT 0 U_VBIN 0 UIL_VBIN 0 Valid colscl values are: Except for U_GAMT, a colscl value of 0 indicates that scale is not applicable to the data type and must be set to 0. 96 The RHLI Structures The dsplen member contains the external display precision, or length, of the column. Valid values are shown in the table. You can also set dsplen to zero (0) which causes the default display precision for that data type to be used. For columns of data type: Valid dsplen values are: Default dsplen value defined as: U_INT 2-5 UDL_INT 0 U_HINT 2-10 UDL_HINT 0 U_GINT 2-20 UDL_GINT 0 U_REAL 1-11 UDL_REAL 0-9 U_FLT 1-17 UDL_FLT 0-15 U_DBL 1–17 UDL_DBL 0-15 U_AMT 4-16 UDL_AMT 2 U_HAMT 4-25 UDL_HAMT 2 U_GAMT 2-22 UDL_GAMT 0-8 U_DATE 8-11 UDL_DATE 0 U_HDATE 8–11 UDL_HDATE 0 U_TIME 5 UDL_TIME 0 U_STR 1–256 UDL_STR 0 U_VTXT 1–256 UDL_VTXT 0 U_VBIN 0 UDL_VBIN 0 U_BYT 1–1024 UDL_BYT 0 Valid dspscl values are: Except for U_GAMT, U_REAL, U_FLT, and U_DBL, a colscl value of 0 indicates that scale is not applicable for the data type and must be set to 0. The colnm member contains a pointer to a buffer containing the name of the column. The name must contain only the column name component; it cannot contain the corresponding schema name or table name prefixes. If a name is not desired, set colnm to (char *)0 or an empty string. The RHLI Structures 97 The coldesc member contains a pointer to a buffer containing the textual description of the column. If a description is not desired, set coldesc to (char *)0 or an empty string. The dsppict member contains a pointer to a buffer containing an RPT-like display picture clause. If a display picture clause is not desired, set dsppict to (char *)0 or an empty string. The colopts member contains the various column processing options: DB_ORDNRY Indicates that the other processing options are not specified (all options are at their default value of FALSE). DB_COLKEY Indicates that the column is a primary key for the table. DB_DELETE Reserved for future use. DB_IMMED Reserved for future use. DB_NOLOG Reserved for future use. DB_NONULL Indicates that the column can only contain nonĆnull values. DB_RDONLY Reserved for future use. DB_UNIQUE Indicates that the column can only contain unique values. DB_NORMAL DB_ORDNRY is the default value used You can specify more than one option by separating the options with the bitwise logical OR operator (|). If the DB_UNIQUE option is specified, you must also specify the DB_NONULL option. If the DB_COLKEY option is specified, you must also specify the DB_UNIQUE option. 98 The RHLI Structures The DBTBLNM Structure: Table Name Information The DBTBLNM structure has the following declaration. typedef struct { UTID tid; /* Table ID to bind name to char * tblnm; /* pointer to table name } DBTBLNM; */ */ The DBTBLNM structure is used by the uaddtnm and udrptnm functions to specify a table name for a table. The tid member contains the table ID which this structure describes. The tblnm member contains the table name string, zero terminated. The RHLI Structures 99 The DBUSER Structure: User and Default Schema Information The DBUSER structure has the following declaration. typedef struct { UOID uoid; UAID aid; char * usernm; } DBUSER; /* User/Owner ID (returned) /* default Authorization ID (or NULL) /* pointer to user name */ */ */ The DBUSER structure is used by the uaddusr and umodusr functions to specify information about a user and a user’s default schema. The uoid member contains the user ID. If the user is being added, Unify DataServer places the user ID in uoid. Otherwise, you must supply an existing user ID. The aid member contains the authorization ID that corresponds to the user’s default schema. If aid is set to UNULLAID, the authorization ID for the PUBLIC schema is used. The usernm member contains a pointer to a buffer containing the name of the user. You must specify a user name. The user name should match the user’s name as it is known to the Unix operating system. 100 The RHLI Structures The DBVOLM Structure: Volume Information The DBVOLM structure has the following declaration. typedef struct { UVID vid; long voloff; long vollen; long volpgsz; UOPTS volopts; char * volfile; char * volname; } DBVOLM; /* Volume ID (returned/specified) /* volume offset (bytes) /* volume length /* volume page size (bytes) /* volume processing options /* pointer to volume file name /* pointer to volume name */ */ */ */ */ */ */ The DBVOLM structure is used by the uaddvol and umodvol functions to specify information about a volume. DBVOLM is also a member of the structure DBINFO. The vid member is used for the volume ID for the volume that this structure describes. If the volume is being added and the DB_SPECIFY option is not specified as a volume processing option, the system places a volume ID in vid. If the volume is being added and the DB_SPECIFY option is specified, you specify the volume ID for the volume in the vid member. The volume ID you specify cannot be an existing volume ID. If the volume is being modified, specify an existing volume ID in the vid member. The voloff member is the offset within the physical media in which the volume resides where data storage for the volume begins. The voloff value is typically used when a single physical device is used for two or more volumes; the offset of the second volume would be the address immediately following the first volume, and so on. The volume offset is the number of bytes from the beginning of the disk partition; this value is converted internally into the corresponding number of blocks. The offset for the root volume must always be zero (0), although a volume offset of zero does not necessarily identify the root volume. The RHLI Structures 101 The vollen member indicates the length of the volume. If DB_BIG is set in the volopts member, vollen represents the number of pages. Otherwise, it represents the number of bytes; this value is converted internally into the corresponding number of blocks. The volume length is used by volume storage routines to determine how much data can fit onto the disk partition. If the volume type is a device or preallocated file, you must specify a non-zero length. If the volume type is a regular file, you can specify a non-zero or zero length. A length of zero indicates that the size is unlimited, A non-zero length indicates a maximum size limit for the regular file; the file grows until it reaches the size. The volpgsz member specifies the operating system dependent page size for the device on which the volume resides. The volpgsz size is expressed in bytes; that is, the value 2048 represents a page whose size is 2048 bytes long. Note that this value is converted to the internal page size, which is an integral number of blocks. The volpgsz value must match the value preset for the Unify DataServer release. If zero is specified, the value specified by the VOLPGSZ configuration variable is used. The volopts member contains various volume processing options. Options are: DB_ORDNRY Indicates that the processing options are normal. That is, the volume is not a device or preĆallocated. This is the default setting. DB_ALLOC Indicates that the volume is a preallocated file. DB_CONTIG Reserved for future use. DB_DEVICE Indicates that the volume is a device (character or blockĆspecial file). If the volume options include DB_DEVICE, the device special file must exist before adding a volume on the device. DB_FORCE 102 Indicates that if the volume is a regular file, the UNIX operating system allocates an entire new segment when the existing segment capacity is exceeded. If this option is not specified, the regular files grow in smaller allocations each time data is added to the file. The RHLI Structures DB_RDONLY Reserved for future use. DB_SPECIFY Indicates that the volume ID is specified (rather than returned) in the vid member. If the volume ID you specify is not currently being used, the volume ID is assigned to the volume being created. DB_BIG Indicates that the vollen specification is in pages rather than bytes. DB_NORMAL DB_ORDNRY is the default value used You can specify more than one option by separating the options with the bitwise logical OR operator (|). The volfile member is the full path name of the file or character device for the volume, including the file name. If the file name does not end with the .db suffix, Unify DataServer appends the suffix to the name. If the volume is a root volume, it must reside in the directory specified by the DBPATH and DBNAME configuration variables. The path and file name can be omitted or equal to null (char *)0 ; in this case the value of the DBPATH and DBNAME configuration variables are used. If the volume is not a root volume, you must specify the file name. The path can be omitted; the default path is determined by the DBPATH configuration variable. The volname member contains a pointer to a buffer containing the name of the volume. If a volume name is not desired, set volname to (char *)0 or a null string (””). If you create a database with a device as the root volume, the device must exist in the directory specified by the DBPATH configuration variable. The device must have the name indicated by the DBNAME configuration variable and the .db file name suffix. This is accomplished by using the mknod utility the directory specified by DBPATH or symlinking the device into DBPATH. To re-use an existing volume that is a device, use the mkvol utility to create the volume. The mkvol utility marks the volume as deleted. The volume size information is maintained in terms of bytes. This is because the user may not know the current database’s block size, and an incorrect calculation on the user’s part cause damage to the database. The RHLI Structures 103 Since the umodvol function can change a volume name, initialize only the volname and vid members only of the DBVOLM structure when calling the umodvol function. Additional Help About 104 See Symlinking Your operating system documentation Configuration variables Unify DataServer: Configuration Variable and Utility Reference The RHLI Structures The UACSINF Structure: Access Method Information The UACSINF structure has the following declaration. typedef struct { int scanid; int acsknd; UACS acsid; } UACSINF; /* scan identifier /* access type (UBTSCN, etc) /* access method info structure */ */ */ The UACSINF structure is used by the uinfacs function to retrieve access method information about an access. The scanid member contains the identifier of the access that the structure describes. The acsknd member contains a value that identifies the type of access method being used by the access. Options are: The btacs Structure USQSCN sequential access type UDKSCN direct key access type UBTSCN BĆtree index access type UHKSCN hash index access type ULKSCN link index access type UNULLSCN no access type, not initialized The btacs structure of type UBTACS contain information if the acsknd member is set to UBTSCN. The btacs structure has the following declaration. typedef struct { UBTID btid; int nbtcol; } UBTACS; /* B-tree Index ID /* # of btree key columns used */ */ The btid member contains the B-tree ID of the B-tree index used by the access. The nbtcol member contains the number of components of the B-tree key used by the access. The RHLI Structures 105 The acid Union The acsid member is a union of type UACS that contains additional information for certain access types. The acsid union has the following declaration. typedef union { UBTACS btacs; UHID hid; ULID lid; } UACS; /* UBTSCN: btree access info /* UHKSCN: Hash Index ID /* ULKSCN: Link Index ID */ */ */ The hid member contains the Hash Table ID of the hash index used by the access. It will be filled in if the value of acsknd is equal to UHKSCN. The lid member contains the Link ID of the link index used by the access if the acsknd member is set to ULKSCN. 106 The RHLI Structures The UATHINF Structure: Schema Information The UATHINF structure has the following declaration. typedef struct { /* information always retrieved */ UAID aid; /* Authorization ID /* information retrieval class: I (UCLASS1) */ UXID xid; /* reserved for future use char authnm[U_ATHNMLN + 1]; /* pointer to the /* schema name } UATHINF; */ */ */ */ The UATHINF structure is used by the uinfath function to retrieve information about a schema. A schema is also called an authorization. The aid member contains the authorization ID of the schema. The xid member is reserved for future use. The authnm member contains the name of the schema as it would be referenced for name-binding. The RHLI Structures 107 The UATHMAP Structure: Authorization ID Mapping The UATHMAP structure has the following declaration. typedef struct { char * athnm; UAID aid; UAID remapid; } UATHMAP; /* pointer to runtime schema name /* compile-time Authorization ID /* runtime Auth. ID (returned) */ */ */ The UATHMAP structure is used by the umapath function to specify information about runtime ID-mapping on a compile-time schema. Tip Use the INIATHMAP mapping macro to initialize the UATHMAP structure. The athnm member contains the runtime name of the schema. To reset the value of a previously mapped authorization ID, set the value of athnm to (char *)0 or the null-string (“”). The aid member contains an authorization ID that uniquely identifies the schema. The authorization ID can be supplied by you or can be generated by the ucc compiler. If you supply the authorization ID (aid), be sure that the number you pick is compatible with the data type UAID. To generate a unique compile-time authorization ID by using ucc, initialize aid with the following compiler directive (athnm is the name of the schema): $(UATH:AID)athnm$ The remapid member, upon successful completion of the ID-mapping operation, contains the runtime ID to which the compile-time ID is mapped. This value is for internal use only. 108 The RHLI Structures The UBTINF Structure: B-tree Information The UBTINF structure has the following declaration. typedef struct { /* information always retrieved */ UBTID btid; /* B-tree this struct describes UBTID remapid; /* logical B-tree ID (if remapped) */ */ /* information retrieval class: I (UCLASS1) */ UOPTS btopts; /* DB_DUPSOK || DB_DUPNOK UTID tid; /* Table ID this B-tree indexes */ */ /* information retrieval class: II (UCLASS2) */ int count; /* length of B-tree column list UBTCINF *btcinf; /* pointer to UBTCINF structure list */ */ /* information retrieval class: III (UCLASS3) */ char btnm[U_NAMELEN + 1]; /* name of B-tree index } UBTINF; */ The UBTINF structure is used by the uinfbt function to retrieve B-tree information. The btid member contains the B-tree index ID that the structure describes. If the ID represents a remapped btree index, the ID is the compile-time B-tree index ID (the ID with which the user remapped the B-tree index). The remapid member contains the logical B-tree index ID that the structure describes. If the ID represents a remapped B-tree index, the ID is the runtime B-tree index ID (the actual Database ID). Otherwise, remapid is identical to the btid member. The btopts member contains the B-tree creation options: DB_DUPNOK no duplicate keys DB_DUPSOK allow duplicate keys DB_NORMAL DB_DUPNOK is the default value used The tid member contains the table ID to which all of the columns comprising the B-tree index belong. The RHLI Structures 109 The count member contains the count of the number of entries in the B-tree index column information list btcinf. The btcinf member contains a pointer to a list of UBTCINF B-tree index column information structures which describes the columns comprising the B-tree index. The list contains ncol entries. See the description of the UBTCINF structure below. The btname member contains the name of the B-tree, if any. If the B-tree index does not have a name, the member contains an empty string. The UBTCINF Structure The UBTCINF structure contains B-tree column information. The UBTCINF structure has the following declaration. typedef struct { int rank; UCID cid; UOPTS btcopts; } UBTCINF; /* 0 top rank – lowest rank /* Column IDs (cid’s) /* DB_ASCEND || DB_DESCND */ */ */ The maximum number of columns in the B-tree is 128. The rank member contains the rank, or position, of the corresponding column in the B-tree index. The rank value starts at zero, indicating the most significant column in the B-tree. The cid member contains the column ID of a column member of the B-tree index. All column IDs in a single B-tree index belong to the same table. The btcopts member contains the B-tree column processing options as they relate to the corresponding column: DB_ASCEND ascending key value DB_DESCND descending key value DB_NORMAL DB_ASCEND is the default value used 110 The RHLI Structures The UBTMAP Structure: B-tree ID Mapping The UBTMAP structure has the following declaration. typedef struct { char * btnm; UBTID bid; UBTID remapid; } UBTMAP; /* pointer to runtime B-tree name /* compile-time B-tree ID /* runtime B-tree ID (returned) */ */ */ The UBTMAP structure is used by the umapbt function to specify runtime ID-mapping on a compile-time B-tree object. Tip Use the INIBTMAP mapping macro to initialize the UBTMAP structure. The btnm member contains the runtime name of the B-tree. The remapping operation fails if the B-tree specified in the btnm member does not exist in the database. The bid member contains a B-tree ID that uniquely identifies the B-tree. The ID can be supplied by you or can be generated by the ucc compiler. If you supply the ID (bid), be sure that the number you pick is compatible with the type UBTID. To generate a unique compile-time B-tree ID using ucc, initialize bid with the following compiler directive (btnm is the name of the B-tree): $(UBT:BID)btnm$ The remapid member, upon successful completion of the ID-mapping operation, contains the runtime ID to which the compile-time ID was mapped. This value is for internal use only. The RHLI Structures 111 The UCGPINF Structure: Column Group Information The UCGPINF structure has the following declaration. typedef struct { /* information always retrieved */ UCID cid; /* its Column ID int coltyp; /* data type of column int ncol; /* # Column IDs in grpcidl UCID * grpcidl; /* list of Column IDs in the group /* OR the groups the column is a /* member of } UCGPINF; */ */ */ */ */ */ The UCGPINF structure is used by the uinfcgp function to retrieve information about a column group and its component columns, or for a column and all of the column groups of which it is a component. The cid member contains the column ID this structure describes. The column ID can describe either a column or column group. The coltyp member contains the column’s Unify DataServer database type, just as it does for the DBTABLE structure. The ncol member contains the number of column IDs in the column group ID list grpcidl. If the value of ncol is zero (0), the grpcidl member is not allocated and should not be referenced. The grpcidl member contains the list of column IDs to which the component column belongs, or the list of column IDs which belong to the column group. For column groups successfully retrieved, storage space is allocated by Unify DataServer for grpcidl. This space should be released after the retrieval operation is complete by calling the free function. 112 The RHLI Structures The UCOLINF Structure: Column Information The UCOLINF structure has the following declaration. typedef struct { /* information always retrieved */ UTID tid; /* table to which column belongs UAID aid; /* authorization/schema of the table UCID cid; /* the column’s Column ID UCID remapid; /* logical Column ID (if remapped) */ */ */ */ /* information retrieval class: I (UCLASS1) */ UDEFID defid; /* column Definition ID UDEFID chgid; /* column Modification ID UMAPID mapid; /* column summation ID int coltyp; /* Unify DataServer data type int collen; /* column storage precision int colscl; /* column storage scale UOPTS colopts; /* column processing options */ */ */ */ */ */ */ /* information retrieval class: II (UCLASS2) */ int dsplen; * column display precision int dspscl; /* column display scale */ */ /* information retrieval class: III (UCLASS3) */ int nbt; /* # of B-trees the column /* participates in int nhsh; /* # of hash-tables the column /* belongs to int nplnk; /* # of links the column /* is parent of int nclnk; /* # of links the column /* is child of (0/1) */ */ */ */ */ */ */ */ /* information retrieval class: IV (UCLASS4) */ char * dsppict; /* pointer to column display picture /* (or NIL) char * coldesc; /* pointer to column description /* (or NIL) } UCOLINF; */ */ */ */ The UCOLINF structure is used by the uinfcol function to retrieve information about a column and its attributes. The cid member contains the column ID that the structure describes. If the ID represents a remapped column, the ID is the compile-time column ID (the ID with which the user remapped the column). The RHLI Structures 113 The remapid member contains the logical column ID that the structure describes. If the ID represents a remapped column, the ID is the runtime column ID (the actual Database ID). Otherwise, remapid is identical to the cid member. The tid member contains the table ID to which the column belongs. The aid member contains the authorization ID(schema) to which the column’s table belongs, and to which the column indirectly belongs. The defid member contains the column’s current Definition ID. The Definition ID changes whenever a physical modification is made to the column, such as modifying the column’s type or length. The chgid member contains the column’s current Change ID. The Change ID changes whenever a logical modification is made to the column, such as modifying the column’s name. The mapid member contains the column’s current Mapping ID. The Mapping ID serves as a checksum of the column’s physical attributes, such as the column’s type and length. The coltyp member contains the column’s database type. Options are listed with the coltyp member of the DBCOL description of the DBTABLE structure on page 92. The collen and colscl members contain the column’s internal storage precision and scale, as follows: U_INT, U_HINT, U_GINT collen is the number of digits as specified (or implied) when the column was created; colscl is 0. U_DATE, U_TIME, U_HDATE, U_VTXT, U_VBIN collen and colscl are both 0. U_REAL, U_FLT, U_DBL collen is the number of bits of floating point representation for the internal, or stored, value: 32 for U_REAL and 64 for U_FLT and U_DBL; colscl is 0. U_BYT, U_STR 114 collen is the number of bytes or characters specified (or implied) when the column was created; colscl is 0. The RHLI Structures U_AMT, U_HAMT collen is the number of digits of precision (possible digits to the left of the radix point plus possible digits to the right of the radix point) as specified. It will be >= to 3. colscl is 2. U_GAMT collen is the number of digits of precision (possible digits to the left of the radix point plus possible digits to the right of the radix point) as specified. It will be > colscl. colscl is the defined scale (0-8); It will be < collen. The dsplen member contains the external display precision or length of the column. Display length can be specified as any length (zero or greater). A zero (0) display length indicates no display length. The colopts member contains the column’s processing options. See page 98. The dspscl member contains the column’s display scale, if applicable, in characters. Valid values allowed for dsplen and dspscl are defined in the rhli.h header file. The nbt member contains the number of B-trees the column participates in. The nhsh member contains the number of hash tables the column participates in. The nplnk member contains the number of links the column is a parent of. The nclnk member contains the number of links the column is a child of. Note that nclnk can only contain the values zero (0) or one (1). The dsppict member contains the column’s display picture clause, if any. The coldesc member contains the column’s description, if any. For columns successfully retrieved by specifying the UCLASS4 or UALLINFO retrieval classes, storage space is allocated by Unify DataServer for coldesc and dsppict members. This space should be released after the retrieval operation is complete by calling the free function. The RHLI Structures 115 The UCOLMAP Structure: Column ID Mapping The UCOLMAP structure has the following declaration. typedef struct { char * colnm; UCID cid; UCID remapid; UDEFID defid; UMAPID mapid; } UCOLMAP; /* /* /* /* /* pointer to runtime column name compile-time Column ID runtime Column ID (returned) compile-time column definition compile-time column summation */ */ */ */ */ The UCOLMAP structure is used by the umapcol function to specify information about runtime ID-mapping on a compile-time column object. Tip It is highly recommended that you use the INICOLMAP mapping macro to initialize the UCOLMAP structure. The colnm member contains the runtime name of the column. The cid member contains the compile-time column ID. The ID can be supplied by you or can be generated by the ucc compiler. If you supply the ID (cid), be sure that the number you pick is compatible with the type UCID. To generate a unique compile-time ID using ucc, initialize cid with the following compiler directive (colnm is the name of the column): $(UCOL:CID)colnm$ The remapid member, upon successful completion of the ID-mapping operation, contains the runtime ID to which the compile-time ID was mapped. This value is for internal use only. The defid member contains the table’s compile-time Definition ID, which is a time-stamped ID that is used by Unify DataServer to verify the accuracy of remapid. 116 The RHLI Structures The Definition ID (defid) must be initialized with the following compiler directive syntax (colnm is the name of the column): $(UCOL:DEFID)colnm$ The mapid member contains the table’s compile-time Mapping ID, which is a checksum value that is used by Unify DataServer to verify the accuracy of remapid. This value is for internal use only. The Mapping ID (mapid) must be initialized with the following compiler directive syntax (colnm is the name of the column): $(UCOL:MAPID)colnm$ The RHLI Structures 117 The UDBINF Structure: Database Information The UDBINF structure has the following declaration. typedef struct { /* information always retrieved */ UDBID dbid; /* database ID int isremote; /* Is database remote? flag */ */ /* information retrieval class: I (UCLASS1) */ UDEFID defid; /* database definition ID */ /* Operating System Specific Values (See: chmod, chown) */ int dbmode; /* database file access modes (u/g/o)*/ int dbusrid; /* database file Owner ID */ int dbgrpid; /* database file Group ID */ /* information retrieval class: II (UCLASS2) */ long nvol; /* # of volumes in the database UVID * uvidlst; /* pointer to volume information */ */ /* information retrieval class: III (UCLASS3) */ char * descrpt; /* pointer to database description (or NIL )*/ char ownernm[U_USRNMLN + 1]; /* buffer for the database creator name */ char dbname[U_NAMELEN + 1]; /* buffer for the database name */ } UDBINF; The UDBINF structure is used by the uinfdb function to retrieve information about a database. The dbid member contains the database ID which is described by the information in this structure. The isremote member specifies if the database to be accessed is a remote database. If the database is remote, isremote will be set to 1; if the database is local, isremote will be set to 0. The defid member contains the current Definition ID for the database. The dbmode member contains the database Unix access modes, as described in section 2 of your Unix system manual. All database files are created with these file access permissions. 118 The RHLI Structures The dbusrid member contains the database Unix owner ID, as described in section 2 of your Unix system manual. All database files are created with these file user ownership. The dbgrpid member contains the database Unix group ID, as described in section 2 of your Unix system manual. All database files are created with these file group ownership. The nvol member contains the number of volumes in the database. A database always includes at least 1 volume: the root volume. The uvidlst member contains a list of volume IDs which belong to the database. The descrpt member contains the database description. The ownernm member contains the name of the user who created the database. The dbname member is the name of the database as it would be referenced for name-binding. The RHLI Structures 119 The UGRTDS Structure: Grant Privilege Descriptor The UGRTDS structure has the following declaration. typedef struct { union { UAID aid; UOID uid; } gte; UCAPS rqstcaps; UOBJDS obj; UCAPS gtecaps; } UGRTDS; /* /* /* /* Grantee Capabilites to be granted Object to grant on Capabilities actually granted */ */ */ */ The UGRTDS structure is used by the uaddprv function to specify information about the privileges to grant. The aid member contains the authorization ID of a schema to which schema privileges are to be granted. The uid member contains the user ID to which permissions are to be granted. The gte member contains the ID of the object to which the user authority is granted. The object to which the privileges are to be granted is known as the grantee. Because gte is a union, either a schema or a user can be granted a privilege, but not both. The rqstcaps member contains the set of permissions that are to be granted. The set of permissions determines whether the type of the grantee is a schema or a user. When you grant permissions, you can grant only those permissions that you have permission to grant. To grant authority, initialize the rqstcaps member as follows: 120 UDBA for DBA authority USCH for SCHEMA authority UDALL for both DBA and SCHEMA authority The RHLI Structures To grant schema privileges, initialize the rqstcaps member as follows: USSEL for SELECT operations USINS for INSERT operations USUPD for UPDATE operations USDEL for DELETE operations USALL for all grantable operations (SELECT, INSERT, UPDATE and DELETE) UGRTCAP to specify that the operation (one of the above) is grantable If a schema grants schema privileges to another schema, all tables in the schema granting the privilege are accessible. However, if a schema grants table privileges to another schema, only the specified tables are accessible. To grant table privileges to a schema, initialize the rqstcaps member as follows: UTSEL for SELECT operations UTINS for INSERT operations UTUPD for UPDATE operations UTDEL for DELETE operations UTALL for all grantable operations (SELECT, INSERT, UPDATE and DELETE) UGRTCAP to specify that the operation (one of the above) is grantable To grant column privileges to a schema, initialize the rqstcaps member as follows: The RHLI Structures UCSEL for SELECT operations UCINS for INSERT operations UCUPD for UPDATE operations UCALL for all grantable operations (SELECT, INSERT and UPDATE) UGRTCAP to specify that the operation (one of the above) is grantable 121 You can specify more than one option by separating the options with the bitwise logical OR operator (|). The obj member is a structure (UOBJDS) that contains a description of the object which the privilege affects. See below for a description of the UOBJDS structure. The gtecaps member contains the set of privileges actually granted to the grantee. That is, if you granted SELECT schema privilege to a table, the value contained in gtecaps would be equivalent to UTSEL; or, if you granted SCHEMA authority to a user, if granted successfully, gtecaps would contain a value equivalent to USCH. Since the grantor may not have the privilege to grant certain privileges, only those privileges actually grantable are granted. For more information about the functionality of each of the operation privileges, see “Controlling Database Object Access” in Unify DataServer: Writing Interactive SQL/A Queries. The UOBJDS Structure The UOBJDS structure contains information about a database object and is defined as: typedef struct { UDBOBJ objid; int objknd; } UOBJDS; /* object ID /* the type of object */ */ The objid member contains a database object ID whose type is described by the objknd member. Options are: UAID Authorization ID UDBID Database ID UCID Column ID UTID Table ID The objknd member contains a value that indicates what type of ID is contained in the objid member. Options are: 122 UDBKND database object USCHKND schema object UTBLKND table object The RHLI Structures UVWKND view object UCOLKND column object If you are granting a user privilege (DBA or SCHEMA authority), objid and objknd must be set to the current Database ID (see the uopndb function) and UDBKND, respectively. The RHLI Structures 123 The UHSHMAP Structure: Hash Table ID Mapping The UHSHMAP structure has the following declaration. typedef struct { char * hshnm; UHID hid; UHID remapid; } UHSHMAP; /* /* /* /* pointer to runtime hash table name compile-time Hash Table ID runtime Hash Table ID (returned) */ */ */ */ The UHSHMAP structure is used by the umaphsh function to specify information about runtime ID-mapping on a compile-time hash table object. It is highly recommended that you use the INIHSHMAP mapping macro to initialize the UHSHMAP structure. The hshnm member contains the runtime name of the hash table. The hid member contains the compile-time hash table ID. The hash table ID (hid) can be hardcoded by the programmer (with a unique custom ID) or be supplied by the Unify DataServer C compiler, ucc. If you choose to hardcoded a unique hash table ID (hid), be sure that the number you pick is compatible with the type UHID. To generate a unique compile-time hash table ID using ucc, initialize hid with the following compiler directive (hshnm is the name of the hash table): $(UHSH:HID)hshnm$ The remapid member, upon successful completion of the ID-mapping operation, contains the runtime ID to which the compile-time ID was mapped. This value is for internal use only. 124 The RHLI Structures The UHTINF Structure: Hash Table Information The UHTINF structure has the following declaration. typedef struct { /* information always retrieved */ UHID hid; /* hash table this struct describes UHID remapid; /* logical hashtb ID (if remapped) */ */ /* information retrieval class: I (UCLASS1) */ long hthibkt; /* highest primary bucket long htsplvl; /* split level long hthshvl; /* primary hash value long htthrsh; /* split threshold value long htovfsz; /* hash table overflow bucket size int htbchsz; /* # of buckets in hash table cache UOPTS htopts; /* DB_DUPSOK || DB_DUPNOK */ */ */ */ */ */ */ /* information retrieval class: II (UCLASS2) */ long count; /* length of hash column list UCID * htloc; /* pointer to Column ID list */ */ /* information retrieval class: III (UCLASS3) */ char htnm[U_NAMELEN + 1]; /* name of hash table } UHTINF; */ The UHTINF structure is used by the uinfhsh function to retrieve information about a hash table. The hid member contains the hash table ID that the structure describes. If the ID represents a remapped hash table, the ID is the compile-time hash table ID (the ID with which the user remapped the hash table). The remapid member contains the logical hash table ID that the structure describes. If the ID represents a remapped hash table, the ID is the runtime hash table ID (the actual Database ID). Otherwise, remapid is identical to the hid member. The hthibkt member contains the current number of primary buckets in the hash table; this values increases as rows are added and decreases as rows are deleted. The value of hthibkt ranges from the value of hthshvl to twice that value. The RHLI Structures 125 The htsplvl member contains the number of times the hash table has doubled in size; the value one (1) means that the hash table is at its original size. The hthshvl member contains the base size of the hash table, which is used to compute the primary bucket address for a given key value. When the value of hthibkt reaches twice the value of hthshvl, the value of htsplvl is incremented and the value of hthshvl is doubled. The htovfsz member contains the size, in blocks, of a segment that contains overflow buckets. This size corresponds to the number of overflow buckets that are allocated as a group. The htthrsh member contains the measurement used to determine when a primary bucket needs to be split. The value indicates the number of internal overflow chains of key values that should have been able to fit into a primary bucket. The htbchsz member contains the number of buckets that can be cached in user memory during an atomic hash table operation. The htopts member contains the hash table processing options. Options are: DB_DUPNOK disallow duplicate keys DB_DUPSOK allow duplicate keys DB_NORMAL DB_DUPNOK is the default value used The count member contains the number of column IDs in the hash table Member ID list htloc. If the value of count is zero (0), the htloc member is not allocated and should not be referenced. The htloc member contains the list of column IDs that comprise the hash table index. The htnm member contains the name of the hash table, if any. If the hash table does not have a name, htnm contains an empty string. 126 The RHLI Structures The UJRNINF Structure: Journal Information The UJRNINF structure has the following declaration. typedef struct { UJRNID ids; UTXID txid; long txdate; UTID tid; URID rid; short opknd; union{ IDML insert; UDML update; DDML delete; } txinfo; } UJRNINF; /* /* /* /* /* /* /* Custom log information. User ID information. Transaction ID. Timestamp of transaction. ID of affected table. ID of affected row. Context for the union: /* Insert DML information. /* Update DML information. /* Delete DML information. */ */ */ */ */ */ */ */ */ */ The UJRNINF structure is used by the ufchjrn function to retrieve information on the next update operation contained in the transaction journal. The UJRNID Structure The ids member of type UJRNID is a structure identifying the operating system user ID, database user ID, and schema ID of the user that performed the update operation. The UJRNID structure has the following declaration. typedef struct { int uid; UOID uoid; UAID uaid; } UJRNID /* /* /* /* User ID information. Unix User ID. Database User ID. Database authorization ID. */ */ */ */ The txid member is the transaction ID of the transaction that performed the update operation. The txdate member is the timestamp at which the transaction performing the update operation was committed. The txdate member is in the same format as a value returned by the operating system time function. The tid member is the table ID of the table updated. The RHLI Structures 127 The rid member is the row ID of the row updated. The opknd member identifies the type of operation and thereby determines the context for the txinfo union. Types of operations consist of insert, update, and delete. These are identified by the constants UINSOPR, UUPDOPR, and UDELOPR respectively and are defined as follows: #define UINSOPR 1 #define UUPDOPR 2 #define UDELOPR 3 The IDML Structure /* Insert DML information. /* New row’s UNIPCOL list. */ */ The update member of type UDML is a structure which contains pointers to UNIPCOL structures for the before and after images of the row updated. The UDML structure has the following declaration. typedef struct { UNIPCOLL *newucol; UNIPCOLL *olducol; } UDML; The DDML Structure */ */ */ The insert member of type IDML is a structure which contains a pointer to a UNIPCOL structure for the after image of the row inserted. The IDML structure has the following declaration. typedef struct { UNIPCOLL *newucol; } IDML; The UDML Structure /* txknd is an insert operation /* txknd is an update operation /* txknd is a delete operation /* Update DML information. /* New row’s UNIPCOL list. /* Old row’s UNIPCOL list. */ */ */ The delete member of type DDML is a structure which contains a pointer to a UNIPCOL structure for the before image of the row deleted. The DDML structure has the following declaration. typedef struct { UNIPCOLL *olducol; } DDML; /* Delete DML information. /* Old row’s UNIPCOL list. */ */ For a description of the UNIPCOL structure, see page 131. 128 The RHLI Structures The ULNKINF Structure: Link Information The ULNKINF structure has the following declaration. typedef struct { /* information always retrieved */ ULID lid; /* Link ID ULID remapid; /* logical Link ID (if remapped) */ */ /* information retrieval class: I (UCLASS1) */ long count; /* # of Column IDs in lists UCID * ploc; /* pointer to the parent list of /* Column IDs UCID * cloc; /* pointer to the child list of /* Column IDs */ */ */ */ */ /* information retrieval class: II (UCLASS2) */ char lnknm[U_NAMELEN + 1]; /* name of link } ULNKINF; */ The ULNKINF structure is used by the uinflnk function to retrieve information about a link. The lid member contains the link ID that the structure describes. If the ID represents a remapped link, the ID is the compile-time Link ID (the ID with which the user remapped the link). The remapid member contains the logical link ID that the structure describes. If the ID represents a remapped link, the ID is the runtime link ID (the actual Database ID). Otherwise, remapid is identical to the lid member. The count member contains the number of column IDs in the link member ID list ploc and cloc. If the value of count is zero (0), the ploc and cloc members are not allocated and should not be referenced. The ploc member contains the list of parent column IDs. The cloc member contains the list of child column IDs. The lnknm member contains the name of the link, if any. If the link does not have a name, the member contains an empty string. The RHLI Structures 129 The ULNKMAP Structure: Link ID Mapping The ULNKMAP structure has the following declaration. typedef struct { char * lnknm; ULID lid; ULID remapid; } ULNKMAP; /* pointer to the runtime link name /* compile-time Link ID /* runtime Link ID (returned) */ */ */ The ULNKMAP structure is used by the umaplnk function to specify information about runtime ID-mapping on a compile-time link object. It is highly recommended that you use the INILNKMAP mapping macro to initialize the ULNKMAP structure. The lnknm member contains the runtime name of the link. The lid member contains the compile-time Link ID. The Link ID (lid) can be hardcoded by the programmer (with a unique custom ID) or be supplied by the Unify DataServer C compiler, ucc. If you choose to hardcoded a unique Link ID (lid), be sure that the number you pick is compatible with the type ULID. To generate a unique compile-time Link ID using ucc, initialize lid with the following compiler directive (lnknm is the name of the hash table): $(ULNK:LID)lnknm$ The remapid member, upon successful completion of the ID-mapping operation, contains the runtime ID to which the compile-time ID was mapped. This value is for internal use only. 130 The RHLI Structures The UNIPCOLL Structure: Column Access Information The UNIPCOLL structure describes how and where column data is stored and manipulated and is used by the following functions: ualctbl uextint ufchlnk ufchrow ufchtbl uinsrow uintext upkfrow uupdrow The UNIPCOLL structure has the following declaration: typedef struct { UCID * cidl; UDBVAL * dbvall; int ncol; } UNIPCOLL; /* list of Column IDs /* list of column values /* # of entries in cidl/dbvall */ */ */ The dbvall member contains members that are structures and unions. The following diagram shows the components of the UNIPCOLL structure. The RHLI Structures 131 UNIPCOLL UCID 0 UCID 1 UCID 2 UDBVAL 1 UDBVAL 2 ... UCID n-1 UCID UDBVAL ncol (=n) UDBVAL 0 UNIP UNIP UNIP UOPTS UOPTS UOPTS UDBVAL n-1 ... UNIP UOPTS UNIP UTP_TIME UTP_DATE UTP_HDTE UTP_AMT UTP_HAMT UTP_INT UTP_HINT UTP_FLT UTP_GINT UTP_STRP UTP_GAMT UTP_VDDS Corresponding Column Data Buffer val scale vd_buf Data Buffer vd_filnm vd_bufnx vd_nbyts vd_mode vd_colnx Data File vd_rbyts vd_colsz TEXT/BINARY Column Buffer 132 The RHLI Structures The ncol member contains the number of entries in the cidl and dbvall lists. The number of entries in the list can be zero, and for variable-length data, can exceed the total number of columns that belong to the table. (See the description for the UTP_VDDS structure on page 136.) The cidl member contains a pointer to a list of column identifiers. The size of the list must be identical to the number of columns specified in the ncol member. All of the identifiers contained in the column identifier list must belong to the same table. If the ncol member is zero (0), the cidl member is not examined and may contain a null pointer. The dbvall member contains a pointer to a list of UDBVAL data value structures. The size of the list must be identical to the number of columns specified in the ncol member. If the member ncol is zero, the dbvall member is not examined and can contain a null pointer. The first entry in the column identifier list (cidl) corresponds to the first entry in the data value list (dbvall), and so forth. The UDBVAL Structure The UDBVAL structure describes the column’s data to be inserted, updated, or retrieved. Each entry in the data value structure list corresponds to the respective identifier entry in the column identifier list. The UDBVAL structure has the following declaration. typedef struct { UNIP unip; UOPTS valopts; } UDBVAL; /* pointer to column’s value buffer /* column value options */ */ The valopts member is a set of options that describe the data in the user data buffer. It is important to initialize valopts for both retrieval and insert/update operations. After retrieval operations, the system reinitializes the valopts member to describe the data that was placed in the user’s buffer. With insert and update operations, however, column values are only verified; valopts is not set. Warning Column values are verified and valopts is set for all fetch (or retrieval) operations unless UIGNORE is specified. This is true even if the application has set valopts explicitly. Fetch operations are upkfchrow, ufchlnk, ufchscn, ufchrow, ufchtbl, and ufchord. The RHLI Structures 133 The valopts member can contain the following options: UNORMAL value is valid & defined UNULLVAL value is null UIGNORE ignore column value; do not use URETVAL return initialized data value UNOTDEF reserved for user application use; not a valid option. Used to indicate that the value is undefined which is distinct from null. To specify more than one option, separate multiple options with the bitwise logical OR operator (|) in the argument list. The UNORMAL option indicates that the user buffer contains valid or null data. The UNULLVAL option indicates that the data in the user buffer is null. The UIGNORE option indicates that the system is to ignore this structure entry completely; no work is to be performed. The UIGNORE option is useful when the user want to selectively activate various columns in the UNIPCOLL structure without rebuilding the structure. Every column can have the UIGNORE option set; this is similar to setting the ncol member’s value to zero. The URETVAL option indicates that the system is to return to the user’s buffer the value of the data that was inserted into the database (after DIS default values have been applied). The URETVAL option is only applicable for insert and update operations. This option cannot be used for columns of type TEXT or BINARY. The URETVAL option can retrieve the default values applied to a column during an insert operation. This way, you do not have to perform a retrieval operation after the insert operation completes. If the URETVAL option is used for this purpose, the UIGNORE option must also be specified. Otherwise, the buffer is assumed to contain valid data, which might not be the case. The option UNORMAL is represented by the bit value 0 and, thus, cannot be used as a bit mask. 134 The RHLI Structures Data is defined to be either null or not null. Therefore, specifying the UNULLVAL option as a bit mask is appropriate, while specifying the UNORMAL option is not correct. For example, the expression valopts & UNORMAL is always false, while the expression valopts & UNULLVAL is true if the data is null, or, false, if the data is not null. The unip Union The unip member is of type UNIP which contains a union of pointers, one for each database data type. The unip points to the user’s buffer where the system either places retrieved data or acquires the data to be inserted or updated. The structure is a union of pointers. The compiler accepts the assignment of the actual pointer to a buffer of the column’s correct data type. The UNIP union has the following declaration. typdef union { UTP_TIME * UTP_DATE * UTP_HDTE * UTP_AMT * UTP_HAMT * UTP_GAMT * UTP_INT * UTP_HINT * UTP_GINT * UTP_FLT * UTP_STRP UTP_BYTP UTP_REAL * UTP_DBL * UTP_VDDS * } UNIP; timep; datep; ldatep; amtp; hamtp; gamtp; intp; hintp; gintp; fltp; strp; bytp; realp; dblp; vdp; /* /* /* /* /* /* /* /* /* /* /* /* /* /* /* ––> ––> ––> ––> ––> ––> ––> ––> ––> ––> ––> ––> ––> ––> ––> ”TIME” column buffer ”DATE” column buffer ”HUGE DATE” column buffer ”AMOUNT” column buffer ”HUGE AMOUNT” column buffer ”CURRENCY” column buffer ”INT” column buffer ”SMALL INT” column buffer ”HUGE INT(64-bit)” col buffer ”FLOAT” column buffer ”STRING” column buffer ”BYTE column buffer ”REAL column buffer ”DOUBLE column buffer ”TEXT/BINARY” column buffer */ */ */ */ */ */ */ */ */ */ */ */ */ */ */ All members of the UNIP union are pointers to a buffer which contains the column’s corresponding data. The pointer of the applicable column’s data type is assigned to a buffer of the same type; mixing pointer and buffer types is not recommended and may be flagged by the compiler as invalid. The RHLI Structures 135 Unify DataServer expects the data buffer identified by the UNIP member to be correctly aligned for the data type of the column. For example, if the column is of type FLOAT, the buffer to contain the float value should be correctly aligned for the FLOAT type. When amtp data is retrieved from a database, the data temporarily loses the scale of 2. Therefore, AMOUNT data stored in amtp variables change their scale to zero and record pennies rather than dollars. If you want to print or display the host variable value, you must divide the value by 100 to restore the default scale of 2. If the host variable value is being applied to a database, you do not need to divide the value by 100; Unify DataServer applies a precision of 2 to AMOUNT values. Except for UTP_VDDS, all members of the UNIP union are atomic data types. For example, int, long and char. UTP_VDDS is a structure that describes variable-length data for use with data type TEXT or BINARY columns. The UTP_VDDS Structure The UTP_VDDS structure has the following declaration. typedef struct { char * vd_buf; char * vd_filnm; long vd_bufnx; long vd_nbyts; int vd_mode; long vd_colnx; long vd_rbyts; long vd_colsz; } UTP_VDDS; /* /* /* /* /* /* /* /* data buffer file to use as data buffer starting offset in buffer # of bytes to fetch/store update mode starting offset in database col # of bytes actually read/written total size of database column */ */ */ */ */ */ */ */ The UTP_VDDS structure allows you to use a memory buffer or file to either store variable length data into or read from when manipulating variable-length data. If you choose to store or read the variable length data into or from a buffer, the member vd_buf must be set to point to the buffer that contains the variable length value. Also, when using a buffer to store or read variable length data to or from, the member vd_filnm must be set to (char *)0. If you want to store or read variable-length data into or from a file instead of a buffer, vd_buf must be set to (char *)0 and the member vd_filnm must point to a file name of an existing file that contains or will contain the variable-length data. 136 The RHLI Structures To assign a null value to a variable-length member, initialize an empty UTP_VDDS structure as follows: dbval.unip.valopts = UNULLVAL; dbval.unip.vds = (UTP_VDDS *)0; Variable text data can only be stored in standard Unix files (as opposed to devices). The vd_bufnx member specifies the byte offset from the beginning of the buffer or file that contains the variable length data. To read the buffer or file from the beginning, set vd_bufnx to 0. The vd_nbyts member specifies the number of bytes of data to read from or write to a buffer or file (specified by vd_buf or vd_filnm from offset position vd_bufnx). The number of bytes are counted beginning at offset position vd_colnx. When writing to a file (and you want to retrieve the entire var_len value), set vd_nbyts to a value that is large enough to fetch all bytes of data. If you do not know the size of the data, set vd_nbyts to an artificially high value, such as 2,147,483,647, to ensure that all data is retrieved. Warning Make sure that the starting offset plus the number of bytes requested (that is, vd_bufnx + vd_nbyts), does not exceed the size of the data buffer. Otherwise, if you read from a buffer that is too small, you will retrieve “garbage” data; or, if you write to a buffer that is too small, you will corrupt memory. Also, an error occurs if vd_bufnx + vd_nbyts exceeds: the actual file size when reading from a file the maximum file size when writing to a file To read a file from the vd_bufnx offset position to “end-of-file,” set vd_nbyts to –1. This option is used when you are reading data from a file for insert or update operations. The –1 option is used when reading from a file only; not to read from a buffer or the database. The vd_mode member specifies the update mode to use when operating on variable length data found in a buffer or file, and the data stored in a database column. The RHLI Structures 137 Valid vd_mode values are: V_TRUNC Truncates (or removes) a database column value starting at offset position vd_colnx to the end of the variable length column value. The removed data is replaced with vd_nbyts of data from a buffer or file. V_OVERWR Replaces (or overwrites) a database column value starting at position vd_colnx for a length of vd_nbyts with data from a buffer or file. V_APPEND Appends (or adds) vd_nbyts of data from a buffer or file to the end of the database column value. V_INSERT Inserts to a database column value, vd_nbyts of data from a buffer or file, before the byte pointed to by vd_colnx. V_DELETE Deletes (or removes) data from a database column value starting at position vd_colnx for a length of vd_nbyts. V_FETCH Fetches (or retrieves) vd_nbyts of data from a database column value starting at position vd_colnx for a length of vd_nbyts, and writes the retrieved data into a buffer or file. Specifying more columns than exist in the table (where the value of ncol exceeds the actual number of columns in the table), is known as overloading an operation. Overloading is most useful for inserting and updating TEXT/BINARY column operations, where the same column is specified multiple times and different vd_mode modes within a single database operation may be applied. The vd_colnx member specifies a byte offset from the beginning of the database column value. To set the offset position at the beginning of the database column value, set vd_colnx to 0. The vd_rbyts member contains the actual number of bytes read from or written to a buffer or file. This value is returned after the operation has completed. For update modes V_TRUNC, V_OVERWR, V_APPEND, V_INSERT, when the update is successful vd_rbyts reflects the value of vd_nbyts. 138 The RHLI Structures For V_DELETE operations, vd_rbyts contains the number of actual bytes deleted from the database column. For the fetch mode, V_FETCH, vd_rbyts will contain the actual number of bytes retrieved. Note that the number of bytes retrieved will be less than the requested number of bytes, vd_nbyts, if the end of the column value is reached. The vd_colsz member contains the number of bytes remaining in the database column after the database operation has completed. The RHLI Structures 139 The UNIPEXT Structure: Column Conversion Information The UNIPEXT structure is used by the uextint and uintext functions to specify column storage locations for conversion from internal to external formats (and vice versa). The UNIPEXT structure has the following declaration. typedef struct { UNIV univ; UOPTS valopts; } UNIPEXT; /* pointer to column’s external value /* column value options */ */ The UNIPEXT structure is laid out identically to the UDBVAL structure, except that the UNIP member is named UNIV. See the UDBVAL structure on page 133. 140 The RHLI Structures The UOACOLS Structure: Ordered Access Row Order The UOACOLS structure has the following declaration. typedef struct { UCID oacid; UOPTS oacopts; } UOACOLS; /* column identifier /* column sort option */ */ The UOACOLS structure is used by the ubegord function to specify the order to be placed on the rows of a table for ordered access. Specify one UOACOLS structures for each column to be part of the order. The oacid member contains the column ID of a column of the order specification. The oacopts member contains the order specification for the corresponding column. Options are: DB_ASCEND ascending column value DB_DESCND descending column value DB_NORMAL DB_ASCEND is the default value used The RHLI Structures 141 The URVKDS Structure: Revoke Privilege Descriptor The URVKDS structure has the following declaration. typedef struct { union { UAID aid; UOID uid; } gte; UCAPS rvkcaps; UOBJDS obj; } URVKDS; /* Grantee /* Privileges to be revoked /* Object to revoke */ */ */ The URVKDS structure is used by the udrpprv function to specify information about the privileges to revoke. The aid member contains the authorization ID from which authorization privileges are to be revoked. The uid member contains the user ID from which permissions are to be revoked. The gte member contains the ID of the object from which the privilege is to be revoked. The object to which privileges have been granted is known as the grantee. Because gte is a union, either a schema or a user can be revoked a privilege, but not both. The rvkcaps member contains the set of privileges that are to be revoked from the specified object. The set of privileges determines whether the type of the grantee is a schema or a user. To revoke authority, initialize the rvkcaps member as follows: 142 UDBA for DBA authority USCH for SCHEMA authority UDALL for both DBA and SCHEMA authority The RHLI Structures To revoke schema privileges to a schema, initialize the rvkcaps member as follows: USSEL for SELECT operations USINS for INSERT operations USUPD for UPDATE operations USDEL for DELETE operations UTALL for all operations (SELECT, INSERT, UPDATE and DELETE) If the schema privileges to a schema are revoked, all tables in the schema are inaccessible. To revoke schema privileges to a table, initialize the rvkcaps member as follows: UTSEL for SELECT operations UTINS for INSERT operations UTUPD for UPDATE operations UTDEL for DELETE operations UTALL for all operations (SELECT, INSERT, UPDATE and DELETE) To revoke schema privileges to a column, initialize the rvkcaps member as follows: UCSEL for SELECT operations UCINS for INSERT operations UCUPD for UPDATE operations UCALL for all operations (SELECT, INSERT and UPDATE) If you want to revoke more than one operation privilege for the same object, separate multiple options with the bitwise logical OR operator (|). The obj member is a structure (UOBJDS) that contains a description of the object that the privilege affects. The RHLI Structures 143 The UOBJDS Structure The UOBJDS structure has the following declaration. typedef struct { UDBOBJ objid; int objknd; } UOBJDS; /* object ID /* the type of object */ */ The objid member contains a database object ID whose type is described by the objknd member. Options are: UAID Authorization ID UDBID Database ID UCID Column ID UTID Table ID The objknd member contains a value that indicates what type of ID is contained in the objid member. Options are: UDBKND database object USCHKND schema object UTBLKND table object UVWKND view object UCOLKND column object If you are revoking a user privilege (DBA or SCHEMA authority), objid and objknd must be set to the current Database ID (see the uopndb function) and UDBKND, respectively. After the privileges revoke operation, the rvkcaps element (in the structure URVKDS), contains the privileges that were revoked. That is, if you revoke SELECT schema privilege from a table, the value contained in rvkcaps is equivalent to UTSEL. If you revoked SCHEMA authority from a user, rvkcaps contains a value equivalent to USCH. Additional Help For information about security and privileges, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 144 The RHLI Structures The UTBLINF Structure: Table Information The UTBLINF structure has the following declaration. typedef struct { /* information always retrieved */ UTID tid; /* table this struct describes UTID remapid; /* logical Table ID (if remapped) UAID aid; /* Authorization ID of this table */ */ */ /* information retrieval class: I (UCLASS1) */ UDEFID defid; /* Table Definition ID UDEFID chgid; /* Table Modification ID UMAPID mapid; /* Table Summation ID long segsz; /* desired segment size (bytes) long tblbase; /* direct table base value UOPTS tblopts; /* table processing options */ */ */ */ */ */ /* information retrieval class: II (UCLASS2) */ int ncol; /* # of columns in table UCID * ucidlst; /* pointer to table Column ID list int nbtree; /* # of btrees in table UBTID * ubidlst; /* pointer to table B-tree ID list int nhsh; /* # of hash-tables in table UHID * uhidlst; /* pointer to the table’s Hash /* Table ID list int nlnk; /* # of links in table ULID * ulidlst; /* pointer to the table’s Link ID /* list int nkey; /* # of columns in key column list UCID * keycidl; /* primary key Column ID */ */ */ */ */ */ */ */ */ */ */ */ /* information retrieval class: III (UCLASS3) */ int nvol; /* # of volumes specified for table UVID * uvidlst; /* pointer to table’s Volume ID list */ */ /* information retrieval class: IV (UCLASS4) */ long expnum; /* expected number of rows char * tbldesc; /* pointer to table description } UTBLINF; */ */ The UTBLINF structure is used by the uinftbl function to retrieve information about a table. The tid member contains the table ID that the structure describes. If the ID represents a remapped table, the ID is the compile-time table ID (the ID with which the user remapped the table). The RHLI Structures 145 The remapid member contains the logical table ID that the structure describes. If the ID represents a remapped table, the ID is the runtime table ID (that is, the actual object ID in the database). Otherwise, the value of remapid is identical to the tid member value. The aid member contains the authorization ID to which this table belongs. The defid member contains the table’s current Definition ID. The Definition ID changes whenever a physical modification is made to the table, such as adding or dropping columns. The chgid member contains the table’s current Change ID. The Change ID changes whenever a logical modification is made to the table, such as modifying the table’s name. The mapid member contains the table’s current Mapping ID. The Mapping ID serves as a checksum of the table’s physical attributes, such as the columns contained in the table. The segsz member contains the table’s designated segment size, if any, in bytes. The tblbase member contains the table’s base value, if the table has been specified as being a direct table. Otherwise, the value of tblbase is zero (0). The tblopts member contains one or more of the following options: DB_ORDNRY create ordinary table DB_DIRECT create directĆkeyed table DB_RDONLY table is readĆonly DB_SCATTR scatter data across volumes DB_USEVOL must use specified preference volumes DB_FIXSIZE table size is fixed at value specified by expnum DB_KEYED create a primary key DB_NORMAL DB_ORDNRY is the default value used The ncol member contains the number of columns in the table column ID list ucidlst. If the value of ncol is zero (0), the ucidlst member is not allocated and should not be referenced. 146 The RHLI Structures The ucidlst member contains a list of column IDs that belong to the table. The list is sorted in the chronological order in which the columns were added to the table. The nbtree member contains the number of B-trees in the B-tree ID list ubidlst. If the value of nbtree is zero (0), the ubidlst member is not allocated and should not be referenced. The ubidlst member contains a list of B-tree IDs that belong to the table. The nhsh member contains the number of hash tables in the hash table ID list uhidlst. If the value of nhsh is zero (0), the uhidlst member is not allocated and should not be referenced. The uhidlst member contains a list of hash table IDs that belong to the table. The nlnk member contains the total number of links in the Link ID list ulidlst. If the value of nlnk is zero (0), the ulidlst member is not allocated and should not be referenced. The ulidlst member contains a list of Link IDs that belong to the table. Note that the list contains both parent and child Link IDs, and does not differentiate between the two. The nkey member contains the number of columns in the table primary key list keycidl. If the value of nkey is zero (0), the keycidl member is not allocated and should not be referenced. The keycidl member contains a list of the column IDs that constitute the designated primary key for the table. The IDs are sorted in their designated order. The nvol member contains the number of volumes in the volume preference list uvidlst. If the value of nvol is zero (0), the uvidlst member is not allocated and should not be referenced. The uvidlst member contains a list of the volume IDs that have been preferenced for the table. The expnum member contains the expected number of records, if any, that was designated when the table was created. The tbldesc member contains the table’s variable length description, if any. The RHLI Structures 147 The UTBLMAP Structure: Table ID Mapping The UTBLMAP structure has the following declaration. typedef struct { char * tblnm; UTID tid; UTID remapid; UDEFID defid; UMAPID mapid; } UTBLMAP; /* /* /* /* /* pointer to the runtime table name compile-time Table ID runtime Table ID (returned) compile-time table Definition ID compile-time table Mapping ID */ */ */ */ */ The UTBLMAP structure is used by the umaptbl function to perform runtime ID-mapping on a compile-time table object. The tblnm member contains the runtime name of the table. The tid member contains the compile-time table ID. The table ID (tid) can be hardcoded by the programmer (with a unique custom ID) or be supplied by the Unify DataServer C compiler, ucc. If you choose to hardcoded a unique table ID (tid), be sure that the number you pick is compatible with the type UTID. To generate a unique compile-time table ID using ucc, initialize tid with the following compiler directive (tblnm is the name of the table): $(UTBL:TID)tblnm$ The remapid member, upon successful completion of the ID-mapping operation, contains the runtime ID to which the compile-time ID was mapped. This value is for internal use only. The defid member contains the table’s compile-time Definition ID, which is a time-stamped ID that is used by Unify DataServer to verify the accuracy of remapid. The Definition ID (defid) must be initialized with the following compiler directive syntax (tblnm is the name of the table): $(UTBL:DEFID)tblnm$ 148 The RHLI Structures The mapid member contains the table’s compile-time Mapping ID, which is a checksum value that is used by Unify DataServer to verify the accuracy of remapid. This value is for internal use only. The Mapping ID (mapid) must be initialized with the following compiler directive syntax (tblnm is the name of the table): $(UTBL:MAPID)tblnm$ Tip It is highly recommended that you use the INITBLMAP mapping macro to initialize the UTBLMAP structure. The RHLI Structures 149 The UTXINF Structure: Transaction Information The UTXINF structure has the following declaration. typedef struct { /* information always retrieved */ UTXID txid; /* Transaction ID int logflag; /* is logging on (1=yes/0=no) /* information long nupd; int txtyp; int state; int lckphas; int nlck; int nlgpg; int opsknd; } UTXINF; */ */ retrieval class: I (UCLASS1) */ /* number of update operations /* transaction type /* current state of the transaction /* Lock phase ascend/descend /* number of locks held by transact. /* number of log pages written /* kind of operations in tx */ */ */ */ */ */ */ The UTXINF structure is used by the uinftx function to retrieve transaction information. The txid member contains the transaction ID described by this structure. The logflag member contains a Boolean value that indicates whether transaction logging is turned on for the user. A value of one (1) indicates that transaction logging is on, a value of zero (0) indicates that transaction logging is off. The nupd member contains the count of the number of update operations that have been performed on behalf of this transaction. An update operation is one that does not retrieve data. The txtyp member contains the transaction type. Options are: 150 UTXTYPE0 Type 0 transactionĊ read only UTXTYPE1 Type 1 transactionĊupdate, limited concurrency UTXTYPE2 Type 2 transactionĊupdate, twoĆphase locking The RHLI Structures The state member contains the current state of the system. Options are: UTXSTRT Start transaction UTXACT Transaction active UTXSUSP Suspend transaction UTXCMT Commit transaction UTXPREP Prepare transaction UTXCMPL Complete transaction UTXEND End transaction UTXABRT Abort transaction The lckphas member contains an indicator of what phase the transaction is currently operating under, if the transaction is a type 2 transaction (two-phase locking). A value of 1 indicates that the transaction is in the unlock phase, a value of 0 (zero) indicates that the transaction is in the lock phase. The nlck member contains the count of the number of row locks, both shared and exclusive, currently held by the transaction. The nlgpg member contains the count of the number of log buffers written to the transaction log for this transaction. This value does not include any log buffers maintained in shared memory. The RHLI Structures 151 The UTXOPT Structure: Begin Transaction Options The UTXOPT structure has the following declaration. typedef struct { long cmtfreq; int txknd; UOPTS waitflg; } UTXOPT; /* # of operations before batch commit */ /* Transaction kind */ /* wait to start transaction log */ The UTXOPT structure is used by the ubegtx and ucbgtx functions to specify options for beginning a transaction. The cmtfreq member specifies the number of operations after which the current batch transaction is to be committed. Also see the description of the UBATCH option of waitflg. The txknd member specifies the transaction level. Options are: UTXTYPE2 Transactions enforce any shared or exclusive locks acquired by the transaction. No locks are released until the transaction ends. UTXTYPE1 Transactions also enforce any shared or exclusive locks acquired by the transaction; however, shared locks can be released before the transaction ends. UTXTYPE0 Transactions are special purpose transactions that allow reading of uncommitted data. No updates are allowed for a transaction of this type. UTXTYPE0 is the only valid value for txknd when the database has been opened in readĆonly mode (see the uopndb function). 152 The RHLI Structures The waitflg member contains a value that controls batch transactions and whether transactions should wait for a resource to become available or to return immediately in case of insufficient resources. Options are: UAUTO Commit a batch transaction if resources are scarce UBATCH Transaction is to operate in batch mode; the cmtfreq member must be specified UCOMMIT Commit transaction on the uclsdb function UNOWAIT Do not wait for a resource before starting a transaction UWAIT Wait for an available resource before starting a transaction DB_NORMAL UNOWAIT is the default value used If you specify the UBATCH option, the UCOMMIT option is implicitly specified. UAUTO can be used in cases where you want to allow the system to react to influences from other users that affect database resources. For example, if shared memory is more than 75% full, or the transaction log space cannot be reserved, the batch transaction is implicitly committed if UAUTO is set. If you specify the UAUTO option, you must also specify the UBATCH option. UBATCH specifies that the transaction operates in batch mode.. The RHLI Structures 153 The UUSRINF Structure: User Information The UUSRINF structure has the following declaration. typedef struct { /* information always retrieved */ UOID uoid; /* User/Owner ID */ /* information retrieval class: I (UCLASS1) */ UAID aid; /* user default Authorization ID UAID curaid; /* user current Authorization ID */ */ /* information retrieval class: II (UCLASS2) */ int nauth; /* # of authorizations in ’aidlst’ UAID * aidlst; /* pointer to list of valid Auth IDs /* (schemas) for user */ */ */ /* information retrieval class: III (UCLASS3) */ char usernm[U_USRNMLN + 1]; /* pointer to user name } UUSRINF; */ The UUSRINF structure is used by the uinfusr function to retrieve information about a user. The uoid member contains the user ID described by the information in this structure. The aid member contains the authorization ID of the user’s default schema upon entry to the system. The curaid member contains the authorization ID of the user’s current schema. The nauth member contains the number of entries in the authorization list aidlst. If the value of nauth is zero (0), aidlst has not been allocated and should not be referenced. The aidlst member contains a list of authorization IDs to which the user has been granted access. The usernm member is the name of the user as it would be referenced for name-binding. 154 The RHLI Structures The UVOLINF Structure: Volume Information The UVOLINF structure has the following declaration. typedef struct { /* information always retrieved */ UVID vid; /* Volume ID UVID remapid; /* logical Volume ID (if remapped) */ */ /* information retrieval class: I (UCLASS1) */ long voloff; /* volume offset (bytes) long vollen; /* volume length long volused; /* allocated (used) space long volpgsz; /* volume page size (bytes) UOPTS volopts; /* volume processing options */ */ */ */ */ /* information retrieval class: II (UCLASS2) */ int ntbl; /* # of tables clustered in volume UTID * utidlst; /* pointer to the clustered Table /* ID list */ */ */ /* information retrieval class: III (UCLASS3) */ char volfile[U_FILELEN + 1]; /* pointer to the volume /* file name char volname[U_NAMELEN + 1]; /* volume name, if any } UVOLINF; */ */ */ The UVOLINF structure is used by the uinfvol function to retrieve information about a volume. The vid member contains the volume ID that the structure describes. If the ID represents a remapped volume, the ID is the compile-time volume ID (the ID with which the user remapped the volume). The remapid member contains the logical volume ID that the structure describes. If the ID represents a remapped volume, the ID is the runtime volume ID (the actual Database ID). Otherwise, the remapid value is identical to the vid member value. The voloff member contains the volume’s initial offset in the volume’s device. If the volume is a regular or contiguous file, voloff is zero (0). A non-zero value indicates the number of bytes into the device at which the device can be found. The RHLI Structures 155 The vollen member indicates the length of the volume. If DB_BIG is set in the volopts member, vollen represents the number of pages. Otherwise, it represents the number of bytes; this value is converted internally into the corresponding number of pages. The volume length is used by volume storage routines to determine how much data can fit onto the disk partition. If the volume is a contiguous file or a device, vollen contains a non-zero value. For a regular file, a zero value (0) indicates that the volume size is unconstrained and can dynamically grow as needed. The volused member contains the amount of the volume that is allocated. The DB_BIG member controls whether this value is specified as pages or bytes, as for vollen. The volpgsz member contains the volume’s page size, in bytes. The volopts member contains the volume’s processing options. Options are: DB_ORDNRY volume is ordinary DB_DEVICE volume is a device DB_RDONLY reserved for future use DB_FORCE force volume preĆallocation DB_ALLOC volume is a preĆallocated file DB_BIG Indicates that the vollen specification is in pages rather than bytes. DB_CONTIG reserved for future use DB_NORMAL DB_ORDNRY is the default value used Your application should test the DB_BIG bit to determine the volume size. For example: 156 The RHLI Structures UVOLINF *volinfl; /* ––> vols info list */ int inx; /* its for–loop index */ ... if ( (volinfl+inx)–>volopts & DB_BIG ) { /* lengths are # of 2k pages */ printf(”vollen = %ld KB, ”, (volinfl+inx)–>vollen * ((volinfl+inx)–>volpgsz/1024)); printf(”volused = %ld KB.”, (volinfl+inx)–>volused * ((volinfl+inx)–>volpgsz/1024)); } else /* volume is not > 2 GB */ { printf(”vollen = %ld bytes, ”, (volinfl+inx)–>vollen); printf(”volused = %ld bytes.”, (volinfl+inx)–>volused); } ... The ntbl member contains the number of entries in the table ID list utidlst. If ntbl contains a zero value (0), utidlst has not been allocated and should not be referenced. The utidlst member contains a list of all of the table IDs that have preferenced this volume. The table’s processing options need to be examined to determine if the volume is a required or advisory preference. The volfile member contains the full path name of the volume file or device. The volname member contains the name of the volume as it would be referenced for name-binding. The RHLI Structures 157 The UVOLMAP Structure: Volume ID Mapping The UVOLMAP structure has the following declaration. typedef struct { char * volnm; UVID vid; UVID remapid; } UVOLMAP; /* pointer to runtime volume name /* compile-time Volume ID /* runtime Volume ID (returned) */ */ */ The UVOLMAP structure is used by the umapvol function to perform runtime ID-mapping on a compile-time volume object. It is highly recommended that you use the INIVOLMAP mapping macro to initialize the UVOLMAP structure. The volnm member contains the runtime name of the volume. The vid member contains the compile-time volume ID. The volume ID (vid) can be hardcoded by the programmer (with a unique custom ID) or be supplied by the Unify DataServer C compiler, ucc. If you choose to hardcoded a unique volume ID (vid), be sure that the number you pick is compatible with the type UVID. To generate a unique compile-time volume ID using ucc, initialize vid with the following compiler directive (volnm is the name of the volume.): $(UVOL:VID)volnm$ The remapid member, upon successful completion of the ID-mapping operation, contains the runtime ID to which the compile-time ID was mapped. This value is for internal use only. 158 The RHLI Structures Function Reference 159 160 Function Reference uabttx Abort a transaction Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uabttx( txid, status ) UTXID txid; USTATUS *status; Arguments Description txid Transaction ID status Returned status of the function Aborts the specified active transaction and rolls back all database operations performed during the transaction. Database operations include insert, delete, and update operations. An active transaction is a transaction that has not been committed or aborted. To use this function, transaction logging must be on (that is, the LOGTX configuration variable must be set to TRUE). This function releases all locks associated with the transaction and closes all scans active in the transaction. Warning Do not abort a transaction by calling the UNIX kill command. This command can cause irreversible damage to the database. Use the uabttx function to abort a transaction. Security No privileges needed. 161 Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also ubegtx, ucbgtx, ucmttx, uinftx For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 162 uaddath Add a schema to a database Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uaddath( txid, nauth, dbauth, statusl, status ) UTXID txid; int nauth; DBAUTH *dbauth; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nauth Number of schemas in the dbauth list dbauth List of DBAUTH structures, each describing a separate schema to add statusl Returned list of status values, one for each dbauth entry status Returned status of the function Validates the schemas described in the dbauth list and adds the schemas to the database. Each DBAUTH structure in the dbauth list contains information about a schema. For a complete description of the DBAUTH structure, see page 80. Security To add a schema, you must have one of the following permissions. DBA authority SCHEMA authority 163 To grant DBA and SCHEMA authority, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also udrpath, uallath, ubndath, ufchath, uaddprv, uinfath, umapath, umodath, unumath For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 164 uaddbt Define a B-tree index Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uaddbt( txid, nbt, dbbtree, statusl, status ) UTXID txid; int nbt; DBBTREE *dbbtree; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nbt Number of B-trees in the dbbtree list dbbtree List of DBBTREE structures, each defining a separate B-tree to add statusl Returned list of status values, one for each dbbtree entry status Returned status of the function Validates the B-trees defined in the dbbtree list and adds the B-trees to the database. B-trees facilitate range searches and random access on one or more columns. B-trees can be built referencing a composite key made up of the maximum number of columns allowed (128.) B-trees are created by specifying any column or combination of columns in a database table as the indexes, except those columns that are of the data type TEXT (U_VTXT) or BINARY (U_VBIN). Each DBBTREE structure in the dbbtree list contains information about a B-tree. 165 For a description of the DBBTREE structure, see page 81. Security To define a B-tree for a table, the table must be in the current schema and you must have one of the following permissions. DBA authority SCHEMA authority To make a schema current, use the ufchath function. To grant DBA and SCHEMA authority, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also udrpbt, uinfbt, umodbt For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about using B-trees, see “Choosing an Access Method” and “Managing the Database Files” in Unify DataServer: Managing a Database. 166 uaddcgp Add a column group to a table Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uaddcgp( txid, ncgrp, dbcgrp, statusl, status ) UTXID txid; int ncgrp; DBCGRP *dbcgrp; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID ncgrp Number of column groups in the dbcgrp list dbcgrp List of DBCGRP structures, each describing a column group to add statusl Returned list of status values, one for each dbcgrp entry status Returned status of the function Validates the column groups listed in dbcgrp and adds the column groups to a table in the database. If a column group already exists, this function combines the requested options with the existing options for that column group. Each DBCGRP structure in the dbcgrp list contains information about a column group. For a description of the DBCGRP structure, see page 83. If the add operation is successful, a column group ID is returned (to DBCGRP member grpcid) for each added column group in the dbcgrp list. You cannot specify uniqueness constraints on a column if the column contains data. 167 Security To add a column group to a table, the table must be in the current schema and you must have one of the following permissions. DBA authority SCHEMA authority To make a schema current, use the ufchath function. To grant DBA and SCHEMA authority, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example The following example uses the uaddcgp function to define a column group for a table. /* initialize the DBCGRP structure */ if ( uaddcgp(txid, ncgrp, dbcgrp, statusl, &status) != USUCCESS ) printf(”Couldn’t add column group: status = %ld\n”, status); See Also udrpcgp, uinfcgp For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 168 uaddcnm Add a column name (or synonym) Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uaddcnm( txid, ncol, dbcolnm, statusl, status ) UTXID txid; int ncol; DBCOLNM *dbcolnm; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID ncol Number of column names in the dbcolnm list dbcolnm List of DBCOLNM structures, each describing a separate column name to add statusl Returned list of status values, one for each dbcolnm entry status Returned status of the function Validates the column names listed in dbcolnm and adds the column names to the database. Once a synonym for a column has been added, name binding operations can be performed by specifying the new column name. Column groups can also be assigned synonyms by specifying the column group ID (in place of the column ID) in the dbcolnm list. Each DBCOLNM structure in the dbcolnm list contains information about a column’s name. 169 For a description of the DBCOLNM structure, see page 85. Column names must be unique within a table. Security To add a column name to a column, the table containing the column must be in the current schema, and you must have one of the following permissions. DBA authority SCHEMA authority To make a schema current, use the ufchath function. To grant DBA and SCHEMA authority, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the uaddcnm function to add a column synonym to the part_number column. dbcolnm.cid = $parts.part_number$; dbcolnm.colnm = ”p_no”; ... if (uaddcnm(txid, 1, &dbcolnm, statusl, &status) != USUCCESS ) { fprintf(stderr, ”Error naming column: status = %ld\n”,status); } See Also udrpcnm For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 170 uadddb Create a new, empty database Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uadddb( dbinfo, status ) DBINFO *dbinfo; USTATUS *status; Arguments Description dbinfo Pointer to a DBINFO structure that describes the database to create status Returned status of the function Creates a root volume and a new, empty database. The new database is described with a DBINFO structure which contains root volume information and other database creation options. For a description of the DBINFO structure, see page 88. Before you create a database: If you want the new database to have a configuration file during the creation process, its configuration file must exist before you create the database. However, anytime you open the database, it will apply the configuration variables defined in the configuration file, even if you created that file after the creating the databse. You can optionally choose a new set of shared memory keys for the database by specifying the keys in the new database’s configuration file. The advantage of choosing a new key is that if one database is corrupted (by a killed process, for example), the corrupted database does not affect other databases that have separate shared memory segments. The disadvantage is that if many databases exist at the same time, they require many shared memory segments. 171 Security To create a new database, you need no special privileges. However, if you want to overwrite an existing database, you must have DBA authority for the database you are overwriting. (An existing database is overwritten by setting the member dbopts in the structure DBINFO to DB_OVERWR.) To grant DBA authority, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the uadddb function to create a database. USTATUS status; DBINFO dbinfo; DBVOLM dbvolm; /* initialize values */ dbvolm.volfile = ”/usr/db/parts.db”; dbvolm.volopts = DB_BIG; dbvolm.voloff = 0; dbvolm.vollen = 20480; /* actual size is 41943040 bytes */ dbvolm.volpgsz = 2048; /* 2k is only supported page size */ dbinfo.rootvol = &dbvolm; dbinfo.descrpt = ”parts database”; dbinfo.dbopts = (DB_CONFIG | DB_OVERWR); /* create the database */ if ( uadddb(&dbinfo, &status) != USUCCESS) fprintf(stdout, ”creation failed; status: %ld\n”, status); See Also uopndb, umoddb, uaddvol For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 172 uaddhsh Define a hash table Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uaddhsh( txid, nhsh, dbhashl, statusl, status ) UTXID txid; int nhsh; DBHASH *dbhashl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Current transaction ID nhsh Number of hash tables in the dbhashl list dbhashl List of DBHASH structures, each describing a separate hash table to add statusl Returned list of status values, one for each dbhashl entry status Returned status of the function Validates the hash tables defined in the dbhashl list and adds the hash tables to the database. Hash table indexes on a set of columns in a table provide very fast access when ordering is not needed. Each DBHASH structure in the dbhashl list contains information about a hash table. For a description of the DBHASH structure, see page 86. To add a hash table to a table that already contains rows, you must first perform a complete scan of the table before adding the hash table. This can be a lengthy operation, depending upon the amount of data in the table. 173 Security To define a hash table index for a table, the table must be in the current schema and you must have one of the following permissions. DBA authority SCHEMA authority To specify a volume for a hash table (the vid member in the DBHASH structure), you must have DBA authority. To make a schema current, use the ufchath function. To grant DBA and SCHEMA authority, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the uaddhsh function to create a hash table index on the manf table. colidl[0] = $manf.mano$; dbhash.htname = “#mano_ht”; dbhash.ncol = 1; dbhash.colidl = colidl; dbhash.htopts = DB_NORMAL; ... if ( uaddhsh(txid, 1, &dbhash, statusl, &status) != USUCCESS ) { printf(”Error adding a hash index: status = %ld\n”, status); printf(”hash index status = %ld\n”, *statusl); } See Also udrphsh, uinfhsh, umodhsh For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about using hash tables, see “Managing the Database Files,” “Choosing an Access Method,” and “Tuning the Access Methods” in Unify DataServer: Managing a Database. 174 uaddlnk Define a link Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uaddlnk( txid, nlnks, dblinkl, statusl, status ) UTXID txid; int nlnks; DBLINK *dblinkl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nlnks Number of links in the dblinkl list dblinkl List of DBLINK structures, each describing a separate link to add statusl Returned list of status values, one for each dblinkl entry status Returned status of the function Validates the link descriptions listed in dblinkl and adds a link for a pair of tables. Each DBLINK structure in the dblinkl list contains information about a link. For a description of the DBLINK structure, see page 90. A link is composed of a parent column (or set of columns) and a corresponding child column (or set of columns). When a link exists between two tables, each row in the child table is associated with a row in the parent table. The association is based on the linked columns’ values. The child row is associated to the parent row that has the same values in the linked columns. 175 If a particular column value does not exist in any parent table row, then you cannot insert a child row with that column value. This characteristic makes links useful for enforcing referential integrity. Also, if a null value is allowed in a child table column, the column will not be associated with a parent column. Performance The link access method can also significantly increase the performance of retrieval operations requiring related information from different tables. If the link is on more than one column, then the order of column IDs in the lists indicates which columns are to be linked between the two tables. A column can be specified only once per link. Each column in the child list must have the same type as the corresponding column in the parent list. Variable-length columns (of data type U_VTXT or U_VBIN) cannot be used in links. A column or set of columns can be the child of one link only. To add a link to a table that already contains rows, you must first perform a complete scan of the table before adding the link. This can be a lengthy operation, depending upon the amount of data in the table. Security To define a link index for a table, you must have one of the following permissions. DBA authority SCHEMA authority Also, the two tables comprising the link definition must be made accessible by one of these methods: Make the schema that contains both linked tables current. If the tables are located in two separate schemas, make the schema that contains the first table current and have schema access permission to the schema that contains the second table. To make a schema current, use the ufchath function. To grant DBA and SCHEMA authority, use the uaddprv function. To grant schema access permission, see the uaddprm function. 176 Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the uaddlnk function to define a link index on columns in a table. pcid = $parts.part_number$; ccid = $orders.p_no$; ... dblink.lnkname = ”@part_num_link”; dblink.ptblid = $parts$; dblink.ctblid = $orders$; dblink.ncol = 1; dblink.pcolidl = &pcid; dblink.ccolidl = &ccid; ... if ( uaddlnk(txid, 1, &dblink, statusl, &status) != USUCCESS ) { printf(”Error adding link: status = %ld\n”, status); printf(”link status = %ld\n”, *statusl); } See Also udrplnk, umodlnk, uinflnk For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about using links, see “Choosing an Access Method” in Unify DataServer: Managing a Database. 177 uaddprf Add a volume preference for a table #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uaddprf( txid, tid, nvol, vidl, statusl, status ) UTXID txid; UTID tid; int nvol; UVID *vidl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of table to add volume preference nvol Number of volumes in the vidl list vidl List of volume IDs, each ring to a separate volume pence to add statusl Returned list of status values, one for each vidl entry status Returned status of the function Registers the volume preferences designated by the list of volumes for the specified table. The table ID (tid) and volume IDs (listed in vidl) must be valid existing entries in the database before calling the uaddprf function. Security To add volume preferences for a table, you must be located in the schema that contains the table, and you must have DBA authority. 178 To make a schema current, use the ufchath function. To grant DBA and SCHEMA authority, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the uaddprf function to add a volume preference to a table. /* Add the volume preferences */ if ( uaddprf(txid, tid, nvol, vidl, statusl, &status) != USUCCESS) { printf(“Unable to add volume preference: %ld\n”, status); for ( i = 0; i < nvol; ++i ) printf(“Entry %d volume: %ld status: %ld\n”, i, vidl[i],statusl[i]); uexit(1); } See Also uadddb,uaddtbl For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about managing database volumes, see “Managing the Database Files” in Unify DataServer: Managing a Database. 179 uaddprm Grant schema access permission #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uaddprm( txid, userid, nauth, authl, statusl, status ) UTXID txid; UOID userid; int nauth; UAID authl[]; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID userid User ID of user to be granted schema access permission nauth Number of authorization IDs (associated with schemas) listed in the authl list authl List of authorization IDs, each referring to a schema for which a user (userid) is to be granted schema access permission statusl Returned list of status values, one for each authl entry status Returned status of the function Grants a user schema access permission to one or more schemas. Schema access permission allows a user (specified by userid) to access the schema (specified in authl). When a user is first added to the database (with the uaddusr function), the user is granted schema access permission to the PUBLIC schema only. To access other schemas in the database, you must use this function to grant schema access permission for each schema the user needs to access. 180 To revoke schema access permission from a user, use the udrpprm function. Security To grant schema access permission to a user, you must DBA authority or both of the following must be true: The schema must be the current schema. You have SCHEMA authority. To make a schema current, use the ufchath function. To grant DBA and SCHEMA authority, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example calls the uaddprm function to grant schema access permission to user lcc: /* define required parameters */ UTXID txid; UOID userid; int nauth; UAID authl; USTATUS statusl; USTATUS * status; /* Retrieve authorization ID (authl) */ ... /* Grant schema access permission */ nauth = 1; userid = ”lcc”; if ( uaddprm(txid, userid, nauth, authl, statusl, &status) != USUCCESS) { printf(“Unable to add privileges: %ld\n”, status); uexit(1); } See Also udrpprm, uaddusr, udrpusr, uaddath, udrpath For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 181 uaddprv Grant user authority or schema privileges #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uaddprv( txid, nrqstds, rqstdsl, statusl, status ) UTXID txid; int nrqstds; UGRTDS rqstdsl[]; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nrqstds Number of privileges described in the rqstdsl list rqstdsl List of UGRTDS structures, each describing a separate privilege to grant statusl Returned list of status values, one for each rqstdsl entry status Returned status of the function Grants one or more of the following permissions: DBA authority SCHEMA authority schema privileges to a schema schema privileges to a table schema privileges to a column DBA authority allows a user to create, drop, access, and modify any object in the database. SCHEMA authority, however, allows a user to create and drop schemas and objects within those schemas only. Both of these privileges are granted to users only. 182 Schema privileges are privileges granted to a schema. The privileges specify access to all objects in another schema or a table or column within the schema. For each set of privileges you want to grant with this function, you must first describe the privileges by initializing the UGRTDS structure. For a description of the UGRTDS structure, see page 120. If a schema grants schema privileges to another schema, all tables in the schema granting the privilege are accessible. If a schema grants schema privileges to a table, only the specified table is accessible. To revoke user or schema privileges, use the udrpprv function. Security To grant DBA or SCHEMA authority, you must have DBA authority. To grant schema privileges to a schema, you must make current the schema that is granting the privileges. To grant schema privileges to a table or column, you must make the schema that contains the table or column to receive the privileges the current schema. To make a schema current, use the ufchath function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example calls the uaddprv function to grant select and insert privileges on the company table to the admin schema: /* define required parameters */ UTXID txid; USATUS status; /* Retrieve authorization ID (aid) for admin schema */ ... /* Retrieve table ID for company table (tidl) in SQL_books schema */ ... /* Initialize UGRTDS structure elements */ rqstdsl.gte.aid = aid; rqstdsl.rqstcaps = UTSEL | UTINS; rqstdsl.obj.objid = tidl.tid; rqstdsl.obj.objknd = UTBLKND; /* Grant privileges */ if ( uaddprv(txid, 1, &rqstdsl, statusl, &status) != USUCCESS) printf(”Unable to grant privileges: status = %ld\n”,status); 183 See Also uaddath, uaddprm, udrpath, udrpprm, udrpprv For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 184 uaddtbl Add a table Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uaddtbl( txid, ntbl, dbtable, statusl, status ) UTXID txid; int ntbl; DBTABLE *dbtable; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID ntbl Number of tables described in the dbtable list dbtable List of DBTABLE structures, each describing a separate table to add statusl Returned list of status values, one for each dbtable entry status Returned status of the function Validates the table described in each DBTABLE structure, pointed to by dbtable, and add the table to the database. For a description of the DBTABLE structure, see page 92. If the add operation is successful, this function returns a table ID (into DBTABLE member tid) for each added table in the dbtable list. Storage space for rows within the added table are allocated either on the default volume or on the preferred volume specified by DBTABLE member vid. The order of the columns added to the table is undefined. The column order returned from the uinftbl routine may differ from the order specified in the DBCOL list (contained in the DBTABLE structure). 185 Security To add a table to a schema, you schema where you want the table to reside must be the current schema, and you must have one of the following authority levels: DBA authority SCHEMA authority To specify a volume for a table (the vid member in the DBTABLE structure), you must have DBA authority. To make a schema current, use the ufchath function. To grant DBA and SCHEMA authority, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the uaddtbl function to add a table. ncol = 2; /* number of DBCOL column entries */ dbcol[0].coltyp = U_HINT; /* column type */ dbcol[0].collen = ULD_HINT; /* column format precision */ dbcol[0].colscl = 0; /* column format scale */ dbcol[0].dsplen = UDL_HINT; /* column display precision */ dbcol[0].dspscl = 0; /* column display scale */ dbcol[0].colopts = DB_NORMAL; /* column processing options */ dbcol[0].colnm = ”part_number”; /* column name */ dbcol[0].coldesc = ”part number”; /* column description */ dbcol[0].dsppict = ””; /* column display picture clause */ dbcol[1].coltyp = U_STR; dbcol[1].collen = UDL_STR; dbcol[1].colscl = 0; dbcol[1].dsplen = UDL_HINT; dbcol[1].dspscl = 0; dbcol[1].colopts = DB_NORMAL; dbcol[1].colnm = ”part_name”; dbcol[1].coldesc = ”part name”; dbcol[1].dsppict = ””; /* ntbl = 1; 186 /* column type */ /* column display length */ /* column format scale */ /* column display precision */ /* column display scale */ /* column processing options */ /* column name */ /* column description */ column display picture clause */ dbtable[0].dbcol = dbcol; /* array of column information */ dbtable[0].vid = UNULLVID; /* initial pence volume */ dbtable[0].expnum = 10000L; /* expected number of rows */ dbtable[0].statusl = statusl; /* array of column status codes */ dbtable[0].tblopts = DB_NORMAL; /* table processing options */ dbtable[0].ncol = ncol; /* # of columns in ’dbcol’ */ dbtable[0].tblnm = ”parts index”; /* table name */ dbtable[0].tbldesc = ”part index”; /* table description */ /* Add the tables */ if (uaddtbl(txid, ntbl, dbtable, statusl, &status) != USUCCESS) { fprintf(stderr,”Unable to create new table: %ld\n”, status); for ( i = 0; i < ntbl; ++i ) for ( j = 0; j < dbtable[i].ncol; ++j ) printf(”Table %d column %d status: %ld\n”, i, j, dbtable[i].statusl[j]); uexit(1); } See also Appendix A for more examples. See Also udrptbl, uinftbl For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 187 uaddtnm Add a table name (or synonym) #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uaddtnm( txid, ntbl, dbtblnm, statusl, status ) UTXID txid; int ntbl; DBTBLNM *dbtblnm; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID ntbl Number of table names in the dbtblnm list dbtblnm List of DBTBLNM structures, each describing a separate table name to add statusl Returned list of status values, one for each dbcolnm entry status Returned status of the function Adds an alternate name (or synonym) for the table specified by a DBTBLNM structure. Synonyms for a table can be used to reference (and bind to) the same table. This feature is useful in applications where you need to reference the same table by different names. Each DBTBLNM structure in the dbcolnm list contains the table ID and for the table and the alternate table name. For a description of the DBTBLNM structure, see page 99. Table names must be unique within a schema. 188 Security To add a synonym to a table, the schema where the table resides must be current, and you must have one of the following permissions. DBA authority SCHEMA authority To make a schema current, use the ufchath function. To grant DBA and SCHEMA authority, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the uaddtnm function to add a table synonym. /* Initialize DBTBLNM structure */ dbtblnm.tid = $parts$; dbtblnm.tblnm = “parts_data”; if ( uaddtnm(txid, 1, &dbtblnm, statusl, &status) != USUCCESS) printf(“Error naming table: status = %ld\n”, status); if ( ubndtbl(txid, 1, &(dbtblnm.tblnm), USLCK, &tid, &status) != USUCCESS) printf( “Error using new table name: status=%ld\n”,status); See Also udrptnm For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 189 uaddusr Add a user to a database and set a user’s default schema #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uaddusr( txid, nuser, dbuser, statusl, status ) UTXID txid; int nuser; DBUSER *dbuser; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nuser Number of users in the dbuser list dbuser List of DBUSER structures, each describing a separate user to add to the database statusl Returned list of status values, one for each dbuser entry status Returned status of the function Adds the users and user schema defaults described by the dbuser list and adds them to the database. Each DBUSER structure in the dbuser list contains information about a user’s name. For a description of the DBUSER structure, see page 100. Security To add a user to a database (grant a user access permission), you must have DBA authority. When a user is added to a database, the user is granted access to the database and schema access permission to the PUBLIC schema. 190 To grant DBA and SCHEMA authority, use the uaddprv function. To grant schema access permission to a schema, use the udrpprm function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example calls the uaddusr function to add two users to the database. /* define required parameters */ UTXID txid; DBUSER * dbuser; USTATUS statusl[2]; USTATUS status; int nuser; /* Initialize DBUSER structure */ dbuser[0].usernm = ”rick”; /* owner (login) name #0 */ dbuser[0].aid = UNULLAID; /* Assume default schema */ dbuser[1].usernm = ”bill”; /* owner (login) name #1 */ dbuser[1].aid = UNULLAID; /* Assume default schema */ /* Add the users */ nuser = 2; if (uaddusr(txid, nuser, dbuser, statusl, &status) != USUCCESS) { printf(”Unable to create new users: %ld\n”, status); for ( i = 0; i < nuser; ++i ) printf(”Entry %d users %s status: %ld\n”, i, dbuser[i].usernm, statusl[i]); exit(1); } See Also uaddprm, udrpprm, uaddath, udrpath, uadddb, udrpusr For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 191 uaddvol Add a volume #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uaddvol( txid, nvol, dbvolm, statusl, status ) UTXID txid; int nvol; DBVOLM *dbvolm; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nvol Number of volumes in the dbvolm list dbvolm List of DBVOLM structures, each describing a separate volume to add to the database statusl Returned list of status values, one for each dbvolm entry status Returned status of the function Adds a new, empty volume to an existing database. Each DBVOLM structure in the dbvolm list contains information about a volume. A volume ID is returned into the DBVOLM vid member, and can be used for future reference to the volume. For a description of the DBVOLM structure, see “The RHLI Structures.” Security To add a volume to a database you must have DBA authority. To grant DBA authority, use the uaddprv function. 192 Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the uaddvol function to add a volume to a database. int nvol; USTATUS status; DBVOLM dbvolm; USTATUS statusl[1]; /* Initialize DBVOLM structure */ nvol = 1; dbvolm.volfile = ”/dev/rdsk3”; dbvolm.volopts = DB_DEVICE | DB_BIG; dbvolm.voloff = 0; dbvolm.vollen = 20480; /* actual size is 41943040 bytes */ dbvolm.volpgsz = 2048; /* 2k is only supported page size */ /* attempt to add volume */ if ( uaddvol(txid, nvol, &dbvolm, statusl, &status) != USUCCESS) fprintf(stdout, ”Volume addition failed. status: %ld\n”,status); See Also umodvol, uaddtbl For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about managing database volumes, see “Managing the Database Files” in Unify DataServer: Managing a Database. 193 ualcscn Allocate a scan information structure #include <include/rhli.h> #include <include/rhlierr.h> Syntax int ualcscn( txid, nsel, nsrt, ncol, scaninf, status ) UTXID txid; int nsel; int nsrt; int ncol; char **scaninf; USTATUS *status; Arguments Description txid Transaction ID. nsel Optional; number of OR-groups in selection criteria; this value represents the number of calls to the uinsor function. nsrt Optional; number of sort criteria. ncol Optional; number of columns in the projection for the scan. scaninf Returned pointer to a scan information structure. status Returned status of the function. Allocates a scan information structure and returns a pointer for future scan initialization calls. Performance If the scan is well defined when the ualcscn function is called, specifying the number of OR-groups (nsel), sort criteria (nsrt), and projection columns (ncol) can greatly speed up the scan specification process for complex selection criteria. If you do not 194 know the values for these parameters, specify 0. In this case, portions of the scan structure are allocated by the uinsor, uinsand, uinssrt, and uinsprj functions. If the numbers specified for nsel, nsrt, or ncol are larger than those actually used by the scan, some user memory will be wasted. If the numbers are smaller, calls to the functions will cause portions of the scan structure to be reallocated. Choosing these numbers is a tradeoff between user memory and the potential CPU overhead of reallocating the structures. When no longer needed, you must free the scan information structure’s memory allocation by using the ufrescn function. Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See appendix A. See Also uinsor, uinsand, uinssrt, uinsprj, ufrescn, ubegscn, uendscn, ufchscn For information about scanning, see page 64. 195 ualctbl Allocate a UNIPCOLL structure #include <include/rhli.h> #include <include/rhlierr.h> Syntax int ualctbl( txid, ntbl, tidl, unipcol, statusl, status ) UTXID txid; int ntbl; UTID *tidl; UNIPCOLL unipcol[]; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID ntbl Number of table IDs in the tidl list tidl An array of table IDs to allocate column data buffers for unipcol An array of UNIPCOLL structures that contain column data buffers for each corresponding table in the tidl list statusl Returned list of status values, one for each tidl entry status Returned status of the function Dynamically allocates the members of a UNIPCOLL structure for columns of each table listed in the tidl list. When you no longer need the UNIPCOLL structure, you must free each column buffer and the dbval and cidl structures. When freeing a dbvall member, be sure to free each data type. Each UNIPCOLL structure in the tidl list contains information about columns for a table. 196 If the ncol member of a UNIPCOLL structure is set to zero, you do not need to free any structures. For a description of the UNIPCOLL structure, see page 131. Security To allocate a UNIPCOLL structure, you must have access to all tables listed in tidl. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example To allocate a UNIPCOLL structure: #define NTBL 3 /* define required parameters */ UTXID tx; /* Transaction ID */ int ntbl; /* # tables listed in tidl */ UTID tidl[NTBL]; /* list of Table IDs */ UNIPCOLL unipcol[NTBL]; /* list of column list info */ USTATUS statusl[NTBL]; /* table status list */ USTATUS status; /* function status buffer */ /* automatically generate a column list */ ntbl = NTBL; tidl[0] = $PUBLIC.manf$; tidl[1] = $PUBLIC.item$; tidl[2] = $PUBLIC.taxes$; if (ualctbl(tx, ntbl, tidl, unipcol, statusl, &status) != USUCCESS) { printf(“Error allocating UNIPCOLL structure: %ld”, status); uexit(1); } 197 To deallocate the allocated UNIPCOLL structure: /* use existing unipcol values */ int tblidx; UNIPCOLL * unipcolp; int colidx; char * charptr; for ( tblidx = 0 ; tblidx < ntbl ; ++tblidx ) { unipcolp = &unipcol[tblidx]; for ( colidx = 0; colidx < unipcolp–>ncol ; ++colidx ) free(unipcolp–>dbvall[colidx].unip.strp); free((char *)unipcolp–>dbvall); free((char *)unipcolp–>cidl); } See Also For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 198 uallath Get a list of authorization IDs (schemas) Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uallath( txid, lcktype, nauth, aidl, status ) UTXID txid; int lcktype; int *nauth; UAID **aidl; USTATUS *status; Arguments Description txid Transaction ID lcktype Lock type to place on the retrieved authorization IDs. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock nauth Number of authorization IDs returned into aidl aidl Returned list of authorization IDs status Returned status of the function Locks and retrieves the authorization IDs for all schemas in the current database. Locking an authorization ID preserves the integrity of the definition for the associated schema while the lock is held. Authorization IDs returned (into aidl) are only valid within the scope of a single transaction. Unify DataServer allocates storage for the returned list of authorization IDs (aidl) by calling the malloc function. After you have retrieved the authorization IDs, you must recover this storage space by calling the free function. 199 If the number of authorization IDs (nauth) returned is zero (0), the malloc function was not performed by Unify DataServer and you do not need to call the free function. Security Authorization IDs are returned only for schemas you have access to. You always have access to the current schema. Additionally, you can access other schemas if you have been granted schema access permission to the schema or if a schema you can access has schema privileges to another schema (or a table or column in the schema). If you have DBA authority, you can access any schema in the database. To make a schema current, use the ufchath function. To grant DBA and SCHEMA authority, use the uaddprv function. To grant schema access permission, use the uaddprm function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example calls the uallath function to retrieve a list of authorization IDs. /* define required parameters */ UAID * aidl; int nauth; /* Retrieve authorization IDs into aidl*/ if ( uallath(txid, USLCK, &nauth, &aidl, &status) != USUCCESS ) printf(”Unable to get all Auth. IDs: status = %ld\n”,status); /* free space used for aidl */ if ( nauth > 0 ) free((UAID *)aidl); See Also uaddath, ubndath, udrpath, ufchath, uinfath, umodath malloc(3) and free(3) in your UNIX system manual For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 200 uallbt Get a list of B-tree IDs Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uallbt( txid, tid, lcktype, nbt, bidl, status ) UTXID txid; UTID tid; int lcktype; int *nbt; UBTID **bidl; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of table containing the B-trees or UNULLTID lcktype Lock type to place on the retrieved B-tree IDs. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock nbt Number of B-tree IDs returned into bidl bidl Returned list of B-tree IDs status Returned status of the function Locks and retrieves a list of all B-tree IDs for the specified table. If tid is set to UNULLTID, B-tree IDs are returned for all tables in the database you have access to. Locking a B-tree ID preserves the integrity of the definition for the B-tree while the lock is held. B-tree IDs returned are valid within the scope of a single transaction only. 201 Unify DataServer allocates storage for the returned list of B-tree IDs (bidl) by calling the malloc function. After you have retrieved the B-tree IDs, you must recover this storage space by calling the free function. If the number of B-tree IDs (nbt) returned is zero (0), the malloc function was not performed by Unify DataServer and you do not need to call the free function. Security B-tree IDs are returned only for B-trees in tables that you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant DBA and SCHEMA authority, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the uallbt function to retrieve a list of B-tree IDs. UBTID * bidl; statusl = (USTATUS *)calloc(1, sizeof(USTATUS)); if ( uallbt(txid, UNULLTID, UNULLLCK, &nbt, &bidl, &status) != USUCCESS) printf(”Unable to get all B-tree ID info: status = %ld\n”,status); if ( nbt > 0 ) free((UBTID *)bidl); See Also ubndbt, ufchath, uinfbt, umodbt, uaddbt For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about using B-trees, see “Choosing an Access Method” and “Managing the Database Files” in Unify DataServer: Managing a Database. 202 uallcgp Get a list of column group IDs Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uallcgp( txid, tid, lcktype, ngrp, cidl, status ) UTXID txid; UTID tid; int lcktype; int *ngrp; UCID *cidl[]; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of the table containing the column groups or UNULLTID lcktype Lock type to place on the retrieved column group IDs. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock ngrp Number of column group IDs returned into cidl cidl Returned list of column group IDs status Returned status of the function Locks and retrieves a list of all column group IDs for the specified table. If tid is set to UNULLTID, column group IDs are returned for all tables in the database that you have access to. 203 Locking a column group ID preserves the integrity of the definition for the column group while the lock is held. Column group IDs returned are valid within the scope of a single transaction only. Unify DataServer allocates storage for the returned list of column group IDs (cidl) by calling the malloc function. After you have retrieved the column group IDs, you must recover this storage space by calling the free function. If the number of column group IDs (ngrp) returned is zero (0), the malloc function was not performed by Unify DataServer and you do not need to call the free function. Security Column group IDs are returned only for column groups in tables that you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant DBA and SCHEMA authority, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the uallcgp function to retrieve a list of column group IDs. UTXID tx; UTID tid; /* Transaction ID */ /* Table ID the col group belongs to /* (or NULLTID) */ int lcktype;/* requested lock type */ int ngrp; /* # col groups returned in cidl */ UCID * cidl; /* list of all Column Group IDs USTATUS status; /* returned status */ int indx; /* temp index variable */ /* attempt to retrieve all column groups for table ’customer’ */ 204 */ */ tid = $customer.$ if (uallcgp(txid, tid, lcktype, &ngrp, &cidl, &status) != USUCCESS) { printf(”Unable to get all column Group ID info: status = %ld\n”,status); uexit(99); } printf(”%d column groups found\n”, ngrp); for ( indx = 0; indx < ngrp; ++indx ) printf(”column Group ID: %ld\n”, cidl[indx]); if ( ngrp > 0 ) free((UCID *)cidl); See Also uaddcgp, udrpcgp, ufchath, uinfcgp, unumcgp malloc(3) and free(3) in your UNIX system manual. For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 205 uallcol Get a list of column IDs #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uallcol( txid, tid, lcktype, ncol, cidl, status ) UTXID txid; UTID tid; int lcktype; int *ncol; UCID **cidl; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of the table containing the columns or UNULLTID lcktype Lock type to place on the retrieved column IDs. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock ncol Number of column IDs returned into cidl cidl Returned list of column IDs status Returned status of the function Locks and retrieves a list of all column IDs for the specified table. If tid is set to UNULLTID, column IDs are returned for all tables in the database that you have access to. Locking a column ID preserves the integrity of the definition for the column while the lock is held. 206 Column IDs returned (into cidl) are valid only within the scope of a single transaction. Unify DataServer allocates storage for the returned list of column IDs (cidl) by calling the malloc function. After you have retrieved the column IDs, you must recover this storage space by calling the free function. If the number of column IDs (ncol) returned is zero (0), the malloc function was not performed by Unify DataServer and you do not need to call the free function. Security Column IDs are returned only for columns in tables that you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant DBA and SCHEMA authority, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the uallcol function to retrieve a list of column IDs. UCID * cidl; if ( uallcol(txid, tid, lcktype, &ncol, &cidl, &status) != USUCCESS ) printf(”Unable to get all Column ID info: status = %ld\n”,status); if ( ncol > 0 ) free((UCID *)cidl); See Also ualltbl, ubndcol, uaddcnm, udrpcnm, uaddtnm, udrptnm, ufchath, uinfcol, uinftbl malloc(3) and free(3) in your UNIX system manual For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 207 uallhsh Get a list of hash table IDs #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uallhsh( txid, tid, lcktype, nhsh, hidl, status ) UTXID txid; UTID tid; int lcktype; int *nhsh; UHID **hidl; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of the table containing the hash tables or UNULLTID lcktype Lock type to place on the retrieved hash table IDs. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock nhsh Number of hash table IDs returned into hidl hidl Returned list of hash table IDs status Returned status of the function Locks and retrieves a list of all hash table IDs for the specified table. If tid is set to UNULLTID, hash table IDs are returned for all tables in the database you have access to. Locking a hash table ID preserves the integrity of the definition for the hash table while the lock is held. Hash table IDs returned (into hidl) are only valid within the scope of a single transaction. 208 Unify DataServer allocates storage for the returned list of hash table IDs (hidl) by calling the malloc function. After you have retrieved the hash table IDs, you must recover this storage space calling the free function. If the number of hash table IDs (nhsh) returned is zero (0), the malloc function was not performed by Unify DataServer and you do not need to call the free function. Security Hash table IDs are returned only for hash indexes in tables that you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This examples uses the uallhsh function to retrieve a list of hash table IDs. UHID * hidl; if (uallhsh(txid, tid, lcktype, &nhsh, &hidl, &status) != USUCCESS) printf(”Unable to get all Hash Table info; status = %ld\n”, status); if ( nhsh > 0 ) free((UHID *)hidl); See Also ubndhsh, uaddhsh, udrphsh, umodhsh, ufchath, uinfhsh malloc(3) and free(3) in your UNIX system manual For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about using hash tables, see “Choosing an Access Method” and “Tuning the Access Methods” in Unify DataServer: Managing a Database. 209 ualllnk Get a list of link IDs #include <include/rhli.h> #include <include/rhlierr.h> Syntax int ualllnk( txid, tid, lcktype, nlid, lidl, status ) UTXID txid; UTID tid; int lcktype; int *nlid; ULID **lidl; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of the table containing the links or UNULLTID lcktype Lock type to place on the retrieved link IDs. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock nlid Number of link IDs returned into lidl lidl Returned list of link IDs status Returned status of the function Locks and retrieves a list of link IDs for the specified table. If tid is set to UNULLTID, link IDs are returned for all tables in the database you have access to. Locking a link ID preserves the integrity of the definition for the link while the lock is held. Link IDs returned (into lidl) are valid only in the scope of a single transaction. 210 Unify DataServer allocates storage for the returned list of link IDs (lidl) by calling the malloc function. After you have retrieved the link IDs, you must recover this storage space by calling the free function. If the number of link IDs (nlid) returned is zero (0), a malloc was not performed by Unify DataServer and you do not need to call the free function. Security Link IDs are returned only for links in tables that you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the ualllnk function to retrieve a list of link index IDs. ULID * lidl; if (ualllnk(txid, tid, lcktype, &nlid, &lidl, &status) != USUCCESS) printf(“Can’t get all Link ID info: status = %ld\n”,status); if ( nlid > 0 ) free((ULID *)lidl); See Also ubndlnk, uaddlnk, udrplnk, umodlnk, ufchath, uinflnk malloc(3) and free(3) in your UNIX system manual For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about using links, see “Choosing an Access Method” in Unify DataServer: Managing a Database. 211 uallpid Get a list of process IDs #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uallpid( dbid, npid, pidl, status ) UDBID dbid; int *npid; UPROCID **pidl; USTATUS *status; Arguments Description dbid Database ID of the database associated with the process IDs npid Number of process IDs returned into pidl pidl Returned list of process IDs status Returned status of the function Retrieves the set of all process IDs of processes registered with the database. A process is registered with the database when the database is opened, and deregistered when the database is closed. Unify DataServer allocates storage for the returned list of process IDs (pidl) by calling the malloc function. After you have retrieved the process IDs, you must recover this storage space by calling the free function. If the number of process IDs (npid) returned is zero (0), the malloc function was not performed by Unify DataServer and you do not need to call the free function. Security If you have DBA authority, you can retrieve all active process IDs. If you don’t have DBA authority, you can retrieve your process ID only. 212 To grant DBA authority, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example calls the uallpid function to retrieve the list of process IDs. /* define require parameters */ UDBID dbid; int npid; UPROCID * pidl; USTATUS status; /* Open the database */ ... /* Retrieve process IDs */ if (uallpid(dbid, &npid, &pidl, &status) != USUCCESS) fprintf(stderr,”Unable to get Process IDs: status = %ld\n”, status); ... / * free space allocated for pidl */ if ( npid > 0 ) free((char *) pidl); See Also ualltx, uinftx malloc(3) and free(3) in your UNIX system manual For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 213 ualltbl Get a list of table IDs #include <include/rhli.h> #include <include/rhlierr.h> Syntax int ualltbl( txid, aid, lcktype, ntbl, tidl, status ) UTXID txid; UAID aid; int lcktype; int *ntbl; UTID **tidl; USTATUS *status; Arguments txid Transaction ID aid Authorization ID of the schema that contains the tables or UNULLAID lcktype Description Lock type to place on the retrieved table IDs. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock ntbl Number of table IDs returned into tidl tidl Returned list of table IDs status Returned status of the function Locks and retrieves a list of table IDs. If aid is set to UNULLAID, table IDs are returned for all schemas in the database you have access to. Locking a table ID preserves the integrity of the definition for the table while the lock is held. 214 Table IDs returned (into tidl) are valid only in the scope of a single transaction. Unify DataServer allocates storage for the returned list of table IDs (tidl) by calling the malloc function. After you have retrieved the table IDs, you must recover this storage space by calling the free function. If the number of table IDs (ntbl) returned is zero (0), the malloc function was not called by Unify DataServer and you do not need to call the free function. Security Table IDs are returned only for tables you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the ualltbl function to retrieve a list of table IDs. UTID * tidl; aid = UNULLAID; if (ualltbl(txid, aid, lcktype, &ntbl, &tidl, &status) != USUCCESS ) fprintf(stderr,”Unable to get all Table ID info: status = %ld\n”,status); if ( ntbl > 0 ) free((UTID *)tidl); See Also uallcol, ubndtbl, uinftbl, uinfcol, uaddtnm, udrptnm, ufchath, ufchtnm malloc(3) and free(3) in your UNIX system manual. For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 215 ualltx Get a list of transaction IDs #include <include/rhli.h> #include <include/rhlierr.h> Syntax int ualltx( dbid, pid, ntx, txidl, status ) UDBID dbid; UPROCID pid; int *ntx; UTXID **txidl; USTATUS *status; Arguments Description dbid Database ID of database associated with the transaction IDs pid UNIX process ID for the current application (obtained from the UNIX system call getpid) or UNULLPID ntx Number of transaction IDs returned into txidl txidl Returned list of transaction IDs status Returned status of the function Retrieves the set of active transaction IDs for the specified process ID. To specify active transactions for the current application, set pid to UNULLPID; otherwise, set pid to a specific process ID. You can use the getpid UNIX function to retrieve the process IDs. Unify DataServer allocates storage for the returned list of transaction IDs (txidl) by calling the malloc function. After you have retrieved the transaction IDs, you must recover this storage space by calling the free function. If the number of transaction IDs (ntx) returned is zero (0), the malloc function was not called by Unify DataServer and you do not need to call the free function. 216 Security If you have DBA authority you can set the process ID (pid) to any valid transaction process ID for any application currently accessing the database. If you don’t have DBA authority, transaction IDs can be returned only for transactions that your application initiated. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also uinftx, ubegtx malloc(3), free(3), and getpid(2) in your UNIX system manual For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 217 uallusr Get a list of user IDs #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uallusr( txid, lcktype, nuser, uoidl, status ) UTXID txid; int lcktype; int *nuser; UOID **uoidl; USTATUS *status; Arguments Description txid Transaction ID lcktype Lock type to place on the retrieved user IDs. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock nuser Number of user IDs returned into uoidl uoidl Returned list of user IDs status Returned status of the function Locks and retrieves user IDs for all users who have access to the current database. Locking a user ID preserves the integrity of the definition for the user while the lock is held. User IDs returned (into uoidl) are valid only in the scope of a single transaction. Unify DataServer allocates storage for the returned list of user IDs (uoidl) by calling the malloc function. After you have retrieved the user IDs, you must recover this storage space by calling the free function. 218 If the number of user IDs (nuser) returned is zero (0), the malloc function was not called by Unify DataServer and you do not need to call the free function. Security User IDs are returned as follows: If you have DBA authority, all user IDs for the database are returned. If you don’t have DBA authority, only your user ID is returned. To grant DBA and SCHEMA authority, use the uaddprv function. To grant schema access permission to a schema, use the udrpprm function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example calls the uallusr function to retrieve the list of database users. / * define required parameters */ UOID * uoidl; UTXID txid; int * nuser; USTATUS status; /* Retrieve user list (uoidl) */ if (uallusr(txid, USLCK, &nuser, &uoidl, &status) != USUCCESS) fprintf(stderr,”Unable to get all User IDs: status = %ld\n”,status); /* free space allocated by uoidl */ if ( nuser > 0 ) free((UOID *)uoidl); See Also uaddusr, ubndusr, udrpusr, uinfusr malloc(3) and free(3) in your UNIX system manual For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 219 uallvol Get a list of volume IDs #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uallvol( txid, lcktype, nvol, vidl, status ) UTXID txid; int lcktype; int *nvol; UVID **vidl; USTATUS *status; Arguments Description txid Transaction ID lcktype Lock type to place on the retrieved volume IDs. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock nvol Number of volume IDs returned into vidl vidl Returned list of volume IDs status Returned status of the function Locks and retrieves the volume IDs for all volumes registered in the current database. Locking a volume ID preserves the integrity of the definition for the volume while the lock is held. Volume IDs returned (into vidl) are valid only in the scope of a single transaction. Unify DataServer allocates storage for the returned list of volume IDs (vidl) by calling the malloc function. After you have retrieved the volume IDs, you must recover this storage space by calling the free function. 220 If the number of volume IDs (nvol) returned is zero (0), the malloc function was not called by Unify DataServer and you do not need to call the free function. Security To get a list of all volume IDs defined for the current database, you must have DBA authority. To grant DBA authority, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the uallvol function to retrieve a list of volume IDs. UVID * vidl; if (uallvol(txid, USLCK, &nvol, &vidl, &status) != USUCCESS) fprintf(stderr,”Unable to get all Volume IDs: status = %ld\n”,status); if ( nvol > 0 ) free((UVID *)vidl); See Also uaddvol, ubndvol, udrpvol, ufchath, uinfvol malloc(3) and free(3) in your UNIX system manual. For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about managing database volumes, see “Managing the Database Files” in Unify DataServer: Managing a Database. 221 ubegord Begin an ordered access #include <include/rhli.h> #include <include/rhlierr.h> Syntax int ubegord (txid, tid, ncol, uoacols, oaccid, statusl, status) UTXID txid; UTID tid; int ncol; UOACOLS *uoacols; UOACCID *oaccid; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of the target table for the ordered access ncol Number of columns in the uoacols list uoacols List of UOACOLS structures, each describing a column comprising the order oaccid Pointer to returned ordered access ID statusl Returned status value for each column listed in the oaccid field of each UOACOLS structure status Returned status of the function Sets up the specified table for ordered access. The table’s rows are set up for access in ascending or descending order on a set of columns as specified by the column order list (uoacols). 222 Currently, the specified order must be supported by an underlying B-tree or direct key. For B-trees, its key columns must match or exceed that specified by the column order list. That is, the column order list can specify a subset of an actual B-tree index key as long as the leading columns match. The B-tree index is not opened at this time. It is opened the first time it is accessed by a ufchord, uposord, or uskrord function call. The order specified by UOACOLS can be in either the same or opposite order of that of the underlying access method. If it is in the opposite order then the B-tree or direct key is traversed in reverse order. The returned ordered access ID is specified in future references to the ordered access. Rows are retrieved from the ordered access with the ufchord function. The current position is set to just before the first row of the order. The current position within an ordered access on a table can be changed with the ufchord, uposord, or uskrord function. When the ordered access is no longer needed, the uendord function closes the ordered access and frees up the resources (such as memory) held by the ordered access. If the transaction ends before the uendord function is called, the ordered access is automatically closed, unless the USCAN option is specified to the ucbgtx function. Security To begin an ordered access, you must be able to access the table. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, see the ufchath function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 223 Example See Appendix A. See Also uendord, ufchord, uposord For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 224 ubegscn Begin a scan Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ubegscn( txid, lcktype, scaninf, scanid, status ) UTXID txid; int lcktype; char *scaninf; int *scanid; USTATUS *status; Arguments Description txid Transaction ID lcktype Lock type to place on the scanned rows. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock scaninf Pointer to a scan information structure scanid Pointer to returned scan ID status Returned status of the function Given a fully specified scan, the ubegscn function optimizes the scan information (contained in scaninf) and returns a scan ID (into scanid). The returned scan ID can be used for future references to the scan. Activities that should occur before beginning a scan include: Allocation of a scan information structure (ualcscn) Establishing the target table (uinstbl) Specifying the scan selection criteria (uinsor, uinsand, uinsprj) Specifying the scan sort criteria (uinssrt) 225 Rows are retrieved from the scan with the ufchscn function. When the scan is complete, the uendscn function is used to close the scan and free up the resources (such as memory) held by the scan. If the transaction ends before the uendscn function is called, the scan is closed. Security To begin a scan, one of the following conditions must be true: You have DBA authority. The target table is in the current schema. You have SELECT schema privilege on the target table. You have SELECT schema privilege on all columns projected from the table. To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also ualcscn, uendscn, ufchscn, ufrescn, uinsand, uinsor, uinsprj, uinssrt, uinstbl For information about scanning, see page 64. For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 226 ubegtx Begin a transaction Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ubegtx( dbid, prtx, txid, txopt, status ) UDBID dbid; UTXID prtx; UTXID *txid; UTXOPT *txopt; USTATUS *status; Arguments Description dbid Database ID (issued by the uopndb function) prtx Must be set to UROOTTXID txid Returned unique transaction ID issued to the current transaction txopt Pointer to UTXOPT structure that describes the new transaction’s options status Returned status of the function Begins a new transaction and returns a unique transaction ID. The transaction ID is used in subsequent function calls. To begin a transaction with this function, a UTXOPT structure (pointed to by txopt) must first be initialized. The UTXOPT structure describes the new transaction. For a complete description of the UTXOPT structure, see page 152. Data definition (DDL) and data manipulation (DML) operations cannot occur in the same transaction. Use separate transactions for DDL and DML operations. To commit one transaction and begin another, use the ucbgtx function. 227 Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also ucbgtx, ucmttx, ualltx, ubegtx, uinftx For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 228 ubndath Bind a schema name to an authorization ID Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ubndath( txid, nauth, authnml, lcktype, aidl, statusl, status ) UTXID txid; int nauth; char * authnml[]; int lcktype; UAID aidl[]; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nauth Number of schema names in the authnml list authnml List of schema names lcktype Lock type to place on the retrieved authorization IDs. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock aidl Returned list of authorization IDs statusl Returned list of status values, one for each aidl entry status Returned status of the function Locks and retrieves an authorization ID for each schema name specified in authnml. Locking an authorization ID preserves the integrity of the name binding while the lock is held. Also see “Transaction Types” on page 62 of this manual. 229 If you use a Type 0 transaction (which allows no locking) when performing name-binding, no definition locks are placed on the schemas you are binding to. Therefore, there is no guarantee that the schemas you bound to will be the same or even exist by the time you commit the transaction. Security To bind a schema name to an authorization ID you must have one of the following permissions: DBA authority (any authorization ID can be returned) Schema access to the schema Access to a schema with a privilege granted on one of the following: – the schema itself (DELETE, INSERT, SELECT, or UPDATE) – a table within the schema (DELETE, INSERT, SELECT, or UPDATE) – a column within the schema (INSERT, SELECT, or UPDATE) To grant DBA authority and schema privileges, use the uaddprv function. To grant access permission to a schema, use the uaddprm function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also uaddath, uallath, udrpath, uinfath, umodath For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. Also see “Transaction Types” on page 62 of this manual. 230 ubndbt Bind a B-tree name to a B-tree ID Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ubndbt( txid, tid, nbt, btnml, lcktype, bidl, statusl, status ) UTXID txid; UTID tid; int nbt; char *btnml[]; int lcktype; UBTID *bidl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of table the B-tree is defined for or UNULLTID nbt The number of B-tree names in the btnml list btnml List of B-tree names lcktype Lock type to place on the retrieved B-tree IDs. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock bidl Returned list of nbt B-tree IDs statusl Returned list of status values, one for each btnml/bidl entry status Returned status of the function Locks and retrieves a B-tree ID for each B-tree name specified in btnml. Locking a B-tree ID preserves the integrity of the name binding while the lock is held. 231 If the table value (tid) is specified as UNULLTID, all tables in the current schema is assumed. If you use a Type 0 transaction (which allows no locking) when performing name-binding, no definition locks are placed on the B-trees you are binding to. Therefore, there is no guarantee that the B-trees you bind to will be the same or even exist by the time you commit the transaction. Security To bind a B-tree name to a B-tree ID one of the following must be true: You have DBA authority (any B-tree ID can be returned). The B-tree is in the current schema. The current schema has been granted one of the following schema privileges: – DELETE, INSERT, SELECT or UPDATE privilege for the schema that contains the B-tree. – DELETE, INSERT, SELECT or UPDATE privilege for the table that contains the B-tree. – INSERT, SELECT or UPDATE privilege for all the columns that comprise the B-tree definition. To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. Example See Appendix A. See Also ualllnk, ubndlnk, uaddlnk, udrplnk, umodlnk, uinflnk For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. Also see “Transaction Types” on page 62 of this manual. 232 ubndcol Bind a column name to a column ID Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ubndcol( txid, tid, ncol, colnml, lcktype, cidl, statusl, status ) UTXID txid; UTID tid; int ncol; char *colnml[]; int lcktype; UCID *cidl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID tid The table ID of table containing the columns, or, UNULLTID ncol The number of column names in the colnml list colnml List of column names lcktype Lock type to place on the retrieved column IDs. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock cidl Returned list of ncol column IDs statusl Returned list of status values, one for each colnml/cidl entry status Returned status of the function Locks and retrieves a column ID for each column name specified in colnml. Locking a column ID preserves the integrity of the name binding while the lock is held. 233 If the table value (tid) is specified as UNULLTID, all tables in the current schema is assumed. If you use a Type 0 transaction (which allows no locking) when performing name-binding, no definition locks are placed on the columns you are binding to. Therefore, there is no guarantee that the columns you bound to will be the same or even exist by the time you commit the transaction. Security To bind a column name to a column ID one of the following must be true: You have DBA authority (any column ID can be returned). The column is in the current schema. The current schema has been granted one of the following schema privileges: – DELETE, INSERT, SELECT or UPDATE privilege for the schema that contains the column. – DELETE, INSERT, SELECT or UPDATE privilege for the table that contains the column. – INSERT, SELECT or UPDATE privilege for the column. To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. To grant schema access permission, use the uaddprm function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also uaddcgp, uaddcnm, uallcol, udrpcgp, udrpcnm, ufchcnm, uinfcol, umapcol, unumcol For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. Also see “Transaction Types” on page 62 of this manual. 234 ubnddb Bind a database name to a database ID Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ubnddb( txid, dbname, lcktype, dbid, status ) UTXID txid; char *dbname; int lcktype; UDBID *dbid; USTATUS *status; Arguments Description txid Transaction ID dbname A database name lcktype Lock type to place on the retrieved database ID. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock dbid Returned database ID status Returned status of the function Locks and retrieves a database ID for the database specified in dbname. Locking a database preserves the integrity of the name binding while the lock is held. Binding to a database name does not open or reopen the database. If you use a Type 0 transaction (which allows no locking) when performing name-binding, no definition locks are placed on the database you are binding to. Therefore, there is no guarantee that the database you bound to (and are currently accessing) will have the same design or layout when you commit the transaction. 235 Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also uadddb, uclsdb, uinfdb, uopndb For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. Also see “Transaction Types” on page 62 of this manual. 236 ubndhsh Bind a hash table name to a hash table ID Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ubndhsh( txid, tid, nhsh, hshnml, lcktype, hidl, statusl, status ) UTXID txid; UTID tid; int nhsh; char *hshnml[]; int lcktype; UHID *hidl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of table the hash table is defined for or UNULLTID nhsh Number of hash table names in the hshnml list hshnml List of hash table names lcktype Lock type to place on the retrieved hash table IDs. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock hidl Returned list of nhsh hash table IDs statusl Returned list of status values, one for each hshnml entry status Returned status of the function Locks and retrieves a hash table ID for each hash table name specified in hshnml. Locking a hash table ID preserves the integrity of the name binding while the lock is held. If the table value (tid) is specified as UNULLTID, all tables in the current schema is assumed. 237 If you use a Type 0 transaction (which allows no locking) when performing name-binding, no definition locks are placed on the hash tables you are binding to. Therefore, there is no guarantee that the hash tables you bound to will be the same or even exist by the time you commit the transaction. Security To bind a hash table name to a hash table ID one of the following must be true: You have DBA authority: any hash table ID can be returned. The hash table is in the current schema. The current schema has been granted one of the following schema privileges: – DELETE, INSERT, SELECT or UPDATE privilege for the schema that contains the hash table. – DELETE, INSERT, SELECT or UPDATE privilege for the table that contains the hash table. – INSERT, SELECT or UPDATE privilege for all the columns that comprise the hash table definition. To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also uaddhsh, uaddprm, uallhsh, udrphsh, uaddprv, uinfhsh, umaphsh, umodhsh For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. Also see “Transaction Types” on page 62 of this manual. For information about using hash tables, see “Managing the Database Files,” “Choosing an Access Method,” and “Tuning the Access Methods” in Unify DataServer: Managing a Database. 238 ubndlnk Bind a link name to a link ID Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ubndlnk( txid, tid, nlnk, lnknml, lcktype, lidl, statusl, status ) UTXID txid; UTID tid; int nlnk; char *lnknml[]; int lcktype; ULID *lidl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of table the links are defined for or UNULLTID nlnk The number of link names in the lnknml list lnknml List of link names lcktype Lock type to place on the retrieved link IDs. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock lidl Returned list of nlnk link IDs statusl Returned list of status values, one for each lnknml/lidl entry status Returned status of the function Locks and retrieves a link ID for each link name specified in lnknml. Locking a link ID preserves the integrity of the name binding while the lock is held. If the table value (tid) is specified as UNULLTID, all tables in the current schema is assumed. 239 If you use a Type 0 transaction (which allows no locking) when performing name-binding, no definition locks are placed on the links you are binding to. Therefore, there is no guarantee that the links you bound to will be the same or even exist by the time you commit the transaction. Security To bind a link name to a link ID one of the following must be true: You have DBA authority: any link ID can be returned. Both tables are in the current schema. Access to a schema with a privilege granted on one of the following: – DELETE, INSERT, SELECT or UPDATE privilege for the schemas that contain the linked tables. – DELETE, INSERT, SELECT or UPDATE privilege for the tables that contain columns comprising the link definition. – INSERT, SELECT or UPDATE privilege for all the columns that comprise the link definition. To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also ualllnk, ubndlnk, uaddlnk, udrplnk, umodlnk, uinflnk For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. Also see “Transaction Types” on page 62 of this manual. For information about using links, see “Choosing an Access Method in Unify DataServer: Managing a Database. 240 ubndtbl Bind a table name to a table ID Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ubndtbl( txid, aid, ntbl, tblnml, lcktype, tidl, statusl, status ) UTXID txid; UAID aid; int ntbl; char* tblnml[]; int lcktype; UTID *tidl; USTATUS statusl[]; USTATUS *status; Arguments txid Transaction ID aid The authorization ID of the schema containing the table IDs or UNULLAID Description ntbl Number of table names in the tblnml list tblnml List of table names lcktype Lock type to place on the retrieved table IDs. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock tidl Returned list of ntbl table IDs statusl Returned list of status values, one for each tblnml entry status Returned status of the function Locks and retrieves a table ID for each table name specified in tblnml. The authorization ID (aid) specifies the schema that contains the table. If aid is specified as UNULLAID, the tables contained in the user’s current schema are used. 241 If a table name (specified in tblnml) contains a prefixed schema name, the schema specified by aid is overridden. Locking a table ID preserves the integrity of the name binding while the lock is held. If you use a Type 0 transaction (which allows no locking) when performing name-binding, no definition locks are placed on the tables you are binding to. Therefore, there is no guarantee that the tables you bound to will be the same or even exist by the time you commit the transaction. Security To bind a table name to a table ID one of the following must be true: DBA authority: any table ID can be returned. The table is in the current schema. Access to a schema with a privilege granted on one of the following: – DELETE, INSERT, SELECT, or UPDATE privilege for the schema that contains the table. – DELETE, INSERT, SELECT, or UPDATE privilege for the table itself. – INSERT, SELECT, or UPDATE privilege for a column in the table. To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also ualltbl, ubndcol, uinftbl For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. Also see “Transaction Types” on page 62 of this manual. 242 ubndusr Bind a user name to a user ID Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ubndusr( txid, nuser, usernml, lcktype, uoidl, statusl, status ) UTXID txid; int nuser; char *usernml[]; int lcktype; UOID *uoidl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nuser Number of user names in the usernml list usernml List of user names lcktype Lock type to place on the retrieved user IDs. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock uoidl Returned list of user IDs statusl Returned list of status values, one for each usernml/uoidl entry status Returned status of the function Locks and retrieves a user ID for each user name specified in usernml. Locking a user ID preserves the integrity of the name binding while the lock is held. If you use a Type 0 transaction (which allows no locking) when performing name-binding, no definition locks are placed on the users you are binding to. Therefore, there is no guarantee that the users you bound to will be the same or even exist by the time you commit the transaction. 243 Security To bind a user name to a user ID, one of the following conditions occur: If you have DBA authority, you can specify any user name in usernml. If you don’t have DBA authority, only your user ID can be returned. To grant DBA authority, use the uaddprv function. To grant access to a database, use the uaddusr function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also uaddusr, uallusr, udrpusr, uinfusr For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. Also see “Transaction Types” on page 62 of this manual. 244 ubndvol Bind a volume name to a volume ID Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ubndvol( txid, nvol, volnml, lcktype, vidl, statusl, status ) UTXID txid; int nvol; char* volnml[]; int lcktype; UVID *vidl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nvol Number of volume names in the volnml list volnml List of volume names lcktype Lock type to place on the retrieved volume IDs. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock vidl Returned list of volume IDs statusl Returned list of status values, one for each volnml/vidl entry status Returned status of the function Locks and retrieves a volume ID for each volume name specified in volnml. Locking a volume ID preserves the integrity of the name binding while the lock is held. If you use a Type 0 transaction (which allows no locking) when performing name-binding, no definition locks are placed on the volumes you are binding to. Therefore, there is no guarantee that the volumes you bound to will be the same or even exist by the time you commit the transaction. 245 Security To bind a volume name to a volume ID, you must have DBA authority. To grant DBA authority, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also uaddvol, uallvol, udrpvol, uinfvol For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. Also see “Transaction Types” on page 62 of this manual. 246 ucbgtx Commit a transaction and begin a new transaction Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ucbgtx( prtx, txid, unlcktype, newtx, txopt, status ) UTXID prtx; UTXID txid; int unlcktype; UTXID *newtx; UTXOPT *txopt; USTATUS *status; Arguments prtx Must be set to UROOTTXID txid The transaction ID of the transaction that is being committed. unlcktype An unlock type for the committing transaction’s locks. The unlcktype member options controls whether locks are released or inherited by the new transaction, or scans are closed or inherited by the new transaction. Options are: URELEASE Release all locks UDEMOTE Inherit all locks but demote exclusive locks to shared locks URETAIN Inherit all locks without demoting exclusive locks to shared locks USCAN Preserve scans and ordered accesses When specifying USCAN, you must also specify either UDEMOTE or URETAIN. newtx Returned unique transaction ID issued to the new transaction txopt Pointer to a UTXOPT structure that describes the new transaction’s options 247 status Description Returned status of the function Commits the current active transaction and begins another transaction against the specified database. To start a transaction with this function, the UTXOPT structure (pointed to by txopt) must first be initialized. UTXOPT describes the new transaction. For a complete description of the UTXOPT structure, see page 152. Data definition (DDL) and data manipulation (DML) operations cannot use the same transaction. Use separate transactions for DDL and DML operations. Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also ubegtx, ucmttx, uinftx, uabttx, ualltx For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 248 uclritr Clear a process interrupt Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uclritr( status ) USTATUS *status; Arguments status Description Allows processing to resume after an interrupt has been set by the usetitr function. This function should not be called from a signal handler routine; instead, it should be called from the main program. Returned status of the function All child processes are terminated at a point where the database is in a known consistent state. Process interruption by any other mechanism may require database recovery and is not recommended. Before clearing an interrupt with the uclritr function, be sure that a database is open. If the DB_NOEXIT option was specified when the current database was opened by calling the uopndb function, all RHLI function calls and Embedded SQL/A statements fail (with a status value of UEINTRPT) when an interrupt signal is received by Unify DataServer until the signal is cleared with the uclritr function. If the database is opened with Embedded SQL/A CONNECT or OPEN DATABASE statements, DB_NOEXIT option is implicit. Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 249 See Also usetitr signal(2) in your UNIX system manual 250 uclsdb Close a database Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uclsdb( dbid, clsopts, status ) UDBID dbid; UOPTS clsopts; USTATUS *status; Arguments dbid The database ID of the database you want to close clsopts The database close options. Valid options are: DB_NORMAL Normal operation DB_ABTTX status Description Abort any active transactions Returned status of the function Closes a database opened by the uopndb function and deallocates all the dynamically allocated memory obtained at the time of the open and during the database access. If clsopts is specified as DB_NORMAL, this function fails if any transactions are still active. However, if the DB_ABTTX option is specified: and transaction logging is on (the LOGTX configuration parameter is set to TRUE), all active transactions are aborted before the database is closed. However, if any transactions were started (by calling the ubegtx or ucbgtx function) with the UCOMMIT option specified (see the UTXOPT structure waitflg member), then only those transactions are committed. and transaction logging is off (the LOGTX configuration variable is set to FALSE), all active transactions are committed before the database is closed. 251 Before opening another database or ending an application program, any previously opened database must be closed with a call to the uclsdb function. Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also uopndb For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. For information about transaction logging, see “Preventing Data Loss” in Unify DataServer: Managing a Database. 252 uclsjrn Close transaction journal file. Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uclsjrn (jrnfd, status) int *jrnfd; USTATUS *status; Arguments jrnfd Pointer to a returned file descriptor for the opened journal file status Returned status of the function Description Closes a transaction journal file. All resources associated with processing a journal file are released. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Security No privileges required. Example See Appendix A. See Also ufchjrn, uopnjrn 253 For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. For information about transaction logging and using a journal, see “Preventing Data Loss” in Unify DataServer: Managing a Database. 254 ucmttx Commit a transaction Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ucmttx( txid, status ) UTXID txid; USTATUS *status; Arguments Description txid Transaction ID status Returned status of the function Commits an active transaction. When a transaction is committed, any changes made during the transaction that affected the database are permanently written to the database. Committing a transaction with this function releases all locks acquired during the transaction and closes any currently active scans against the transaction. To commit one transaction and begin another, use the ucbgtx function. Warning Once a transaction is committed, you cannot rollback any changes that were made to the database during the transaction. Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 255 Example See Appendix A. See Also uabttx, ubegtx, ucbgtx, uinftx, ualltx For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 256 ucrypwd, Encrypt a password Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ucrypwd( pwd, crywdp, status ) char * pwd; char **crywdp; USTATUS * status; Arguments Description pwd Password to encrypt. crywdp Returned pointer to the encrypted password. The ucrypwd function accepts a password and encrypts it in a format that can be used for the database user identity, as in, for example, the DBUSER configuration variable or the dbname argument to the uopndb function. If the function executes successfully, the memory used for storing the encrypted password has been allocated using malloc. You must recover the storage space using free. If the function does not execute successfully, the memory used for the encrypted password has not been allocated. In this case, do not call free. Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 257 Example /* encrypt password */ if (ucrypwd(inpass, &outpass, &status) != USUCCESS ) { (void) printf(”ucrypwd error;”); if ((errmsg = ufchmsg(status, &msgstat)) == NULL) printf(”error %ld fetching message \n”, (long) msgstat); else printf(”error is ’%s’\n”, errmsg); exit(1); } /* endif */ /* put encrypted password to standard output */ (void) printf(”encrypted password is ’%s’\n”, outpass); free(outpass); outpass = NULL; See Also “Accessing a Remote Database” and “Using the uopndb dbname Argument” in Unify/Net Guide. 258 udelrow Delete a row Syntax #include <include/rhli.h> #include <include/rhlierr.h> int udelrow( txid, tid, rid, status ) UTXID txid; UTID tid; URID rid; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of the table that contains the row rid Row ID of the row to delete status Returned status of the function Deletes the row requested (rid) from the table specified (tid). A row deletion can fail for any of the following reasons: The row is referenced as a parent in a link index The lock on the row could not be obtained The transaction log is full Before you can delete a parent row of a link index, you must drop or change the value of any related rows in the child table. For more information about links, see the uaddlnk function. Security To delete a row from a table, one of the following conditions must be true: You have DBA authority. The table is in the current schema. You have DELETE schema privilege on the table that contains the row. 259 To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also uinsrow, uupdrow For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about using links, see “Choosing an Access Method” in Unify DataServer: Managing a Database. 260 udrpath Drop a schema Syntax #include <include/rhli.h> #include <include/rhlierr.h> int udrpath( txid, nauth, authl, statusl, status ) UTXID txid; int nauth; UAID *authl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nauth Number of authorization IDs (schemas) listed in authl authl List of authorization IDs associated with the schemas to be dropped statusl Returned list of status values, one for each authl list entry status Returned status of the function Validates the authorization IDs listed in authl and drops the associated schemas from the database. Before dropping a schema, all tables and views within the schema must be dropped. When you drop a schema, all privileges associated with the schema are also dropped. You cannot drop a schema that is in use. For instance, you cannot drop the current schema or any schema that has privileges being used. If the schema that is dropped is the default schema for any user, the PUBLIC schema becomes the new default schema for the user. Function Reference 261 The PUBLIC schema cannot be dropped from a database. The PUBLIC schema is removed when the database is removed. Before dropping any schemas, the udrpath function first attempts to acquire an exclusive lock on the specified schemas. If the schema is in use, the udrpath function cannot acquire an exclusive lock on the schema and the drop operation fails. Security To drop a schema, you must have one of the following permissions: DBA authority SCHEMA authority and schema access permission For information about granting DBA and SCHEMA authority, see uaddprv; for information about granting schema access permission, see uaddprm. To grant DBA and SCHEMA authority, use the uaddprv function. To grant schema access permission, use the uaddprm function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also uaddath For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 262 Function Reference udrpbt Drop a B-tree Syntax #include <include/rhli.h> #include <include/rhlierr.h> int udrpbt( txid, nbt, bidl, statusl, status ) UTXID txid; int nbt; UBTID bidl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nbt Number of B-tree IDs listed in bidl bidl List of B-tree IDs for B-trees to drop statusl Returned list of status values, one for each bidl list entry status Returned status of the function Validates the B-tree IDs listed in bidl and drops the B-trees from the database. Warning For B-trees that are stored in separate files instead of the database file, when a B-tree is dropped, Unify DataServer removes the external B-tree file (named btxxx.idx) and creates a temporary file (named plxxxxxx) where xxx/xxxxxx are internal index numbers. Do not remove this temporary file because database corruption can occur. Security To drop a B-tree, the schema containing the B-tree must be current, and you must have one of the following permissions: DBA authority SCHEMA authority Function Reference 263 To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also uaddbt For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about using B-trees, see “Managing the Database Files” and “Choosing an Access Method” in Unify DataServer: Managing a Database. 264 Function Reference udrpcgp Drop a column group Syntax #include <include/rhli.h> #include <include/rhlierr.h> int udrpcgp( txid, ncgrp, dbcgrp, statusl, status ) UTXID txid; int ncgrp; DBCGRP *dbcgrp; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID ncgrp Number of column groups in the dbcgrp list dbcgrp List of DBCGRP structures, each describing a separate column group statusl Returned list of status values, one for each dbcgrp entry status Returned status of the function Drops the column groups specified in the dbcgrp list from the database. Also, this function can be used to modify column group options for a column group. If all options for a column group are dropped, then the column group is dropped from the database table. Each DBCGRP structure in the dbcgrp list contains column group information. For a description of the DBCGRP structure, see page 83. Before dropping a column group, the udrpcgp function first acquires an exclusive lock on each column member of the column group and then drops the column group specification, leaving the underlying columns unaffected. If any member column is in use, the exclusive lock attempt fails, and the drop column group operation also fails. Function Reference 265 Security To drop a column group, the schema containing the column group definition must be current, and, you must have one of the following permissions: DBA authority SCHEMA authority To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the udrpcgp function to drop a column group described in the dbcgrp argument. if (udrpcgp(txid, ncgrp, dbcgrp, statusl, &status) != USUCCESS) { fprintf(stderr, ”Couldn’t drop column groups: status = %ld\n”,status); } See Also uaddcgp For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 266 Function Reference udrpcnm Drop a column name (or synonym) Syntax #include <include/rhli.h> #include <include/rhlierr.h> int udrpcnm( txid, ncol, dbcolnm, statusl, status ) UTXID txid; int ncol; DBCOLNM *dbcolnm; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID ncol Number of column name descriptors in dbcolnm dbcolnm List of DBCOLNM structures, each containing a separate column name (or synonym) to drop statusl Returned list of status values, one for each dbcolnm entry status Returned status of the function Validates the column synonyms listed in dbcolnm and drops the column names from the database. Once a column name is dropped, any future reference to the dropped column name fails. Each DBCOLNM structure in the dbcolnm list contains column name information. Column group synonyms can also be dropped by specifying the column group ID (in place of the column ID) in the dbcolnm list. For a description of the DBCOLNM structure, see page 85. Function Reference 267 Security To drop a column name (or synonym), the schema containing the column must be current and you must have one of the following privileges: DBA authority SCHEMA authority To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also uaddcnm, ufchcnm For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 268 Function Reference udrphsh Drop a hash table Syntax #include <include/rhli.h> #include <include/rhlierr.h> int udrphsh( txid, nhsh, hidl, statusl, status ) UTXID txid; int nhsh; UHID *hidl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nhsh Number of hash tables in the hidl list hidl List of hash table IDs of the hash tables to drop statusl Returned list of status values, one for each hidl entry status Returned status of the function Validates the hash table IDs listed in hidl and drops the associated hash tables from the database. Before dropping a hash table, this function attempts to acquire an exclusive lock on the hash table. If the hash table is in use, the udrphsh function cannot acquire an exclusive lock on the hash table and the drop operation fails. Adding or dropping a hash table on a column can have a significant effect on retrieval times of unordered accesses. Security To drop a hash table, the schema containing the hash table definition must be current, and, you must have one of the following permissions: DBA authority SCHEMA authority To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. Function Reference 269 Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the udrphsh function to drop a hash table index. if ( udrphsh(txid, 1, &hid, statusl, &status) != USUCCESS ) printf(”Failure dropping a hash table: status = %ld\n”,status); See Also uaddhsh For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about using hash tables, see “Managing the Database Files,” “Choosing an Access Method,” and “Tuning the Access Methods” in Unify DataServer: Managing a Database. 270 Function Reference udrplnk Drop a link Syntax #include <include/rhli.h> #include <include/rhlierr.h> int udrplnk( txid, nlnks, lidl, statusl, status ) UTXID txid; int nlnks; ULID *lidl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nlnks Number of link IDs in the lidl list lidl List of link IDs to drop statusl Returned list of status values, one for each hidl entry status Returned status of the function Validates the link IDs listed in lidl and drops the associated links for the parent and child tables/columns. Once dropped, a link no longer enforces referential integrity across the two tables. Before you can delete a parent row, you must first drop or change the value of any related rows in the child table. Rows are related through links. For more information about links, see the uaddlnk function. To delete a row, use the udelrow function. When a link is dropped, Unify DataServer places an exclusive lock on both the linked parent and the child columns. Function Reference 271 Security To drop a link, the current schema must contain one or both of the linked tables and you must have one of the following permissions: DBA authority Schema access permission and SCHEMA authority on the schemas that contain the tables that comprise the link definition To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. To grant schema access permission, use the uaddprm function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This function uses the udrplnk function to drop a link index on a table. if ( udrplnk(txid, 1, &lid, statusl, &status) != USUCCESS ) printf(”Error dropping a link: status = %ld\n”,status); See Also uaddlnk For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about using links, see “Choosing an Access Method” in Unify DataServer: Managing a Database. 272 Function Reference udrpprf Drop a volume preference for a table Syntax #include <include/rhli.h> #include <include/rhlierr.h> int udrpprf( txid, tid, nvol, vidl, statusl, status ) UTXID txid; UTID tid; int nvol; UVID *vidl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of table to drop volume preferences for nvol Number of volume IDs in the vidl list vidl List of volume IDs describing the volume preferences to drop statusl Returned list of status values, one for each vidl list entry status Returned status of the function Validates the volume IDs listed in vidl for table tid, and drops the associated volume preferences from the database. The table ID and volume preferences listed must be valid existing entries in the database before the call to the udrpprf function. Security Function Reference To drop a volume preference for a table, the schema containing the table must be current, and, you must have DBA authority. 273 To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the udrpprf function to drop a volume preference from a table. /* Drop the volume preferences */ if (udrpprf(txid, tid, nvol, vidl, statusl, &status) != USUCCESS) { fprintf(stderr, ”Unable to drop volume preferences: %ld\n”, status); for ( i = 0; i < nvol; i++ ) printf(stderr, ”Entry %d volume: %ld status: %ld\n”, i, vidl[i], statusl[i]); exit(1); } See Also uadddb, uaddprf, uaddtbl, uaddvol, udrpvol For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about managing database volumes, see “Managing the Database Files” in Unify DataServer: Managing a Database. 274 Function Reference udrpprm Revoke schema access permission Syntax #include <include/rhli.h> #include <include/rhlierr.h> int udrpprm( txid, userid, nauth, authl, statusl, status ) UTXID txid; UOID userid; int nauth; UAID authl[]; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID userid User ID of user from whom to drop schema access permission nauth Number of authorization IDs listed in authl authl List of authorization IDs for which schema access permission is dropped statusl Returned list of status values, one for each authl list entry status Returned status of the function Revokes schema access permission from a user for the authorization IDs (schemas) listed in authl. Schema access permission allows a user to access a specified schema. When a user is first added to the database (by calling the uaddusr function), the user is granted schema access permission to the PUBLIC schema. To access other schemas in the database, however, a user must be granted schema access permission for each schema the user needs to access. Function Reference 275 If the schema for which schema access permission is being revoked is a user’s default schema, that user’s default schema is reset to the PUBLIC schema. Security To revoke schema access permission from a user, you must have DBA authority or both of the following must be true: The schema is current. You have SCHEMA authority. To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. To grant schema access permission, use the uaddprm function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example calls the udrpprm function to revoke schema access permission from a user. /* define required parameters */ UTXID txid; UOID userid; UAID athl; USTATUS statusl; STATUS * status; int nauth; /* Revoke schema access permission from a user */ if ( udrpprm(txid, userid, nauth, authl, statusl, &status) != USUCCESS ) { fprintf(stderr, ”Unable to drop privileges: %ld\n”, status); exit(1); } See Also uadddb, uaddprm, uaddath For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 276 Function Reference udrpprv Revoke user authority or schema privileges Syntax #include <include/rhli.h> #include <include/rhlierr.h> int udrpprv( txid, nrqstds, rqstdsl, statusl, status ) UTXID txid; int nrqstds; URVKDS rqstdsl[]; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nrqstds Number of privileges described in the rqstdsl list rqstdsl List of URVKDS structures, each describing a separate privilege to revoke statusl Returned list of status values, one for each rqstdsl entry status Returned status of the function Revokes one or more of the following permissions: DBA authority SCHEMA authority Schema privileges to a schema Schema privileges to a table Schema privileges to a column DBA authority allows a user to access and modify any object in the database. SCHEMA authority, however, allows a user to create and drop schemas only. Both of these privileges are granted to users only. Function Reference 277 Schema privileges are privileges granted to a schema. The privileges specify access to another schema or a table or column within the schema. For each set of privileges you want to revoke with this function, you must first describe the privileges by initializing the URVKDS structure. For a description of the URVKDS structure, see page 142. To grant user or schema privileges, use the uaddprv function. Security To revoke DBA or SCHEMA authority, you must have DBA authority. To revoke schema privileges, you must make the schema that granted the privileges the current schema. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example calls the udrpprv function to revoke schema privileges granted to the company table. /* define required parameters */ UTXID txid; UOID userid; UAID uathl; USTATUS statusl; STATUS * status; int nauth; /* Retrieve the authorization ID (aid) for schema ”admin” */ ... /* Retrieve the table id (tid) for the company table */ ... /* Initialize URVKDS structure elements */ rqstdsl.aid = aid; rqstdsl.rvkcaps = UTSEL; rqstdsl.obj.objknd = UTBLKND; rqstdsl.obj.objid = tidl.tid; /* Revoke from schema ”admin” the privilege to perform SELECT operations on table ”company” */ if (udrpprv(txid, 1, &rqstdsl, statusl, &status) != USUCCESS ) { fprintf(stderr, ”Unable to revoke SELECT privilege: status = %ld\n”, status); } 278 Function Reference See Also uaddprv, uaddath, uaddprm, uaddusr, udrpath, udrpprm, udrpusr For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. Function Reference 279 udrptbl Drop a table Syntax #include <include/rhli.h> #include <include/rhlierr.h> int udrptbl( txid, ntbl, tidl, statusl, status ) UTXID txid; int ntbl; UTID *tidl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID ntbl Number of table IDs listed in the tidl list tidl List of table IDs to drop statusl Returned list of status values, one for each tidl list entry status Returned status of the function Validates the table IDs listed in tidl and drops the associated tables from the database. When a table is dropped, all privileges, B-trees, hash tables, and links defined for the table are also dropped. Before dropping a table, you must remove all Data Integrity Subsystem (DIS) definitions for the table. For information about editing a DIS file, see Unify DataServer: Managing a Database When dropping a table, Unify DataServer acquires an exclusive lock on the specified tables, then drops them. If the table is in use, this function cannot acquire an exclusive lock on the table. 280 Function Reference Security To drop a table, the schema containing the table must be current, and you must have one of the following privileges: DBA authority SCHEMA authority To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the udrptbl function to drop a table from the database. ntbl = 3; tidl[0] = $qualified_vendors.$; tidl[1] = $vendor_info.$; tidl[2] = $vendor_transactions.$; /* Drop the tables */ if ( udrptbl(txid, ntbl, tidl, statusl, &status) != USUCCESS ) { fprintf(stderr, ”Unable to drop table: %ld\n”, status); for ( j = 0; j < ntbl; ++j ) printf(stderr, ”Table %ld status: %ld\n”,tidl[j], statusl[j]); exit(1); } See Also uaddtbl For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. Function Reference 281 udrptnm Drop a table name (or synonym) Syntax #include <include/rhli.h> #include <include/rhlierr.h> int udrptnm( txid, ntbl, dbtblnm, statusl, status ) UTXID txid; int ntbl; DBTBLNM *dbtblnm; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID ntbl Number of table name descriptors in dbtblnm list dbtblnm List of DBTBLNM structures, each containing a separate table name to drop statusl Returned list of status values, one for each dbtblnm entry status Returned status of the function Drops the names (or synonyms) for a table. After dropping a table name, any future references to the dropped table name fail. If all names for a table are dropped, the table can be referenced by its table ID only. Each DBTBLNM structure in the dbauth list contains information about each table name. For a description of the DBTBLNM structure, see page 99. Security To drop a table name, the schema containing the table must be current, and, you must have one of the following privileges: DBA authority SCHEMA authority 282 Function Reference To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. To grant schema access permission, use the uaddprm function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the udrptnm function to drop a table synonym. /* Initialize the DBTBLNM structure */ dbtblnm.tblnm = ”partsdata.”; dbtblnm.tid = $parts.$; if (udrptnm(txid, 1, &dbtblnm, statusl, &status) != USUCCESS) printf(”Error dropping a table name: status = %ld\n”,status); See Also uaddtnm For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. Function Reference 283 udrpusr Drop a user Syntax #include <include/rhli.h> #include <include/rhlierr.h> int udrpusr( txid, nuser, useridl, statusl, status ) UTXID txid; int nuser; UOID *useridl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nuser Number of user IDs listed in the useridl list useridl List of user IDs to drop statusl Returned list of status values, one for each useridl list entry status Returned status of the function Validates the user IDs listed in useridl and drops the associated users from the database. You cannot drop yourself or any other users currently accessing the database. Security To drop a user from the database, you must have DBA authority. To grant DBA authority, use the uaddprv function. Status Values 284 For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Function Reference Example This example calls the udrpusr function to remove a user from the database. /* define required parameters */ UTXID txid; int nuser, i; USTATUS statusl; USTATUS status; UOID * useridl; /* Initialize the useridl structure */ ... useridl[0] = usrinf[0].uoid; /* Returned from uinfusr() */ useridl[1] = usrinf[1].uoid; /* Drop the users */ nuser = 2; if ( udrpusr(txid, nuser, useridl, statusl, &status) != USUCCESS) { fprintf(stderr, ”Unable to drop users: %ld\n”, status); for ( i = 0; i < nuser; ++i ) printf(stderr, ”Entry %d user %s status: %ld\n”, i, dbuser[i].usernm, statusl[i]); exit(1); } See Also uaddusr, uaddath, uadddb, uaddprm, uaddusr, udrpath, udrpprm, uinfusr For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. Function Reference 285 udrpvol Drop a volume Syntax #include <include/rhli.h> #include <include/rhlierr.h> int udrpvol( txid, nvol, vidl, statusl, status ) UTXID txid; int nvol; UVID *vidl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nvol Number of volume IDs listed in the vidl list vidl List of volume IDs to drop statusl Returned list of status values, one for each vidl list entry status Returned status of the function Validates the volumes listed in vidl and drops the associated volumes from the database. The root volume cannot be dropped, and any volume which has active storage allocated cannot be dropped. Use the uinfvol function to determine if the volume has any active storage. Volumes designated as a preference volume for a table cannot be dropped. To drop a preference volume, you must first drop all of the preferences by calling the udrpprf function. Warning When a volume that is defined as a file is dropped, Unify DataServer removes the volume file and creates a temporary file (named plxxxxxx, where xxxxxx is an internal index number). Do not remove this temporary file; database corruption will occur. 286 Function Reference Security To drop a volume definition from a database, you must have DBA authority. To grant DBA authority, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the udrpvol function to drop a volume from the database. /* Initialize volume specifications */ nvol = 1; /* number of volumes to drop */ vidl[0] = 2; /* drop volume #2 */ ... /* Drop the volume */ if (udrpvol(txid, nvol, vidl, statusl, &status) != USUCCESS) { fprintf(stderr, ”Unable to drop volumes: %ld\n”, status); for ( i = 0; i < nvol; ++i ) printf(”Entry %d Volume ID %ld status: %ld\n”,i, vidl[i], statusl[i]); uexit(1); } See Also uadddb, uaddvol For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about managing database volumes, see “Managing the Database Files” in Unify DataServer: Managing a Database. Function Reference 287 uendord End an ordered access #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uendord (oaccid, status) UOACCID oaccid; USTATUS *status; Arguments Description oaccid Ordered access ID of the ordered access to end, as returned from the ubegord function status Returned status of the function Ends (or deactivates) an ordered access (specified by oaccid) and releases temporary memory allocated for the ordered access. If a transaction that contains active ordered accesses is ended (with the uabttx, ucbgtx or ucmttx functions), the active ordered accesses are automatically ended. Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also ubegord, ufchord, uposord, uskrord For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 288 uendscn End a scan Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uendscn( scnid, status ) int scnid; USTATUS *status; Arguments Description scnid Scan ID of the scan to end, as returned from the ubegscn function status Returned status of the function Ends (or deactivates) a scan (specified by scnid) and releases temporary memory used for the scan. If a transaction that contains active scans is ended (by calling the uabttx, ucbgtx or ucmttx function), the active scans are ended. To deallocate a scan information structure, use the ufrescn function. Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. 289 See Also uabttx, ualcscn, ubegscn, ucbgtx, ucmttx, ufchscn, ufrescn For information about scanning, see page 64. For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 290 uexit Terminate a process Syntax #include <include/rhli.h> #include <include/rhlierr.h> void uexit( exit_code ) int exit_code; Arguments exit_code Description Terminates a process with the following consequences: An integer exit code. This value is passed to the exit operating system function. All active read-only (type 0) transactions in the calling process are committed. All other active update (type 1 or 2) transactions in the calling process are rolled back. The open database in the calling process is closed. The exit operating system function is called. Failure to use this function to terminate a process will result in the above activities being performed at a later time. This function, as well as the exit operating system function, does not return control to the calling procedure. Security No privileges required. Example See Appendix A. See Also fork(2), wait(2), and exit(3) in your UNIX operating system manual For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 291 uextint Convert column values from external format to internal format #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uextint( txid, tid, unipext, unipcoll, statusl, status ) UTXID txid; UTID tid; UNIPEXT *unipext; UNIPCOLL *unipcoll; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of the table columns contained in UNIPEXT unipext A pointer to a UNIPEXT structure that contains column values in external-format to be converted into internal format unipcoll A pointer to a UNIPCOLL structure that will contain the converted values statusl Returned status value for each column listed in the UNIPCOLL column list cidl status Returned status of the function Converts a list of data values from an external (display) format to an internal database format. This function must be used before inserting or updating data into columns of data type U_DATE, U_HDATE and U_TIME . To read and display date and time data in a Unify DataServer database, you must first convert the data into external format by calling the uintext function. 292 For each column in the external format list described by the UNIPEXT structure (pointed to by unipext), each external column data value is converted to internal format and stored in a UNIPCOLL structure (pointed to by unipcoll). The entries in the unipext list must contain the same number of entries as in the unipcoll list. For a description of the UNIPEXT and UNIPCOLL structures, see pages 140 and 131. All columns specified in each UNIPEXT structure must belong to the specified table ID (tid). Any valopts member contained in the UNIPCOLL structure that contains a value of UIGNORE is by-passed and the entry for statusl is set to UENOWRK. All column values must be converted to internal format before being inserted into the database. See the uinsrow or uupdrow function for more information. If the destination column structure value pointer is null, storage space is allocated by Unify DataServer. This space must later be deallocated by calling the system subroutine free function. If the value pointer is not null, it is assumed that enough space exists for the converted value. A value that cannot be converted is rejected. Check each status value in statusl to determine which values could not be converted. Security To convert column values, you must have access to the column that you want to convert. You can always access all columns in the tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 293 Example See Appendix A. See Also uintext, uinsrow, ufchrow, uupdrow, udelrow For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 294 ufchath Change the current schema Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ufchath( txid, authid, status ) UTXID txid; UAID authid; USTATUS *status; Arguments Description txid Transaction ID authid Authorization ID (for a schema) obtained from a previous call to the ubndath function status Returned status of the function Changes a user’s current schema to the schema specified by authid. The current schema is the schema available to a user. Only one schema can be current at any given time. If the user has adequate access permission, the user can make another schema the current schema. Security To change a current schema, you must have one of the following permissions. DBA authority Schema access permission for the schema you are changing to To grant DBA authority, use the uaddprv function. To grant schema access permission, use the uaddprm function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 295 Example This example calls the ufchath function to make the PUBLIC schema the current schema. /* define required parameters */ UTXID txid; USTATUS * status; /* open the database and begin a transaction */ ... /* make the PUBLIC schema the current schema */ if ( ufchath(txid, $PUBLIC..$, &status) != USUCCESS) printf(”Could not set the current schema: status = %ld\n”, status); See Appendix A for more examples. See Also uaddath, uallath, ubndath, udrpath, uinfath, umapath, umodath, unumath For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 296 ufchcnm Fetch column names (or synonyms) Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ufchcnm( txid, cid, ncnames, colnml, status ) UTXID txid; UCID cid; int *ncnames; char ***colnml; USTATUS *status; Arguments Description txid Transaction ID cid Column ID of the column to fetch the name and synonyms from ncnames Number of column names returned into the colnml list colnml Returned list of column names status Returned status of the function Validates the given column ID and retrieves all names assigned to the column. Each column can have one or more names associated with the column. See uaddcnm for more information about adding column names (or synonyms). All names contained in colnml are zero-byte terminated strings. Also, colnml should be accessed by specifying an argument like argv[ ][ ]. Unify DataServer allocates storage for the returned list of column names (colnml) by calling the malloc function. After you have retrieved the column names, you must recover this storage space by calling the free function. 297 If the number of column names (ncnames) returned is zero (0), the malloc function was not performed by Unify DataServer and you do not need to call the free function. Security Column names are returned only for columns in tables that you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the ufchcnm function to retrieve all column names associated with the co_id column in the company table. /* define required parameters */ char ** colnml; int * ncnames; TXID txid; USTATUS status; ... /* find all the names the ”co_id” column is known by */ if (ufchcnm(txid, $co_id$, &ncnames, &colnml, &status) != USUCCESS) { printf(”Error fetching column names: status = %ld\n”,status); exit( 3 ); } ... /* print the column names */ for( idx = 0; idx < ncnames; ++idx ) printf(”%s\n”, colnml[idx]); ... /* free the space used by colnm */ if ( ncnames > 0 ) free((char *)colnml); See Also uaddtnm, udrptnm, uaddcnm, udrpcnm, uaddtbl 298 free(3) in your UNIX system manual For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 299 ufchjrn Retrieve journal information structure Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ufchjrn (jrnfd, jrninfo, status) int jrnfd; UJRNINF *jrninfo; USTATUS *status; Arguments Description jrnfd The file descriptor of the opened journal file jrninfo A pointer to a UJRNINF structure that contains the returned journal information on the next update operation contained in the journal; for a complete description of the UJRNINF structure, see page 127. status Returned status of the function Retrieves the journal information structure for the next update operation contained in the journal. For delete and update operations, the before image of the row is returned in a UNIPCOL structure. For add and update operations, an after image is returned. For a description of the UNIPCOL structure, see page 131. The UNIPCOL structures for the before and after row images are dynamically allocated by the ufchjrn function. These structures should be freed using the C function, free, to prevent running out of memory. Data for variable length data type fields is not returned by the ufchjrn function. All variable length data is skipped. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 300 Security To process a journal file, one of the following conditions must be true: You have DBA authority. You have schema privilege on all the tables and columns that are referenced in the journal. Example See Appendix A. See Also uopnjrn, uclsjrn For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about transaction management and using the transaction journal, see “Managing Transactions and Locks”and “Preventing Data Loss” in Unify DataServer: Managing a Database. 301 ufchlnk Fetch the parent row through a link index #include <include/rhli.h> #include <include/rhlierr.h> Syntax int ufchlnk( txid, crid, lid, lcktype, rid, unipcoll, statusl, status ) UTXID txid; URID crid; ULID lid; int lcktype; URID *rid; UNIPCOLL *unipcoll; USTATUS statusl[]; USTATUS *status; Arguments 302 txid Transaction ID crid Row ID of a child row of a link lid Link ID of the table to search lcktype Lock type to place on the retrieved row ID. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock rid Returned as the row ID of the parent row retrieved unipcoll A pointer to a UNIPCOLL structure specifying the column retrieval list that corresponds to the parent column statusl Returned list of status values, one for each column described in unipcoll status Returned status of the function Description Retrieves columns of the parent row from the implied table through a link access method. All columns specified in the UNIPCOLL structure must belong to the parent table. Before fetching the parent row, you must obtain the child row ID and specify this value in crid. The UNIPCOLL structure describes the data being retrieved from the database. This structure contains several lists which describe both the column and the column’s value. For a description of the UNIPCOLL structure, see page 131. Warning Column values are verified and valopts is set for all fetch (or retrieval) operations unless UIGNORE is specified. This is true even if the application has set valopts explicitly. Security A row can be returned for linked tables that you have access to only. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 303 Example This example uses the ufchlnk function to fetch a row. #define NCOL 3 /* define required parameters */ UTXID txid; URID crid; URID rid; USTATUS statusl[NCOL]; USTATUS status; UNIPCOLL unipcoll; ... /* retrieve crid */ ... /* retrieve the row */ if ( ufchlnk(txid, crid, $ULNK:LIDparts_used$, USLCK, &rid, &unipcoll, statusl, &status) != USUCCESS) { printf(”Cannot fetch parent row: status=%ld\n”, status); uexit(2); } See Also uaddlnk, ufchtbl, ufchrow, upkfrow For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about using links, see “Choosing an Access Method” in Unify DataServer: Managing a Database. 304 ufchmsg Retrieve the message text corresponding to a status error code Syntax #include <include/rhli.h> #include <include/rhlierr.h> char *ufchmsg( errnum, status ) USTATUS errnum; USTATUS *status; Arguments Description errnum Status error code of the message to fetch status Returned status of the function Fetches the error message text for the status error code specified by errnum. The status error code is usually the status code returned by a previous RHLI function call. Most RHLI functions return a status code into the status argument. This status code reflects the condition of the function operation. A UENORM status code indicates the function operated successfully; all other status codes indicate an error condition. For a listing of status codes and associated mnemonics, see the rhlierr.h header file. For a complete description of error messages, see Unify DataServer: Error Messages. If the ufchmsg function is unable to fetch a message for the error code provided, a status code describing the error is returned into the status argument. Security No privileges required. Return Values Returns a pointer to the error message text. If this function cannot retrieve the error message text, the value (char*) 0 is returned. 305 Example See Appendix A. See Also uinimsg, ulogmsg For information about error handling, see page 51. 306 ufchord Fetch a row from within an ordered access Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ufchord (txid, oaccid, dir, lcktype, ridp, unipcoll, statusl, status) UTXID txid; UOACCID oaccid; int dir; int lcktype; URID *ridp; UNIPCOLL *unipcoll; USTATUS statusl[]; USTATUS *status; Arguments txid Transaction ID oaccid An ordered access ID as returned from the ubegord function dir Direction to fetch the row, relative to the current position. Options are: lcktype USAME Get same (current) row UFIRST Get first row UNEXT Get next row UPREV Get previous row ULAST Get last row Lock type to place on retrieved row. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock 307 Description ridp Returned as the row ID of the retrieved row unipcoll A pointer to a UNIPCOLL structure specifying the column retrieval list statusl Returned list of status values, one for each column value returned into the unipcoll list status Returned status of the function Retrieves columns of the row indicated by dir from the specified ordered access. The order in which the rows are retrieved is specified by the column order list (UOACOLS) for the order access (initiated by the ubegord function). Warning Column values are verified and valopts is set for all fetch (or retrieval) operations unless UIGNORE is specified. This is true even if the application has set valopts explicitly. The UNIPCOLL structure describes the data being retrieved from the database. This structure contains several lists which describe both the column and the column’s value. If the function returns USUCCESS, the row ID and requested columns are returned with a lock no lower than the requested lock. The current position within the ordered access is set to the retrieved row. The UNEXT option can be specified (instead of UFIRST) immediately after the ubegord function call to fetch the first row of the table. The txid specified to the ufchord function may differ from the txid used by the ubegord function. If a different txid is used for the ufchord function, then the transaction can be committed to release locks on the current record without affecting the ordered access. The direction parameter is the order specified by the UOACOLS structure, not the physical order of the underlying access method. Therefore, for an ascending order, UFIRST refers to the row whose columns have the lowest value and ULAST to the row whose columns have the highest value. For a descending order, UFIRST refers to the row whose columns have the highest value and ULAST to the row whose columns have the lowest value. 308 Columns specified in UNIPCOLL must belong to the table associated with the ordered access. If the function fails due to the inability to obtain the specified row lock, the current position is set to the fetched row. Additional Help See “Function Structures,” for a description of the structure UNIPCOLL. Security To retrieve the column values for the row, one of the following conditions must be true: You have DBA authority. The target table is in the current schema. You have SELECT schema privilege on the target table. You have SELECT schema privilege on all columns to be retrieved. To make a schema current, see the ufchath function; for information about granting DBA authority and schema privileges, see the uaddprv function. UEDBCLS Database is closed UEBDATYP Bad access type (not an ordered access ID) UEBDSCN Invalid ordered access ID UENODIR Invalid direction specification Example See Appendix A. See Also ubegord, uendord, uposord, uskrord For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 309 ufchrow Fetch a row #include <include/rhli.h> #include <include/rhlierr.h> Syntax int ufchrow( txid, tid, lcktype, rid, unipcoll, statusl, status ) UTXID txid; UTID tid; int lcktype; URID rid; UNIPCOLL *unipcoll; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of the table to fetch from lcktype Lock type to place on the retrieved row ID. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock rid Row ID of the row to retrieve unipcoll A pointer to a UNIPCOLL structure specifying the column retrieval list statusl Returned list of status values, one for each column specified in the unipcoll column list status Returned status of the function Given valid table and row IDs, this function retrieves specified columns of the row. Columns specified in UNIPCOLL must belong to the specified table ID. Warning Column values are verified and valopts is set for all fetch (or retrieval) operations unless UIGNORE is specified. This is true even if the application has set valopts explicitly. 310 The UNIPCOLL structure describes the data retrieved from the database. This structure contains several lists describing the column and its value. For a description of the UNIPCOLL structure, see page 131. Security A row can only be returned for tables you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the ufchrow function to fetch a row from the company table. #define NCOL 3 /* define required parameters */ UTXID txid; URID rid; /* Row ID */ USTATUS status; /* function status */ USTATUS statusl[NCOL]; /* status value list */ UNIPCOLL unipcoll; ... /* Retrieve rowid of row to be fetched */ ... /* Fetch the row */ if (ufchrow(txid, $company.$, USLCK, rid, &unipcoll, statusl, &status) != USUCCESS ) { printf(”data retrieval failed: %ld\n”, status); uexit(101); } See Appendix A for more examples. See Also upkfrow, ufchlnk For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 311 ufchscn Retrieve the next row from a scan #include <include/rhli.h> #include <include/rhlierr.h> Syntax int ufchscn( scanid, ridp, dbvall, prevlck, statusl, status ) int scanid; URID *ridp; UDBVAL *dbvall; int *prevlck; USTATUS statusl[]; USTATUS *status; Arguments Description scanid A scan ID as returned from the ubegscn function ridp Returned as a row ID of the retrieved row dbvall Pointer to a list of UDBVAL structures, each specifying a different column retrieval buffer prevlck Returned as the previous lock status of the retrieved row statusl Returned list of status values, one for each column value returned into the dbvall list status Returned status of the function Retrieves the next row along the specified scan. The order in which the rows are retrieved is specified by the sort criteria for the scan. Warning Column values are verified and valopts is set for all fetch (or retrieval) operations unless UIGNORE is specified. This is true even if the application has set valopts explicitly. 312 A list of UDBVAL structures should be allocated to correspond to the elements of the previously specified projection list. The order of the columns returned is the same as that of the projection list. The UDBVAL structure can be allocated by the application or through the uinfscn function. For a description of UDBVAL, found in the structure UNIPCOLL, see page 131. If the function returns USUCCESS, the row ID and requested columns are returned with a lock no lower than the requested lock. The prevlck member is set to the lock previously held on the row by the calling transaction. If the function returns UFAILURE, the rowid and column values are undefined. If a row is passed but the lock could not be granted (STATUS = UELMOUT) then prevlck is set to the conflicting lock type. If the application did not call uinsprj (the number of projected columns is zero) then dbvall and statusl may be null. When inserting into or updating columns of type U_STR (string), Unify DataServer adds trailing blanks to values less than the total size defined for the column. Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also ualcscn, uendscn, ubegscn, ufrescn, uinfscn, ufchscn, uinsprj, uinssrt, uinsor For information about scanning, see page 64. 313 ufchtbl Fetch a row sequentially #include <include/rhli.h> #include <include/rhlierr.h> Syntax int ufchtbl( txid, tid, dir, lcktype, rid, unipcoll, statusl, status ) UTXID tx; UTID tid; int dir; int lcktype; URID *rid; UNIPCOLL *unipcoll; USTATUS statusl[]; USTATUS *status; Arguments txid Transaction ID tid Table ID of the table to fetch from dir Direction to fetch the row, relative to the current position. Options are: lcktype rid 314 UFIRST Get first row UNEXT Get next row UPREV Get previous row ULAST Get last row Lock type to place on the retrieved row ID. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock The row ID. Used for UNEXT and UPREV operations (specified by dir), the next or previous row is relative to the row ID passed in. For all operations the row ID of the row retrieved is returned. Description unipcoll A pointer to a UNIPCOLL structure specifying the column retrieval list statusl Returned list of status values, one for each column specified in the unipcoll column list status Returned status of the function Retrieves columns of the indicated row from the specified table. Columns specified in UNIPCOLL must belong to the specified table ID. The UFIRST option must be used when retrieving the first row of the table. This sets the current position for subsequent retrievals. Warning Column values are verified and valopts is set for all fetch (or retrieval) operations unless UIGNORE is specified. This is true even if the application has set valopts explicitly. The UNIPCOLL structure describes the data retrieved from the database. This structure contains several lists which describe both the column and the column’s value. For a description of the UNIPCOLL structure, see page 131. Security A row can be returned only for tables you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant DBA authority and schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 315 Example This example uses the ufchtbl function to fetch a row from the company table. #define NCOL 2 /* define required parameters */ int dir; /* fetch direction */ UTXID txid; URID rid; /* Row ID */ USTATUS status; /* function status */ USTATUS statusl[NCOL]; /* status value list */ UNIPCOLL unipcoll; ... /* Retrieve rowid of row to be fetched (rid) */ ... /* Fetch the row */ dir = UFIRST; if (ufchtbl(txid, $company.$, dir, USLCK, &rid, &unipcoll, statusl, &status) != USUCCESS) { printf(”data retrieval failed: %ld\n”, status); uexit(101); } See Also ufchlnk, ufchrow, upkfrow For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 316 ufchtnm Fetch table names (or synonyms) Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ufchtnm( txid, tid, ntnames, tblnml, status ) UTXID txid; UTID tid; int *ntnames; char ***tblnml; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of the table to fetch names from ntnames Number of table names returned to tblnml tblnml Returned list of pointers to table names status Returned status of the function Validates the table (specified by tid) and retrieves all names assigned to the table. Each table can have one or more names associated with the table. See the uaddtnm function for more information about adding table names (or synonyms). All table names returned into tblnml are zero byte-terminated strings. Unify DataServer allocates storage for the returned list of table names (tblnml) by calling the malloc function. After you have retrieved the table name, you must recover this storage space by calling the free function. If the number of table names (ntnames) returned is zero (0), the malloc function was not called by Unify DataServer and you do not need to call the free function. 317 Security Table names are returned only for tables you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the ufchtnm function to fetch and print all names that the company table is known by. #define NCOL 2 /* define required parameters */ int * ntnames; char ** tblnml; UTXID txid; USTATUS status; /* function status */ USTATUS statusl[NCOL]; /* status value list */ ... } /* find all the names the ”company” table is known by */ if (ufchtnm(txid, $company$, &ntnames, &tblnml, &status) != USUCCESS) { fprintf(stderr,”Error fetching table names: status = %ld\n”,status); uexit(99); } for( idx = 0; idx < ntnames; ++idx ) printf(”%s\n”, tblnml[idx]); /* free the space used by tblnml */ if ( ntnames > 0 ) free((char *)tblnml); See Also uaddcnm, uaddtnm, udrpcnm, udrptnm, ufchcnm free(3) in your UNIX system manual For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 318 ufrescn Free or clear a scan information structure Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ufrescn( txid, scaninf, rlsopts, status ) UTXID txid; char **scaninf; UOPTS rlsopts; USTATUS *status; Arguments txid Transaction ID scaninf A scan information structure to free as returned from the ualcscn function rlsopts Operation indicator; options are: DB_RESET Reset the scan structure; do not release. The DB_RESET option is used to reuse the scan structure and should not be used with any other options. status DB_RELEAS Release the scan information structure. DB_FREPRJ Release only the projection specification. DB_FRESEL Release only the selection criteria. DB_FRESRT Release only the sort order specification. Returned status of the function Description Frees the memory belonging to the scan information structure indicated. This function does not deactivate the scan. A scan stays active until a call is made to the uendscn function. Security No privileges required. 319 Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also ualcscn, ubegscn, uendscn, ufchscn, uinssrt, uinsprj, uinsor, uinsand For information about scanning, see page 64. 320 uinfacs Retrieve access information for a scan or ordered access Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uinfacs( scnid, acsinfp, status ) int scnid; UACSINF *acsinfp; USTATUS *status; Arguments Description scnid The ID returned by either the ubegscn or ubegord function acsinfp UACSINF structure to hold access information status Contains a value that indicates the status of the function’s operation. The uinfacs function should be called after the ubegscn or ubegord function is called, and before the ufchscn or ufchord function is called. When a scan or ordered access is initiated with the ubegscn or ubegord function, the optimizer analyzes the selection and database criteria, and chooses the most efficient access method. The uinfacs function retrieves this access type information by filling a UACSINF structure for the given ID. The information includes the access type chosen by the optimizer, and can be one of the following: USQSCN Sequential access UDKSCN Direct key access UBTSCN BĆtree index access UHKSCN Hash index access ULKSCN Link index access UNULLSCN No access type, not initialized 321 UFAILURE A failure occurred during the operation; check status value for the error. And, depending on access type, may include other information: UBTSCN BĆtree ID and number of components used. UHKSCN Hash table ID ULKSCN Link ID Other functions may then be used to retrieve additional information about the particular index chosen by the optimizer. For example: uinfbt BĆtree information uinfhsh Hash table information uinflnk Link information. Security Subject to the same restrictions as the scan or ordered access in question. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also uinfbt, uinfhsh, uinflnk For information about scanning, see page 64. For information about access methods, see “Choosing an Access Method” in Unify DataServer: Managing a Database. 322 uinfath Retrieve schema information Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uinfath( txid, nauth, athidl, infclas, athinfl, statusl, status ) UTXID txid; int nauth; UAID *athidl; UOPTS infclas; UATHINF *athinfl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nauth Number of authorization IDs in the athidl list athidl List of authorization IDs (associated with schemas) to retrieve information about infclas Type of information requested. Options are: UNOCLASS Retrieve essential information only UCLASS1 Retrieve Class 1 information only UALLINFO Retrieve all information athinfl List of UATHINF structures, each containing retrieved information about each authorization ID listed in athidl statusl Returned list of status values, one for each athidl/athinfl entry status Returned status of the function Returns a UATHINF schema information structure for each authorization ID listed in athidl. For a description of the UATHINF structure, see “The RHLI Structures” chapter. The parameter infclas controls how much schema information is returned to each UATHINF structure. 323 Information retrieval classes (specified by inflclas) are ordered according to system overhead usage: UNOCLASS information uses the least amount of overhead, while UALLINFO uses the most. The UNOCLASS option specifies that only essential information defined by the UATHINF structure is to be retrieved. No optional information is retrieved. The UNOCLASS information is always retrieved, no matter which retrieval class you choose. The UCLASS1 category (which is defined in the structure UATHINF) retrieves the specified information for that class only. The UALLINFO option specifies that all the information defined for the UATHINF structure is to be retrieved. You can specify more than one retrieval class by separating options with the bitwise logical OR operator (|). Security Schema information is returned only for schemas that you have access to. You always have access to the current schema. Additionally, you can access other schemas if you have been granted schema access permission to the schema or if a schema you can access has schema privileges to another schema (or a table or column in the schema). If you have DBA authority, you can access any schema in the database. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. To grant schema access, use the uaddprm function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 324 Example This example calls the uinfath function to retrieve information about the current schema. /* define required parameters */ UTXID txid; int nauth; UAID * athidl; UATHINF * athinfl; USTATUS statusl; STATUS status; /* retrieve authorization ID (athidl) */ ... /* retrieve authorization information */ if ( uinfath(txid, 2, athidl, UALLINFO, athinfl, statusl,&status) != USUCCESS) printf(”Unable to get info on schema: status = %ld\n”,status); See Also uaddath, uallath, ubndath, udrpath, ufchath, umapath, umodath, unumath For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 325 uinfbt Retrieve B-tree information for a list of B-trees #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uinfbt( txid, nbt, btidl, infclas, btinfl, statusl, status ) UTXID txid; int nbt; UBTID *btidl; UOPTS infclas; UBTINF *btinfl; USTATUS statusl[]; USTATUS *status; Arguments 326 txid Transaction ID nbt Number of B-tree IDs in the btidl list btidl List of B-tree IDs to retrieve information about infclas Information retrieval class. Options are: UNOCLASS Retrieve essential information only UCLASS1 Retrieve Class 1 information only UCLASS2 Retrieve Class 2 information only UCLASS3 Retrieve Class 3 information only UALLINFO Retrieve all information btinfl Returned list of UBTINF structures, each containing information about a separate B-tree (corresponding to the B-tree IDs listed in btidl) statusl Returned list of status values, one for each btinfl entry status Returned status of the function Description Returns a UBTINF B-tree information structure for each B-tree ID listed in btidl. The parameter infclas controls how much B-tree information is returned to each UBTINF structure. You can specify more than one option by separating the options with the bitwise logical operator (|). Information retrieval classes (specified by infclas) are ordered according to system overhead usage: UNOCLASS information uses the least amount of overhead, while UALLINFO uses the most. The UNOCLASS option specifies that only essential information defined by the UBTINF structure is to be retrieved. No optional information is desired. The UNOCLASS option information is always retrieved, no matter which retrieval class you choose. The UCLASS1, UCLASS2 and UCLASS3 categories (which are described in the structure UBTINF), retrieve the specified information for that class only. The UALLINFO option specifies that all the information defined for the UBTINF structure is to be retrieved. For a description of the UBTINF structure, see “The RHLI Structures” chapter. Security B-tree information can be retrieved only for B-trees on tables that you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Volume information is only returned if you have DBA authority. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. To grant schema access, use the uaddprm function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 327 Example This example uses the uinfbt function to retrieve B-tree information. if ((btidl = (UBTID *)calloc(2, sizeof(UBTID))) == NULL) uexit(1); if (ubndbt(txid, $parts$, 1, ”part_number”, &btidl[0], &statusl[0],&status) != USUCCESS) { free (btidl); uexit(1); } if (ubndbt(txid, $parts$, 1, ”ord_number”, &btidl[1], &statusl[1], &status) != USUCCESS) { free (btidl); uexit(1); } if ((btinfl = (UBTINF *)calloc(2, sizeof(UBTINF))) == NULL) { free (btidl); uexit(1); } if (uinfbt(txid, 2, btidl, (UCLASS1|UCLASS2), btinfl, statl, &status) != USUCCESS) { printf(”Unable to get info on B-trees: status = %ld\n”,status); free (btidl); free (btinfl); uexit(1); } See Also ubndbt For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about using B-trees, see “Managing the Database Files” and “Choosing an Access Method” in Unify DataServer: Managing a Database. 328 uinfcgp Retrieve column group information for a list of columns Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uinfcgp( txid, ncol, cidl, infclas, colinfl, statusl, status ) UTXID txid; int ncol; UCID *cidl; UOPTS infclas; UCGPINF *cgpinfl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID ncol Number of column IDs in the cidl list cidl List of column IDs to retrieve information about infclas Information retrieval class. Must be set to UNOCLASS. cgpinfl Returned list of UCGPINF structures, each containing information about a separate column group (corresponding to the cidl list) statusl Returned list of status values, one for each cgpinfl entry status Returned status of the function Returns a UCGPINF column group information structure for each column ID listed in cidl. Any column ID listed in cidl can be either a group member or a group leader. If a column’s type is U_CGP, the returned list of column IDs is of the component group members. Otherwise, the list of column IDs is of the column groups the component participates in. 329 The parameter infclas should be set to UNOCLASS. This indicates that essential information is to be retrieved. All column group information is essential, and therefore is retrieved. The parameter cginfl points to a list of UCGPINF structures that contain the returned column group information. For a description of the structure UCGPINF, see “The RHLI Structures” chapter. For successfully retrieved information structures where the number of column IDs returned is non-zero, the UCGPINF structure member grpcidl, is allocated storage space by calling the malloc function. This space should be released when the information is no longer needed by calling the free function. Security Column group information can be retrieved only for columns in tables that you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the uinfcgp function to retrieve information about a column group. pcol[0] = ”part_number”; pcol[1] = ”part_name”; /* Retrieve the Column IDs to get info on */ if (ubndcol(txid, $parts$, 2, pcol, USLCK, &cidl, statusl, &status) != USUCCESS) uexit(1); if ( uinfcgp(txid,2,cidl,UALLINFO,cgpinfl,statusl,&status) != USUCCESS ) { printf(”Unable to get info on col. group: status = %ld\n”,status); uexit(1); } 330 See Also uaddcgp, uallcgp, udrpcgp, unumcgp malloc(3) and free(3) in your UNIX system manual For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 331 uinfcol Retrieve column information for a list of columns #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uinfcol( txid, ncol, cidl, infclas, colinfl, statusl, status ) UTXID txid; int ncol; UCID *cidl; UOPTS infclas; UCOLINF *colinfl; USTATUS statusl[]; USTATUS *status; Arguments 332 txid Transaction ID. ncol Number of column IDs in the cidl list cidl List of column IDs to retrieve information about infclas Information retrieval class. Options are: UNOCLASS Retrieve essential information only UCLASS1 Retrieve Class 1 information only UCLASS2 Retrieve Class 2 information only UCLASS3 Retrieve Class 3 information only UCLASS4 Retrieve Class 4 information only UALLINFO Retrieve all information colinfl Returned list of UCOLINF structures, each containing information about a separate column (corresponding to the cidl list) statusl Returned list of status values, one for each colinfl entry status Returned status of the function Description Returns a UCOLINF column information structure for each column ID listed in cidl. The parameter infclas controls how much column information is returned to each UCOLINF structure. You can specify more than one option by separating the options with the bitwise logical operator (|). Information retrieval classes (specified by infclas) are ordered according to system overhead usage: UNOCLASS information uses the least amount of overhead, while UALLINFO uses the most. The UNOCLASS option specifies that only essential information defined by the UCOLINF structure is to be retrieved. No optional information is retrieved. The UNOCLASS information is always retrieved, no matter which retrieval class you choose. The UCLASS1, UCLASS2, UCLASS3 and UCLASS4 categories (which are defined in the structure UCOLINF), retrieve the specified information for that class only. The UALLINFO option specifies that all the information defined for the UCOLINF structure is to be retrieved. For a description of the UCOLINF structure, see “The RHLI Structures” chapter. For columns successfully retrieved by specifying UCLASS4 or UALLINFO retrieval classes, storage space is allocated by Unify DataServer for UCOLINF structure members coldesc and dsppict calling the malloc function. This space should be released when the information is no longer needed by calling the free function. Security Column information can be retrieved only for columns you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 333 Example This example uses the uinfcol function to retrieve information about a column. pcol[0] = ”part_number”; pcol[1] = ”part_name”; /* Retrieve the Column IDs to get info on */ if (ubndcol(txid, $parts$, 2, pcol, USLCK, &cidl, statl, &status) != USUCCESS) uexit(1); if (uinfcol(txid,2,cidl,UALLINFO,colinfl,statusl,&status) != USUCCESS) { printf(”Unable to get info on columns: status = %ld\n”,status); uexit(1); } See Also ubndtbl, ubndcol malloc(3) and free(3) in your UNIX system manual For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 334 uinfdb Retrieve database information Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uinfdb( txid, ndb, dbidl, infclas, dbinfl, statusl, status ) UTXID txid; int ndb; UDBID *dbidl; UOPTS infclas; UDBINF *dbinfl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID ndb The number of database IDs in the dbidl list dbidl List of database IDs to retrieve information about infclas Information retrieval class. Options are: UNOCLASS Retrieve essential information only UCLASS1 Retrieve Class 1 information only UCLASS2 Retrieve Class 2 information only UCLASS3 Retrieve Class 3 information only UALLINFO Retrieve all information dbinfl Returned list of UDBINF structures, each containing information about a separate database (corresponding to the dbidl list) statusl Returned list of status values, one for each dbinfl entry status Returned status of the function Returns a UDBINF database information structure for each valid database ID. The parameter infclas controls how much database information is returned to each UDBINF structure. You can specify more than one option by separating the options with the bitwise logical operator (|). 335 Information retrieval classes are ordered according to system overhead usage: UNOCLASS information uses the least amount of overhead, while UALLINFO uses the most. The UNOCLASS option specifies that only essential information defined by the UDBINF structure is to be retrieved. No optional information is retrieved. The UNOCLASS information is always retrieved, no matter which retrieval class you choose. The UCLASS1, UCLASS2 and UCLASS3 categories, which are defined in the structure UDBINF, retrieve the specified information for that class only. The UALLINFO option specifies that all the information defined for the UDBINF structure is to be retrieved. For a description of the UDBIN structure, see “The RHLI Structures” chapter. For columns successfully retrieved by specifying UCLASS2, UCLASS3 or UALLINFO retrieval classes, storage space is allocated by Unify DataServer for UDBINF structure members uvidlst and descrpt by calling malloc function. This space should be released when the information is no longer needed by calling the free function. Security No privileges required. However, volume information is only returned if you have DBA authority. For information about granting DBA authority, see uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 336 Example This example uses the uinfdb function to retrieve information about the current database. if (uinfdb(txid,ndb,dbidl,UALLINFO,dbinfl,statusl,&status) != USUCCESS) { printf(”Unable to get info on database: status = %ld\n”,status); uexit(1); } else { for(dbindx=0; dbindx < ndb; ++dbindx) { prdbinf(&dbinfl[dbindx]); for(volindx=0; volindx < dbinfl[dbindx].nvol; ++volindx) prvolin(&(dbinfl[dbindx].uinfvol[volindx])); if ( ndb > 0 ) free(dbinfl[dbindx].uinfvol); } } See Also uadddb, ubnddb, uclsdb, ulckdb, umoddb, uopndb chmod(2), chown(2), malloc(3), and free(3) in your UNIX system manual For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about managing database volumes, see “Managing the Database Files” in Unify DataServer: Managing a Database. 337 uinfhsh Retrieve hash index information #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uinfhsh( txid, nht, hidl, infclas, hinfl, statusl, status ) UTXID txid; int nht; UHID *hidl; UOPTS infclas; UHTINF *hinfl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nht Number of hash table IDs in the hidl list hidl List of hash table IDs to retrieve information about infclas Information retrieval class. Options are: UNOCLASS Retrieve essential information only UCLASS1 Retrieve Class 1 information only UCLASS2 Retrieve Class 2 information only UCLASS3 Retrieve Class 3 information only UALLINFO Retrieve all information hinfl Returned list of UHTINF structures, each containing information about a separate hash table (corresponding to the hidl list) statusl Returned list of status values, one for each hinfl entry status Returned status of the function Returns a UHTINF hash table information structure for each hash table ID listed in hidl. The parameter infclas controls how much hash table information is returned to each UHTINF structure. You can specify more than one option by separating the options with the bitwise logical operator (|). 338 Information retrieval classes (specified by infclas) are ordered according to system overhead usage: UNOCLASS information uses the least amount of overhead, while UALLINFO uses the most. The UNOCLASS option specifies that only essential information defined by the UHTINF structure is to be retrieved. No optional information is retrieved. The UNOCLASS information is always retrieved, no matter which retrieval class you choose. The UCLASS1, UCLASS2 and UCLASS3 categories (which are defined in the structure UHTINF), retrieve the specified information for that class only. The UALLINFO option specifies that all the information defined for the UHTINF structure is to be retrieved. For a description of the UHTINF structure, see “The RHLI Structures” chapter. For columns successfully retrieved by specifying UCLASS2 or UALLINFO retrieval classes, storage space is allocated by Unify DataServer for UHTINF structure member htloc by calling the malloc function. This space should be released (by calling the free function) when the information is no longer needed. Security Hash table information can be retrieved only for hash indexes on tables that you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Volume information is only returned if you have DBA authority. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 339 Example This example uses the uinfhsh function to retrieve information about a hash index. if (ubndhsh(txid, $parts$, 1, ”part_number”, &hidl[0], &statusl[0], &status) != USUCCESS) uexit(1); if (ubndhsh(txid, $parts$, 1, ”ord_number”, &hidl[1], &statusl[1], &status) != USUCCESS) uexit(1); if (uinfhsh(txid,2,hidl,UALLINFO,hinfl,statusl,&status) != USUCCESS) { printf(”Unable to get info on hash tables: status = %ld\n”,status); uexit(1); } See Also uaddhsh, uallhsh, ubndath, udrphsh, umaphsh, umodhsh malloc(3) and free(3) in your UNIX system manual For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about using hash tables, see “Managing the Database Files,” “Choosing an Access Method,” and “Tuning the Access Methods” in Unify DataServer: Managing a Database. 340 uinflck Retrieve data lock information Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uinflck( txid, tid, rid, lcktype, status ) UTXID txid; UTID tid; URID rid; int *lcktype; USTATUS *status; Arguments txid The current transaction ID tid The table ID of the row’s table or UNULLTID rid The row ID of the row whose locking status is checked or UNULLRID lcktype status Description Returned lock type of the object. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock Returned status of the function Retrieves data lock information for a row, table, or database. The lock type is returned into lcktype. To retrieve lock information for a row, set tid to the row table ID and rid to the row ID. To retrieve data lock information for a table, set tid to the row’s table ID and rid to UNULLRID. 341 To retrieve data lock information for a database, set tid to UNULLTID and rid to UNULLRID. The lock information being retrieved must match exactly with the lock placed on the object. For example, if a row lock is promoted to a table lock and you retrieve lock information for a row in that table, uinflck function returns an error. To retrieve definition lock information for a table, use the uinfsch function. Security Table data lock information can be retrieved only for tables you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the uinflck function to retrieve information about row, table, and database locks. /* define required parameters */ int lcktype URID rid; int prevlck; USTATUS fstatus; UTXID txid; USTATUS status; ... /* open the database */ ... /* retrieve row ID */ ... /* retrieve row lock information */ if (uinflck(txid, $company.$, rid, &lcktype, &status) != USUCCESS) { printf(”row is not locked\n”); uexit(101); } printf(”row is locked with type %d\n”, lcktype); 342 /* retrieve table lock information */ if (uinflck(txid, $company.$, UNULLRID, &lcktype, &status) != USUCCESS) { printf(”table is not locked\n”); uexit(102); } printf(”table is locked with type %d\n”, lcktype); /* retrieve database lock information */ if (uinflck(txid, UNULLTID, UNULLRID, &lcktype, &status) != USUCCESS) { printf(”database is not locked\n”); uexit(103); } printf(”database is locked with type %d\n”, lcktype); See Also uinfsch, ulckdb, ulckrow, ulcksch, ulcktbl, uulkdb, uulkrow, uulksch, uulktbl For information about lock management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 343 uinflnk Retrieve link information #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uinflnk( txid, nlnk, lidl, infclas, lnkinfl, statusl, status ) UTXID txid; int nlnk; ULID *lidl; UOPTS infclas; ULNKINF *lnkinfl; USTATUS statusl[]; USTATUS *status; Arguments 344 txid Transaction ID nlnk Number of Link IDs in the lidl list lidl List of Link IDs to retrieve information about infclas Information retrieval class. Options are: UNOCLASS Retrieve essential information only UCLASS1 Retrieve Class 1 information only UCLASS2 Retrieve Class 2 information only UALLINFO Retrieve all information lnkinfl Returned list of ULNKINF structures, each containing information about a separate link (corresponding to the lidl list) statusl Returned list of status values, one for each lnkinfl entry status Returned status of the function Description Returns a ULNKINF link information structure for each valid Link ID listed in lidl. The parameter infclas controls how much link information is returned to each ULNKINF structure. You can specify more than one option by separating the options with the bitwise logical operator (|). Information retrieval classes (specified by infclas) are ordered according to system overhead usage: UNOCLASS information uses the least amount of overhead, while UALLINFO uses the most. The UNOCLASS option specifies that only essential information defined by the ULNKINF structure is to be retrieved. No optional information is retrieved. The UNOCLASS information is always retrieved, no matter which retrieval class you choose. The UCLASS1 and UCLASS2 categories (which are defined in the structure ULNKINF), retrieve the specified information for that class only. The UALLINFO option specifies that all the information defined for the ULNKINF structure is to be retrieved. For a description of the ULNKINF structure, see “The RHLI Structures” chapter. For columns successfully retrieved by specifying UCLASS1 or UALLINFO retrieval classes, storage space is allocated by Unify DataServer for UHTINF structure members ploc and cloc by calling the malloc function. This space should be released (by calling the free function) when the information is no longer needed. Security Link information is returned only for links on tables that you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 345 Example This example uses the uinflnk function to retrieve link information. if (ubndlnk(txid, $parts$, 1, ”part_number”, &lidl[0], &statusl[0],&status) != USUCCESS) { uexit(1); } if (ubndlnk(txid, $parts$, 1, ”ord_number”, &lidl[1], &statusl[1], &status) != USUCCESS) if ((lnkinfl = (ULNKINF *)calloc(2, sizeof(ULNKINF))) == NULL) { uexit(1); { if (uinflnk(txid,2,lidl,UALLINFO,lnkinfl,statusl,&status) != USUCCESS ) { fprintf(stderr,”Unable to get info on links: status = %ld\n”,status); uexit(1); } See Also uaddlnk, ualllnk, ubndlnk, udrplnk, ufchlnk, umaplnk, umodlnk malloc(3) and free(3) in your UNIX system manual For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about using links, see “Choosing an Access Method” in Unify DataServer: Managing a Database. 346 uinfsch Retrieve definition lock information for a table Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uinfsch( txid, tid, lcktype, status ) UTXID txid; UTID tid; int *lcktype; USTATUS *status; Arguments txid The current transaction ID tid The table ID for the table you are inquiring about lcktype Returned lock type of the table. Options are: status Description UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock Returned status of the function Retrieves the current definition lock type (lcktype) on the specified table (tid). A definition lock prevents any other transactions from changing the definition of the table. A data lock, however, prevents other operations from changing data contained in the table rows. To retrieve data lock information for a table, use the uinflck function. Security Table definition lock information can be retrieved only for tables you have access to. 347 You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the uinfsch function to retrieve information about object definition locks on the company table. The lock type is printed. /* define required parameters */ int lcktype UTXID txid; USTATUS status; ... /* retrieve table definition lock information */ if (uinfsch(txid, $company.$, &lcktype, &status) != USUCCESS ) printf(”Table definition is not locked\n”); else printf(”Table definition is locked with type %d\n”, lcktype); See Also uinflck, ulcksch, uulksch For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about lock management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 348 uinfscn Create a list used for scan data retrieval Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uinfscn( txid, scaninf, dbvallp, status ) UTXID txid; char *scaninf; UDBVAL **dbvallp; USTATUS *status; Arguments Description txid Transaction ID scaninf Pointer to a scan information structure dbvallp Location for storing the pointer to the allocated UDBVAL list status Returned status of the function Builds a database value list from a pre-established scan projection list. The scaninf member points to the scan information structure created by the ualcscn function. The projection criteria for the scan must be specified (by calling the uinsprj function) before calling the uinfscn function. The uinfscn function allocates a UDBVAL list (and column value buffers) that matches the projection criteria specified in scaninf. The pointer to the allocated UDBVAL list is returned in the location pointed to by dbvallp. The UDBVAL items appear in the same order that the columns were specified in the uinsproj function reference. The allocated UDBVAL list is used in future calls to the ufchscn function. When the scan is completed, you must free the memory that was allocated for the UDBVAL list. For a description of the UDBVAL structure, see the UNIPCOLL structure description on “The RHLI Structures” chapter. 349 Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also uinsor, uinsand, uinssrt, uinsprj, ufchscn, ufrescn, ubegscn, uendscn, ufchscn For information about scanning, see page 64. 350 uinftbl Retrieve table information Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uinftbl( txid, ntbl, tidl, infclas, tblinfl, statusl, status ) UTXID txid; int ntbl; UTID *tidl; UOPTS infclas; UTBLINF *tblinfl; USTATUS statusl[]; USTATUS *status; Arguments txid Transaction ID ntbl Number of table IDs in the tidl list tidl List of table IDs to retrieve information about infclas Information retrieval class. Options are: UNOCLASS Retrieve essential information only UCLASS1 Retrieve Class 1 information only UCLASS2 Retrieve Class 2 information only UCLASS3 Retrieve Class 3 information only UCLASS4 Retrieve Class 4 information only UALLINFO Retrieve all information tblinfl Returned list of UTBLINF structures, each containing information about a separate table (corresponding to the tidl list) statusl Returned list of status values, one for each tblinfl entry status Returned status of the function 351 Description Returns a UTBLINF table information structure for each valid table ID in the tidl list. The parameter infclas controls how much table information is returned to each UTBLINF structure. You can specify more than one option by separating the options with the bitwise logical operator (|). Information retrieval classes (specified by infclas) are ordered according to system overhead usage: UNOCLASS information uses the least amount of overhead, while UALLINFO uses the most. The UNOCLASS option specifies that only essential information defined by the UTBLINF structure is to be retrieved. No optional information is retrieved. The UNOCLASS option information is always retrieved, no matter which retrieval class you choose. The UCLASS1, UCLASS2, UCLASS3 and UCLASS4 categories, (which are defined in the structure UTBLINF), retrieve the specified information for that class only. The UALLINFO option specifies that all the information defined for the UTBLINF structure is to be retrieved. For a description of the UTBLINF structure, see “The RHLI Structures” chapter. For table successfully retrieved by specifying UCLASS2, UCLASS3, UCLASS4 or UALLINFO retrieval classes, storage space is allocated byUnify DataServer for UTBLINF structure members ucidlst, ubidlst, uhidlst, ulidlst, keycidl and uvidlst by calling the malloc function. This space should be released (by calling free function) when the information is no longer needed. Security Table information can be retrieved only for tables and access methods you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Volume information is only returned if you have DBA authority. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 352 Example The following example uses the uinftbl function to retrieve table information. #define FREEIT(nitem, sptr) if (nitem > 0) free(sptr); if (ualltbl(txid, UNULLAID, USLCK, &ntbl, tidl, &status) != USUCCESS) uexit(1); if ((tblinfl = (UTBLINF *)calloc(ntbl, sizeof(UTBLINF))) == NULL) uexit(1);} if (uinftbl(txid, ntbl, tidl, UALLINFO, tblinfl, statusl, &status)!=USUCCESS) { printf(”Unable to retrieve info on tables: status = %ld\n”,status); FREEIT(ntbl, tblinfl) FREEIT(ntbl, statusl) uexit(1); } /* Print contents of TBLINF structure and deallocate space */ for ( tblindx = 0; tblindx < ntbl; ++tblindx ) { /* Print table information */ prtblinf(&tblinfl[tblindx]); for ( colindx = 0; colindx < tblinfl[tblindx].ncol; ++colindx ) /* Print column information */ prcolin(tblinfl[tblindx].ucidlst[colindx]); FREEIT(colindx, tblinfl[tblindx].ucidlst) for ( volindx = 0; volindx < tblinfl[tblindx].nvol; ++volindx ) /* Print volume information */ prvolin(tblinfl[tblindx].uvidlst[volindx])); FREEIT(volindx, tblinfl[tblindx].uvidlst) FREEIT(tblinfl[tblindx].nbtree, tblinfl[tblindx].ubidlst) FREEIT(tblinfl[tblindx].nhsh, tblinfl[tblindx].uhidlst) FREEIT(tblinfl[tblindx].nlnk, tblinfl[tblindx].ulidlst) FREEIT(tblinfl[tblindx].nkey, tblinfl[tblindx].keycidl) FREEIT(ntbl, statusl) } See Also uaddtbl, ualltbl, ubndtbl, udrptbl, ufchtbl, uinfsch, uinstbl, ulcktbl, umaptbl, unumtbl, uulktbl malloc(3) and free(3) in your UNIX system manual 353 For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about managing database volumes, see “Managing the Database Files” in Unify DataServer: Managing a Database. 354 uinftx Retrieve transaction information Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uinftx( dbid, ntx, txl, infclas, txinfl, statusl, status ) UDBID dbid; int ntx; UTXID *txl; UOPTS infclas; UTXINF *txinfl; USTATUS statusl[]; USTATUS *status; Arguments Description dbid The database ID (issued by the uopndb function) ntx Number of transaction IDs in the txl list txl List of transaction IDs to retrieve information about infclas Information retrieval class. Options are: UNOCLASS Retrieve transaction ID and whether logging is on UCLASS1 Retrieve Class 1 information UALLINFO Retrieve all information txinfl Returned list of UTXINF structures, each containing information about a separate transaction (corresponding to the txl list) for the current database (specified by dbid) statusl Returned list of status values, one for each txinfl entry status Returned status of the function Returns a UTXINF transaction information structure for each valid transaction ID in the txl list. The infclas parameter controls how much transaction information is returned to each UTXINF structure. You can specify more than one option by separating the options with the bitwise logical OR operator (|). 355 The UNOCLASS information is always retrieved, no matter which retrieval class you choose. The UCLASS1 category, retrieves the specified information for that class only. See the UTXINF structure for a list of the UCLASS1 retrieved information. The information retrieval classes (specified by infclas) require different amounts of system overhead usage: UNOCLASS information uses the least amount of overhead, while UALLINFO uses the most. For a description of the UTXINF structure, see “The RHLI Structures” chapter. You must allocate the storage space required for the transaction information structures. When the space is no longer needed, you must free the space. Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also uabttx, ualltx, ubegtx, ucbgtx, ucmttx For information about transaction management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 356 uinfusr Retrieve user information Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uinfusr( txid, nuser, usridl, infclas, usrinfl, statusl, status ) UTXID txid; int nuser; UOID *usridl; UOPTS infclas; UUSRINF *usrinfl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nuser Number of user IDs in the usridl list usridl List of user IDs to retrieve information about or UNULLUID infclas Information retrieval class. Options are: UNOCLASS Retrieve essential information only UCLASS1 Retrieve Class 1 information only UCLASS2 Retrieve Class 2 information only UCLASS3 Retrieve Class 3 information only UALLINFO Retrieve all information usrinfl Returned list of UUSRINF structures, each containing information about a separate user (corresponding to the usridl list) statusl Returned list of status values, one for each usrinfl entry status Returned status of the function Returns a UUSRINF user information structure for each valid user ID. Returns curaid in usrinfl only for the current user, and UNULLAID for other users. For a description of the UUSRINF structure, see “The RHLI Structures” chapter. The parameter infclas controls how much user information is returned to each UUSRINF structure. 357 Information retrieval classes (specified by infclas) are ordered according to system overhead usage: UNOCLASS information uses the least amount of overhead, while UALLINFO uses the most. The UNOCLASS option specifies that only essential information defined by the UUSRINF structure is to be retrieved. No optional information is retrieved. The UNOCLASS information is always retrieved, no matter which retrieval class you choose. The UCLASS1, UCLASS2 and UCLASS3 categories (which are defined in the structure UUSRINF), retrieve the specified information for that class only. The UALLINFO option specifies that all the information defined for the UUSRINF structure is to be retrieved. You can combine retrieval classes by OR’ing the infclas values together, separated by single bars (|). If the usridl member points to UNULLUID, this function returns information about the current user. For columns successfully retrieved by specifying UCLASS2 or UALLINFO retrieval classes, storage space is allocated by Unify DataServer for UUSRINF structure member aidlst by calling malloc function. This space should be released (by calling the free function) when the information is no longer needed. Security User information is returned as follows: If you have DBA authority, information about any user is available. If you don’t have DBA authority, only information about your user ID is available. For information about granting DBA authority, see the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 358 Example This example calls the uinfusr function to retrieve user information. /* define required parameters */ UTXID txid; UOID * usridl; UUSRINF * usrinfl; USTATUS statusl; USTATUS status; /* Retrieve user ID (usridl) */ ... /* Fetch user information */ if (uinfusr(txid, 2, usridl, UALLINFO, usrinfl, statusl, &status) != USUCCESS) fprintf(stderr,”Unable to get info on user: status = %ld\n”, status); See Also uaddusr, uallusr, ubndusr, udrpusr, umodusr, unumusr For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 359 uinfvol Retrieve volume information #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uinfvol( txid, nvol, vidl, infclas, volinfl, statusl, status ) UTXID txid; int nvol; UVID *vidl; UOPTS infclas; UVOLINF *volinfl; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID nvol Number of volume IDs in the vidl list vidl List of volume IDs to retrieve information about infclas Information retrieval class. Options are: UNOCLASS Retrieve essential information only UCLASS1 Retrieve Class 1 information only UCLASS2 Retrieve Class 2 information only UCLASS3 Retrieve Class 3 information only UALLINFO Retrieve all information volinfl Returned list of UVOLINF structures, each containing information about a separate volume (corresponding to the vidl list) statusl Returned list of status values, one for each volinfl entry status Returned status of the function Returns a UVOLINF volume information structure for each valid volume ID. The parameter infclas controls how much volume information is returned to each UVOLINF structure. You can specify more than one option by separating the options with the bitwise logical operator (|). 360 Information retrieval classes (specified by infclas) are ordered according to system overhead usage: UNOCLASS information uses the least amount of overhead, while UALLINFO uses the most. The UNOCLASS option specifies that only essential information defined by the UVOLINF structure is to be retrieved. no optional information is retrieved. The UNOCLASS information is always retrieved, no matter which retrieval class you choose. The UCLASS1, UCLASS2 and UCLASS3 categories, which are defined in the structure UVOLINF, retrieve the specified information for that class only. The UALLINFO option specifies that all the information defined for the UVOLINF structure is to be retrieved. For a description of the UVOLINF structure, see “The RHLI Structures” chapter. For columns successfully retrieved by specifying UCLASS2 or UALLINFO retrieval classes, storage space is allocated by Unify DataServer for UVOLINF structure member utidlst by calling the malloc function. This space should be released (by calling the free function) when the information is no longer needed. The volume length number (vollen) of volinf will be in bytes, wherever the volume size is less than 2 GB. Otherwise, DB_BIG will be set in volopts, and vollen will represent the number of pages. Security You must have DBA authority to retrieve volume information. For information about granting DBA authority, see the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example The following example uses the uinfvol function to retrieve volume information. if (uinfvol(txid, nvol, vidl, UALLINFO, volinfl, statusl,&status) != USUCCESS) printf(“Unable to retrieve info on volumes: status = %ld\n”,status); 361 See Also uadddb, uaddvol, uallvol, ubndvol, udrpvol, umapvol, umodvol, unumvol malloc(3) and free(3) in your UNIX system manual For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about managing database volumes, see “Managing the Database Files” in Unify DataServer: Managing a Database. 362 uinimsg Initialize the message-logging facilities Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uinimsg( prgnm, status ) char *prgnm; USTATUS *status; Arguments Description prgnm The name of the current application’s executable, typically argv[0] status Returned status of the function Initializes the error-logging facilities of the database. Error logging is explicitly accomplished by calling the ulogmsg function. Error logging also happens when a serious error is encountered while performing a database operation. If you do not call the uinimsg function, messages written to the error log do not contain the executable name. You normally call the uinimsg function once to perform one-time-only initialization operations. Subsequent calls to the uinimsg function can change the current executable name. The uinimsg function should be called prior to opening a database. Security No privileges required. 363 Example This example uses the uinimsg function to initialize the error-logging file (errlog). Error messages are logged to the errlog file by using the ulogmsg function. /* define required RHLI parameters */ USTATUS status; USTATUS lstats; ... /* initialize the logging facilities */ if ( uinimsg(argv[0], &status) != USUCCESS)) { printf (”error initializing error logging with program name;”); if ((errmsg = ufchmsg(status, &fstatus)) == NULL) printf(”error fetching message\n”); else printf(”error is : %s\n”,errmsg); } /* open the database */ if ( uopndb ( dbname, DB_NORMAL, &dbid, &status) != USUCCESS ) { printf (”unable to open the database;”); if ((errmsg = ufchmsg(status, &fstatus)) == NULL) printf(”error fetching message\n”); else printf(”error is : %s\n”,errmsg); } /* log error messages ... */ See Appendix A for more examples. See Also ufchmsg, ulogmsg For information about message handling, see “Error Handling” on page 51. 364 uinsand Allocate a criteria structure for a scan Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uinsand( txid, scaninf, numexp, status ) UTXID txid; char *scaninf; int numexp; USTATUS *status; Arguments Description txid Transaction ID scaninf Pointer to the scan information structure numexp Number of expressions in conditions associated with the query that builds the scan status Returned status of the function Allocates the storage for a criteria structure. The criteria structure is determined by the uinsor function. You must call this function before you call the uinsor function. The numexp parameter specifies the number of expressions in the OR-groups associated with the scan. Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 365 Example See Appendix A. See Also uinsprj, ualcscn, ufrescn, uinssrt, uinsor 366 uinsbuf Allocate buffer Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uinsbuf( txid, scaninf, rmtbufsz, status ) UTXID txid; char * scaninf; long rmtbufsz; USTATUS * status; Arguments Description txid The current Transaction ID. scaninf Pointer to the scan information structure rmtbufsz Remote row buffer size in number of rows status Pointer to the value that indicates the status of the function’s operation The remote buffer size is the number of rows to buffer. If uinsbuf is not used to specify the remote row buffer size, the value of the shared configuration variable RMTROWBUFSZ is used to determine this size. Any required locking occurs when the rows are placed in the buffer. Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 367 Example The following example shows how to use uinsbuf. USTATUS status; /* USTATUS fchstat; /* UDBID dbid; /* UTXID txid; /* char * scaninf; /* UDBVAL tstval; UDBVAL * prjvall; USTATUS statusl[4]; UTP_HINT monumv; int scnid; URID rowid; int prevlck; status returned by RHLI functions */ status returned by ufchmsg( ) */ Database ID from uopndb( ) */ Transaction ID */ scan descriptor */ /* Allocating a Scan Structure */ if ( ualcscn(txid, 2, 0, 3, &scaninf, &status) == UFAILURE ) { printf(”ualcscn( ) failed;\n”); if ((errmsg = ufchmsg(status, &fchstat)) == NULL) printf(”error fetching message\n”); else printf(”error is %s\n”, errmsg); uexit(1); } /* Adding selection criteria */ if ( uinstbl(txid, scaninf, $model.$, &status) == UFAILURE ) { printf(”uinstbl( ) failed: \n”); if ((errmsg = ufchmsg(status, &fchstat)) == NULL) printf(”error fetching message\n”); else printf(”error is %s\n”, errmsg); uexit(1); } /* set remote scan buffer size */ if ( uinsbuf(txid, scaninf, 20L, &status) == UFAILURE ) { printf(”uinsbuf( ) failed: \n”); if ((errmsg = ufchmsg(status, &fchstat)) == NULL) printf(”error fetching message\n”); else printf(”error is %s\n”, errmsg); uexit(1); } if ( (uinsprj(txid,scaninf,$model.monum$,&status) == UFAILURE) || (uinsprj(txid,scaninf,$model.momano$,&status) == UFAILURE) || (uinsprj(txid,scaninf,$model.mdes$,&status) == UFAILURE) ) 368 { printf(”uinsprj( ) failed: \n”); if ((errmsg = ufchmsg(status, &fchstat)) == NULL) printf(”error fetching message\n”); else printf(”error is %s\n”, errmsg); uexit(1); } if ( uinsand(txid, scaninf, 1, &status) == UFAILURE ) { printf(”uinsand( ) failed: \n”); if ((errmsg = ufchmsg(status, &fchstat)) == NULL) printf(”error fetching message\n”); else printf(”error is %s\n”, errmsg); uexit(1); } tstval.unip.hintp = &monumv; tst.val.unip.valopt = UNORMAL monumv = 5; if ( uinsor(txid, scaninf, $model.monum$, UGTTST, &tstval, (UDBVAL *)0, &status) == UFAILURE ) { printf(”uinsor( ) failed: \n”); if ((errmsg = ufchmsg(status, &fchstat)) == NULL) printf(”error fetching message\n”); else printf(”error is %s\n”, errmsg); uexit(1); } if ( uinsand(txid, scaninf, 1, &status) == UFAILURE ) { printf(”uinsand( ) failed:\n”); if ((errmsg = ufchmsg(status, &fchstat)) == NULL) printf(”error fetching message\n”); else printf(”error is %s\n”, errmsg); uexit(1); } monumv = 17; if ( uinsor(txid, scaninf, $model.monum$, ULTTST, &tstval, (UDBVAL *)0, &status) == UFAILURE ) { printf(”uinsor( ) failed: \n”); if ((errmsg = ufchmsg(status, &fchstat)) == NULL) printf(”error fetching message\n”); else printf(”error is %s\n”, errmsg); uexit(1); } 369 if ( uinfscn(txid, scaninf, &prjvall, &status) == UFAILURE ) { printf(”uinfscn( ) failed: \n”); if ((errmsg = ufchmsg(status, &fchstat)) == NULL) printf(”error fetching message\n”); else printf(”error is %s\n”, errmsg); uexit(1); } if ( ubegscn(txid,USLCK,scaninf,&scnid,&status) == UFAILURE ) { printf(”ubegscn( ) failed: \n”); if ((errmsg = ufchmsg(status, &fchstat)) == NULL) printf(”error fetching message\n”); else printf(”error is %s\n”, errmsg); uexit(1); } printf(”Scanning for ’monum’ values between 5 and 17\n”); while ( ufchscn(scnid, &rowid, prjvall, &prevlck, statusl, &status) != UFAILURE ) { printf(”rowid = %ld\n”, rowid); printf(”monum = %ld\n”, *prjvall[0].unip.hintp); printf(”momano = %d\n”, *prjvall[1].unip.intp); } if ( uendscn(scnid, &status) == UFAILURE ) { printf(”uendscn( ) failed: \n”); if ((errmsg = ufchmsg(status, &fchstat)) == NULL) printf(”error fetching message\n”); else printf(”error is %s\n”, errmsg); uexit(1); } free((char free((char free((char free((char *)prjvall[0].unip.hintp); *)prjvall[1].unip.intp); *)prjvall[2].unip.strp); *)prjvall); /* Resetting the Scan Info. Structure */ if ( ufrescn(txid, scaninf, DB_RESET, &status) == UFAILURE ) { printf(”ufrescn( ) failed: \n”); if ((errmsg = ufchmsg(status, &fchstat)) == NULL) printf(”error fetching message\n”); else printf(”error is %s\n”, errmsg); uexit(1); } 370 See Also ualcscn, ufrescn, uinsand, uinsor, uinsprj, uinssrt, uinstbl, ualcscn. RMTROWBUFSZ configuration variable description in Unify DataServer: Configuration Variable and Utility Reference. 371 uinsor Insert an expression into an OR group of the scan selection criteria #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uinsor( txid, scaninf, cid, rop, val1, val2, status ) UTXID txid; char *scaninf; UCID cid; int rop; UDBVAL *val1; UDBVAL *val2; USTATUS *status; Arguments Description txid Transaction ID scaninf Pointer to the scan information structure cid Column ID of the column to be compared rop A value indicating the relational operator for the comparison val1 Pointer to a UDBVAL structure that contains the first value to compare against column values for the selection criteria val2 Pointer to a UDBVAL structure that contains the second value to compare against column values for the selection criteria status Returned status of the function Adds a unary, binary, or range expression to the selection criteria for a column. The values cid, rop, and optionally val1 and val2 make up a selection criterion. If you specify more than one selection criterion on the same column, the criteria are grouped to form a logical OR-group. 372 If more than one OR-group exists, the OR-groups are logically ANDed to comprise the selection criteria. To initialize the selection criteria for a scan, call the uinsand function. The target table must be established by calling the uinstbl function before calling the uinsor function. By using this function, you can represent most database queries. Queries that can be expressed on a single form of an ACCELL/SQL application can be performed using a single scan. Other queries require more than one scan. Performance If the number of expressions passed into the uinsand function (numexp) is greater than zero (which reflects the number of calls to the uinsor function), all the storage space is pre-allocated for maximum performance. To further enhance performance, specify the total number of OR-groups when calling the ualcscn function. The value of rop indicates a unary, binary or a range selection criterion. Unary operators are the following null value tests: UISNULL Null value UNOTNULL Not a null value UTTST Always true UFTST Always false Binary operators require only val1 to be passed in. Valid binary operators for rop are: UEQTST Equal UNETST Not equal ULTTST Less than ULETST Less than or equal UGTTST Greater than UGETST Greater than or equal ULIKE Like pattern matching %" matches 0 or more characters _" matches a single character \" escapes any character UREGEXP Regular expression pattern matching 373 UGLOB GLOB pattern matching (GLOB is a UNIX pattern matching operation) For more information about pattern matching, see the “Selecting Rows and Columns” chapter in Unify DataServer: Writing Interactive SQL/A Queries. Range operators require both val1 and val2 and specification of end point inclusion. Valid range operators for rop are: URNGOO Range: points(open,open) URNGOC Range: points(open,closed) URNGCO Range: points(closed,open) URNGCC Range: points(closed,closed) Open end points mean that the end point value is not included in the test range. Closed end points mean that the end point value is included in the test range. Parameters val1 and val2 are pointers to separate UDBVAL structures that contain comparison values. For a description of the UDBVAL structure, see “The RHLI Structures” chapter. Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See appendix A. See Also ualcscn, ufrescn, uinsand, uinsprj, uinssrt, uinstbl 374 uinsprj Add a column to the projection list of a scan Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uinsprj( txid, scaninf, ncol, cidl, statusl, status ) UTXID txid; char *scaninf; int ncol; UCID *cidl; USTATUS *statusl; USTATUS *status; Arguments Description txid Transaction ID scaninf Pointer to the scan information structure ncol Number of columns in the column ID list cidl Column ID list to add in the projection status Returned status of the function Adds a column to the projection list of columns from rows returned from a scan. The uinsprj function must be used for each column that is to be included in the projection. The order in which columns are added to the projection by calling the uinsprj function defines the order of the columns in the rows returned from the scan. Multiple calls to the same scan are allowed. The target table must be established by calling the uinstbl function before calling the uinsprj function. Security No privileges required. 375 Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also uinssrt, ufrescn, ualcscn, uinsor, uinsand 376 uinsrow Insert a row into a table Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uinsrow( txid, tid, rid, unipcoll, statusl, status ) UTXID txid; UTID tid; URID *rid; UNIPCOLL *unipcoll; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of the table rid Row ID of the newly created row unipcoll A pointer to a UNIPCOLL structure which specifies the values of the columns of the row to insert statusl Returned status value for each column listed in the UNIPCOLL column list cidl status Returned status of the function. Creates a new row and initializes the row’s column values with the values passed in from the UNIPCOLL structure. The UNIPCOLL structure (pointed to by unipcoll) describes the data being inserted into or retrieved from the database. This structure contains several lists which describe both the column and the column’s value. All of the columns specified in UNIPCOLL must belong to the table ID specified in tid. For a description of the UNIPCOLL structure, see “The RHLI Structures” chapter. 377 If no initial value is specified for a column and the column is not defined as a direct key, Unify DataServer checks the Data Integrity Subsystem (DIS) file for a default value. For more information about DIS, see Unify DataServer: Managing a Database. If the non-initialized column is defined as a direct key, Unify DataServer assigns the next consecutive value to the column. If no DIS value is defined for column (or the column is not a direct key), then the non-initialized column is set to a null value. If a primary key has been specified for the table, a value for the key column must be defined (in UNIPCOLL) or the new row is rejected. If any portion of the row insert is in error, the entire row is rejected and an error is generated. When inserting into or updating columns of type U_STR (string), Unify DataServer adds trailing blanks to values less than the total size defined for the column. Security A row can be inserted into a table that you have access to only. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted insert schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also udelrow For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about the data integrity subsystem (DIS), see “Initializing a Database” in Unify DataServer: Managing a Database. 378 uinssrt Insert a column to the sort criteria Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uinssrt( txid, scaninf, cid, ascflg, status ) UTXID txid; char *scaninf; UCID cid; int ascflg; USTATUS *status; Arguments txid Transaction ID scaninf Pointer to the scan information structure cid Column ID of the column to be a sort key ascflg Specifies the order of comparison for the key. Options are: status Description DB_ASCEND Ascending key value DB_DESCND Descending key value Returned status of the function Adds a column to the sort criteria, specifying the comparison order. The column ID specified (as cid) is added as a sort key of the order in which the ufchscn function returns a projection from a scan. The uinssrt function adds each column to the sort criteria in calling order. The first call to the uinssrt function defines the primary sort key, the second call defines the secondary sort key, and so on. This function must be used for each column that is to be included in the sort key. Before calling the uinssrt function, you must establish the target table by calling the uinstbl function. 379 Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also uinsprj, ualcscn, ufrescn, uinsor, uinsand For information about scanning, see page 64. 380 uinstbl Specify the target table for a scan. Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uinstbl( txid, scaninf, tid, status ) UTXID txid; char *scaninf; UTID tid; USTATUS *status; Arguments Description txid Transaction ID scaninf Pointer to the scan information structure tid Table ID of the target table for the scan status Returned status of the function Establishes the target table for a scan. If a target table has already been set, an error occurs. The uinstbl function must be called before any calls to the uinsor, uinsprj, uinssrt, or ubegscn functions. Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also ualcscn, ufrescn, uinsand, uinsor, uinsprj, uinssrt For information about scanning, see page 64. 381 uintext Convert column values from internal format to external format #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uintext( txid, tid, unipcoll, unipext, statusl, status ) UTXID txid; UTID tid; UNIPCOLL *unipcoll; UNIPEXT *unipext; USTATUS statusl[]; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of the table columns contained in UNIPCOLL unipcoll A pointer to a UNIPCOLL structure containing column values in internal format to be converted into external format unipext A pointer to a UNIPEXT structure that is to contain the converted column values from internal to external format. This list must contain the same number of entries as the number of internal format entries in UNIPCOLL (pointed to by unipcoll). statusl Returned status value for each column listed in the UNIPCOLL column list cidl status Returned status of the function Converts a list of data values from an internal format to the external format used by the database. To read and display data contained in columns of data type U_DATE, U_HDATE and U_TIME it is necessary to first convert the data from internal (database) format to external (display) format. 382 To update or insert date and time data into a Unify DataServer database, you must first convert the data into internal format by calling the uextint function. For each column in the internal format list described by the structure UNIPCOLL (pointed to by unipcoll), each internal column data value is converted to external format and stored in a UNIPEXT structure (pointed to by unipext). For a description of the UNIPEXT and UNIPCOLL structures, see “The RHLI Structures” chapter. All columns specified in each UNIPCOLL structure must belong to the specified table ID (tid). Any valopts member contained in the UNIPCOLL structure that contains a value of UIGNORE is bypassed; no value is converted for that entry. Also, the entry for statusl is set to UENOWRK. All column values must be converted to external format after being retrieved from the database. See the ufchrow function for more information. If the destination column structure value pointer is null, storage space is allocated by Unify DataServer. This space must then be deallocated by calling the system subroutine free function. If the value pointer is not null, it is assumed that enough space exists for the converted value. A value that cannot be converted is rejected. Check each status value in statusl to determine which values could not be converted. Security To convert column values, you must have access to the column value that you want to convert. You can always access all columns in the tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 383 Example See Appendix A. See Also uextint, uinsrow, ufchrow, uupdrow free(3) in your UNIX system manual For information about converting from internal to external formats, see page 35. 384 ulckdb Lock a database Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ulckdb( txid, lcktype, nretry, prevlck, status ) UTXID txid; int lcktype; long nretry; int *prevlck; USTATUS *status; Arguments Description txid Transaction ID lcktype The lock type to place on the database. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock nretry The number of times to attempt acquiring the lock before failing; if set to 0L, then the value set by the LMNRETRY configuration variable is used prevlck Pointer to the previous lock type. If the database lock was not granted, the lock type returned is the conflicting lock value. status Returned status of the function Locks the database associated with the transaction. No changes can be made to the database by other transactions while the lock is held. A database lock locks all tables and rows for that database. If the database lock is granted, the previous lock type is returned into prevlck; otherwise, the conflicting lock value is returned. 385 You cannot acquire an exclusive lock on the database unless all users have opened the database (by calling the uopndb function) with the DB_NOLOCK option. You may not be able to acquire an exclusive lock if data definition locks placed by Unify DataServer conflict with database lock requests. Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the ulckdb function to acquire a shared lock on the database associated with the dbid value. /* define required parameters */ UTXID txid; UDBID dbid; int lcktype; long nretry; int prevlck; USTATUS status; ... /* open the database and begin a transaction */ ... /* lock the database */ lcktype = USLCK; nretry = 10; if ( ulckdb(txid,lcktype, nretry, &prevlck, &status) != USUCCESS ) { printf(”Unable to get database lock: status = %ld\n”,status); uexit(99); } See Also ulcksch, uulksch, ulckrow, uulkrow, ulcktbl, uulktbl, uulkdb For more information about setting configuration variables, see Unify DataServer: Configuration Variable and Utility Reference. For information about lock management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 386 ulckrow Lock a row Syntax #include <include/rhli.h> #include <include/rhlierr.h> int ulckrow(txid, tid, rid, lcktype, nretry, prevlck, status) UTXID txid; UTID tid; URID rid; int lcktype; long nretry; int *prevlck; USTATUS *status; Arguments txid Transaction ID tid Table of the row to lock rid Row ID of the row to lock lcktype The lock type to place on the row. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock nretry The number of times to attempt acquiring the lock before failing; if set to 0L, then the value set by the LMNRETRY configuration variable is used prevlck Pointer to the previous lock type. If the lock was not granted, the lock type returned is the conflicting lock value. status Returned status of the function. 387 Description Locks the row associated with the row ID for the specified table. No changes can be made to the row data by other transactions while the lock is held. If the row lock is granted, the previous lock type is returned into prevlck; otherwise, the conflicting lock value is returned. Security Rows can be locked only for tables you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the ulckrow function to acquire a shared lock on the row in the company table with a rowid of 3. /* define required parameters */ UTXID txid; URID rid; int lcktype; long nretry; int prevlck; USTATUS status; ... /* open the database and begin a transaction */ ... /* lock the row */ lcktype = USLCK; nretry = 10; rid = 3; if ( ulckrow(txid, $company.$, rid, lcktype, nretry, &prevlck,&status) != USUCCESS ) { printf(”Unable to get a row lock: status = %ld\n”,status); uexit(99); } ... 388 See Also uulkrow, ulcktbl, uulktbl, ulckdb, uulkdb For more information about setting configuration variables, see Unify DataServer: Configuration Variable and Utility Reference. For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about lock management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 389 ulcksch Lock a table definition #include <include/rhli.h> #include <include/rhlierr.h> Syntax bool ulcksch( txid, tid, lcktype, nretry, prevlck, status ) UTXID txid; UTID tid; int lcktype; long nretry; int *prevlck; USTATUS *status; Arguments Description txid Current transaction ID tid Table ID of the table to lock lcktype The lock type to place on the table. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock nretry The number of times to attempt acquiring the lock before failing; if set to 0L, then the value set by the LMNRETRY configuration variable is used prevlck Pointer to the previous lock type. If the lock was not granted, the lock type returned is the conflicting lock value. status Returned status of the function Locks the definition for a table. A definition lock prevents any other transaction from changing or deleting the definition for a table while table data is being accessed. A table definition lock is separate from and does not interfere with data locks. To acquire a data lock see the ulcktbl or the ulckrow function. 390 If the table definition lock is granted, the previous lock type is returned into prevlck; otherwise, the conflicting lock value is returned. To remove a table definition lock, call the uulksch function. Security Definition locks can only be acquired on tables you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the ulcksch function to acquire an object definition lock on the company table. /* define required parameters */ UTXID txid; int lcktype; int prevlck; USTATUS status; ... /* lock the table definition */ if (ulcksch(txid, $company.$, USLCK, 0L, &prevlck, &status) != USUCCESS) { printf(”SLOCK of company table failed—%ld\n”,status); uexit(99); } See Also uulksch For more information about setting configuration variables, see the Unify DataServer: Configuration Variable and Utility Reference. For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about lock management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 391 ulcktbl Lock a table #include <include/rhli.h> #include <include/rhlierr.h> Syntax int ulcktbl( txid, tid, lcktype, nretry, prevlck, status ) UTXID txid; UTID tid; int lcktype; long nretry; int *prevlck; USTATUS *status; Arguments Description txid Transaction ID tid Table ID of the table to lock lcktype The lock type to place on the table. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock nretry The number of times to attempt acquiring the lock before failing; if set to 0L, then the value set by the LMNRETRY configuration variable is used prevlck Pointer to the previous lock type. If the lock was not granted, the lock type returned is the conflicting lock value. status Returned status of the function Locks all rows in the table. No changes can be made to the table data by other transactions while the lock is held. If the table lock is granted, the previous lock type is returned into prevlck; otherwise, the conflicting lock value is returned. 392 Security Only tables you have access to can be data locked. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the ulcktbl function to acquire a lock on the company table. /* define RHLI interface required parameters */ UTXID txid; UDBID dbid; int lcktype; long nretry; int prevlck; USTATUS status; ... /* open the database and begin a transaction */ ... /* lock the table */ lcktype = USLCK; nretry = 10; if ( ulcktbl(txid, $parts.$, lcktype, nretry, &prevlck, &status) != USUCCESS ) { printf(”Unable to get table lock: status = %ld\n”, status); uexit(100); } See Also uulkrow, ulckrow, uulktbl, ulckdb, uulkdb For more information about setting configuration variables, see the Unify DataServer: Configuration Variable and Utility Reference. For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about lock management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 393 ulogmsg Append a message to the error log #include <include/rhli.h> #include <include/rhlierr.h> Syntax int ulogmsg( callfnm, errnum, xtrainfo, status ) char *callfnm; USTATUS errnum; char *xtrainfo; USTATUS *status; Arguments Description txid Transaction ID callfnm Name of the function from which the error logging is performed errnum The status error code xtrainfo Pointer to a character string containing any other relevant information, or a null pointer if there is no other information status Returned status of this function Appends the status error code (errnum), a corresponding error message, and related information to the error log file DBNAME.err. The error log file is searched for in the directory specified by the DBPATH configuration variable. If the file is not found, error messages are written to the file errlog. The error file DBNAME.err is usually linked to the file errlog. If errlog cannot be found or cannot be opened or written to, this function fails. Use the uinimsg function to initialize the error logging facilities. Security No privileges required. 394 Example This example uses the ulogmsg function to log error messages to the errlog file. The errlog file is initialized by using the uinimsg function. See function uinimsg for the code. See Appendix A for more examples. See Also ufchmsg, uinimsg For information about error handling, see page 51. 395 umapath Map a compile-time authorization ID to its runtime equivalent #include <include/rhli.h> #include <include/rhlierr.h> Syntax int umapath( dbid, nath, athmapl, statusl, status ) UDBID dbid; int nath; UATHMAP athmapl[]; USTATUS statusl[]; USTATUS *status; Arguments Description dbid Database ID issued by the uopndb function nath Number of authorization IDs in the athmapl list athmapl List of UATHMAP structures, each containing information about a separate schema statusl Returned list of status values, one for each athmapl entry status Returned status of the function Maps the compile-time authorization IDs stored in each UATHMAP structure to runtime authorization IDs. Each UATHMAP structure must be initialized with schema information. To initialize the UATHMAP structure, use the mapping macro command INIATHMAP. For a description of the UATHMAP structure, see “The RHLI Structures” chapter. The mapping operation fails in these cases: The schema specified in the athnm member of the UATHMAP structure does not exist in the database. The schema’s runtime mapping ID specified in the member remapid does not match the compile-time mapping ID. Multiple values are mapped to the same runtime authorization ID. 396 To map the value of an authorization ID that is already mapped, you must reset the athnm member to (char *)0 or the null-string (“”). You cannot reset the ID of a schema which has not been mapped. Cross-mapping of authorization IDs is allowed. The runtime authorization ID to which a compile-time authorization ID has been mapped can also be obtained by calling the uinfath function. If the database was opened with the DB_REMAP or DB_PARTMAP options, compileĆtime IDs are automatically mapped to their runtime equivalents for all database objects. Therefore, you do not need to map the object IDs by calling the mapping functions. If you do call a mapping function for an already mapped object, the UEISMAP error is returned. You can ignore this error. Security Authorization IDs can be re-mapped only for schemas you have access to. You always have access to the current schema. Additionally, you can access other schemas if you have been granted schema access permission to the schema or if a schema you can access has schema privileges to another schema (or a table or column in the schema). If you have DBA authority, you can access any schema in the database. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. To grant schema access, use the uaddprm function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also uaddath, uallath, ubndath, udrpath, ufchath, uinfath, umodath, unumath For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about ID mapping, see page 42. 397 umapbt Map a compile-time B-tree ID to its runtime equivalent #include <include/rhli.h> #include <include/rhlierr.h> Syntax int umapbt( dbid, nbt, btmapl, statusl, status ) UDBID dbid; int nbt; UBTMAP btmapl[]; USTATUS statusl[]; USTATUS *status; Arguments Description dbid Database ID issued by calling the uopndb function nbt Number of B-tree IDs in the btmapl list btmapl List of UBTMAP structures, each containing information about a separate B-tree statusl Returned list of status values, one for each btmapl entry status Returned status of the function. Maps the compile-time B-tree IDs stored in each UBTMAP structure to runtime B-tree IDs. Each UBTMAP structure must be initialized with B-tree information. To initialize UBTMAP, use the mapping macro command INIBTMAP. For a description of the UBTMAP structure, see “The RHLI Structures” chapter. The mapping operation fails in these cases: The B-tree specified in the UBTMAP member btnm does not exist in the database. The B-tree’s runtime mapping ID specified in the member remapid does not match the compile-time mapping ID. Multiple values are mapped to the same runtime B-tree ID. 398 You cannot map the value of a B-tree that is already mapped, or reset the ID of a B-tree which has not been mapped. To reset the value of a previously mapped B-tree ID, set the value of UBTMAP member btnm to (char *)0 or the null-string (“”). Cross-mapping of B-tree IDs is allowed. The runtime B-tree ID to which a compile-time B-tree ID has been mapped can also be obtained by calling the uinfbt function. If the database was opened with the DB_REMAP or DB_PARTMAP options, compileĆtime IDs are automatically mapped to their runtime equivalents for all database objects. Therefore, you do not need to map the object IDs by calling the mapping functions. If you do call a mapping function for an already mapped object, the UEISMAP error is returned. You can ignore this error. Security B-tree IDs can be re-mapped only for B-trees on tables that you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also uaddbt, uallbt, ubndbt, udrpbt, uinfbt, umodbt For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about ID mapping, see page 42. 399 umapcol Map a compile-time column ID to its runtime equivalent #include <include/rhli.h> #include <include/rhlierr.h> Syntax int umapcol( dbid, ncol, colmapl, statusl, status ) UDBID dbid; int ncol; UCOLMAP colmapl[]; USTATUS statusl[]; USTATUS *status; Arguments Description dbid Database ID issued by the uopndb function ncol Number of column IDs in the colmapl list colmapl List of UCOLMAP structures, each containing information about a separate column statusl Returned list of status values, one for each colmapl entry status Returned status of the function Maps the compile-time column IDs stored in each UCOLMAP structure to runtime column IDs. Each UCOLMAP structure must be initialized with column information that is used to re-map a compile-time column ID to its runtime equivalent. To initialize UCOLMAP, use the mapping macro command INICOLMAP. For a description of the UCOLMAP structure, see “The RHLI Structures” chapter. The mapping operation fails in these cases: The column specified in the UCOLMAP member colnm does not exist in the database. The column’s runtime mapping ID specified in the member remapid does not match the compile-time mapping ID. Multiple values are mapped to the same runtime authorization ID. 400 You cannot map the value of a column that is already mapped, or reset the ID of a column which has not been mapped. To reset the value of a previously mapped column ID, set the value of UCOLMAP member colnm to (char *)0 or the null-string (“”) . Cross-mapping of column IDs is allowed. The runtime column ID to which a compile-time column ID has been mapped can also be obtained by calling the uinfcol function. If the database was opened with the DB_REMAP or DB_PARTMAP options, compileĆtime IDs are automatically mapped to their runtime equivalents for all database objects. Therefore, you do not need to map the object IDs by calling the mapping functions. If you do call a mapping function for an already mapped object, the UEISMAP error is returned. You can ignore this error. Security Column IDs can be re-mapped only for columns in tables that you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also uaddcnm, uallcol, ubndcol, udrpcnm, uextint, ufchcnm, uinfcol, uintext, umapcol, unumcol For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about ID mapping, see page 42. 401 umaphsh Map a compile-time hash table ID to its runtime equivalent #include <include/rhli.h> #include <include/rhlierr.h> Syntax int umaphsh( dbid, nhsh, hshmapl, statusl, status ) UDBID dbid; int nhsh; UHSHMAP hshmapl[]; USTATUS statusl[]; USTATUS *status; Arguments Description dbid Database ID issued by the uopndb function nhsh Number of hash table IDs in the hshmapl list hshmapl List of UHSHMAP structures, each containing information about a separate hash table statusl Returned list of status values, one for each hshmapl entry status Returned status of the function Maps the compile-time hash table IDs stored in each UHSHMAP structure to runtime hash table IDs. Each UHSHMAP structure must be initialized with hash table information that is used to re-map a compile-time hash table ID to its runtime equivalent. To initialize UHSHMAP, use the mapping macro command INIHSHMAP. For a description of the UHSHMAP structure, see “The RHLI Structures” chapter. The mapping operation fails in these cases: The hash table specified in the UHSHMAP member hshnm does not exist in the database. The hash table’s runtime mapping ID specified in the member remapid does not match the compile-time mapping ID. Multiple values are mapped to the same runtime hash table ID. 402 You cannot map the value of a hash table that is already mapped, or reset the ID of a hash table which has not been mapped. To reset the value of a previously mapped hash table IDset the value of UHSHMAP member hshnm to (char *)0 or the null-string (””). Cross-mapping of hash table IDs is allowed. The runtime hash table ID to which a compile-time hash table ID has been mapped can also be obtained by calling the uinfhsh function. If the database was opened with the DB_REMAP or DB_PARTMAP options, compileĆtime IDs are automatically mapped to their runtime equivalents for all database objects. Therefore, you do not need to map the object IDs by calling the mapping functions. If you do call a mapping function for an already mapped object, the UEISMAP error is returned. You can ignore this error. Security Hash table IDs can be re-mapped only for hash tables you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also uaddhsh, uallhsh, undhsh, udrphsh, uinfhsh, umodhsh For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about ID mapping, see page 42. 403 umaplnk Map a compile-time link ID to its runtime equivalent #include <include/rhli.h> #include <include/rhlierr.h> Syntax int umaplnk( dbid, nlnk, lnkmapl, statusl, status ) UDBID dbid; int nlnk; ULNKMAP lnkmapl[]; USTATUS statusl[]; USTATUS *status; Arguments Description dbid Database ID issued by the uopndb function nlnk Number of Link IDs in the lnkmapl list lnkmapl List of ULNKMAP structures, each containing information about a separate link statusl Returned list of status values, one for each lnkmapl entry status Returned status of the function Maps the compile-time Link IDs stored in each ULNKMAP structure to runtime Link IDs. Each ULNKMAP structure must be initialized with link information. To initialize ULNKMAP, use the mapping macro command INILNKMAP. For a description of the ULNKMAP structure, see “The RHLI Structures” chapter. The mapping operation fails in these cases: The link specified in the ULNKMAP member lnknm does not exist in the database. The link’s runtime mapping ID specified in the member remapid does not match the compile-time mapping ID. Multiple values are mapped to the same runtime link ID. 404 You cannot map the value of a link that is already mapped, or reset the ID of a link which has not been mapped. To reset the value of a previously mapped link ID, set the value of ULNKMAP member lnknm to (char *)0 or the null-string (“”) . Cross-mapping of Link IDs is allowed. The runtime Link ID to which a compile-time Link ID has been mapped can also be obtained by calling the uinflnk function. If the database was opened with the DB_REMAP or DB_PARTMAP options, compileĆtime IDs are automatically mapped to their runtime equivalents for all database objects. Therefore, you do not need to map the object IDs by calling the mapping functions. If you do call a mapping function for an already mapped object, the UEISMAP error is returned. You can ignore this error. Security Link IDs can be re-mapped only for links in tables that you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also uaddlnk, ualllnk, ubndlnk, udrplnk, ufchlnk, uinflnk, umodlnk For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about ID mapping, see page 42. 405 umaptbl Map a compile-time table ID to its runtime equivalent #include <include/rhli.h> #include <include/rhlierr.h> Syntax int umaptbl( dbid, ntbl, tblmapl, statusl, status ) UDBID dbid; int ntbl; UTBLMAP tblmapl[]; USTATUS statusl[]; USTATUS *status; Arguments Description dbid Database ID issued by the uopndb function ntbl Number of table IDs in the tblmapl list tblmapl List of UTBLMAP structures, each containing information about a separate table statusl Returned list of status values, one for each tblmapl entry status Returned status of the function Maps the compile-time table IDs stored in each UTBLMAP structure to runtime table IDs. Each UTBLMAP structure must be initialized with table information. To initialize UTBLMAP, use the mapping macro command INITBLMAP. For a description of the UTBLMAP structure, see “The RHLI Structures” chapter. The mapping operation fails in these cases: The table specified in the UTBLMAP member tblnm does not exist in the database. The table’s runtime mapping ID specified in the member remapid does not match the compile-time mapping ID. Multiple values are mapped to the same runtime table ID. 406 You cannot map the value of a table that is already mapped, or reset the ID of a table which has not been mapped. To reset the value of a previously mapped table ID, set the value of UTBLMAP member tblnm to (char *)0 or the null-string (“”) . Cross-mapping of table IDs is allowed. The runtime table ID to which a compile-time table ID has been mapped can also be obtained by calling the uinftbl function. If the database was opened with the DB_REMAP or DB_PARTMAP options, compileĆtime IDs are automatically mapped to their runtime equivalents for all database objects. Therefore, you do not need to map the object IDs by calling the mapping functions. If you do call a mapping function for an already mapped object, the UEISMAP error is returned. You can ignore this error. Security Table IDs can be re-mapped only for tables you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also uaddtbl, uaddtnm, ualltbl, ubndtbl, udrptbl, udrptnm, ufchtnm, uinftbl, ulcksch, ulcktbl, unumtbl, uulksch, uulktbl For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about ID mapping, see page 42. 407 umapvol Map a compile-time volume ID to its runtime equivalent #include <include/rhli.h> #include <include/rhlierr.h> Syntax int umapvol( dbid, nvol, volmapl, statusl, status ) UDBID dbid; int nvol; UVOLMAP volmapl[]; USTATUS statusl[]; USTATUS *status; Arguments Description dbid Database ID issued by uopndb function ntbl Number of volume IDs in the volmapl list volmapl List of UVOLMAP structures, each containing information about a separate table statusl Returned list of status values, one for each volmapl entry status Returned status of the function Maps the compile-time volume IDs stored in each UVOLMAP structure to runtime volume IDs. Each UVOLMAP structure must be initialized with volume information. To initialize UVOLMAP, use the mapping macro command INIVOLMAP. For a description of the UVOLMAP structure, see “The RHLI Structures” chapter. The mapping operation fails in these cases: The volume specified in the UVOLMAP member volnm does not exist in the database. The volume’s runtime mapping ID specified in the member remapid does not match the compile-time mapping ID. Multiple values are mapped to the same runtime volume ID. 408 You cannot map the value of a volume that is already mapped, or reset the ID of a volume which has not been mapped. To reset the value of a previously mapped volume ID, set the value of UVOLMAP member volnm to (char *)0 or the null-string (“”) . Cross-mapping of volume IDs is allowed. The runtime volume ID to which a compile-time volume ID has been mapped can also be obtained by calling the uinfvol function. If the database was opened with the DB_REMAP or DB_PARTMAP options, compileĆtime IDs are automatically mapped to their runtime equivalents for all database objects. Therefore, you do not need to map the object IDs by calling the mapping functions. If you do call a mapping function for an already mapped object, the UEISMAP error is returned. You can ignore this error. Security You must have DBA authority to re-map a volume ID. For information about granting DBA authority, see uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also uaddprf, uaddvol, uallvol, ubndvol, udrpprf, udrpvol, uinfvol, umodvol, unumvol For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about ID mapping, see page 42. 409 umodath Modify a schema name #include <include/rhli.h> #include <include/rhlierr.h> Syntax int umodath( txid, nauth, dbauthl, statusl, status ) UTXID txid; int nauth; DBAUTH *dbauthl; USTATUS statusl[]; USTATUS *status; Arguments Description txid The transaction ID nauth Number of schemas in the dbauthl list dbauthl List of DBAUTH structures, each describing a separate schema to modify statusl Returned list of status values, one for each dbauthl entry status Returned status of the function Modifies the schema name for one or more schemas listed in the dbauthl list. Each DBAUTH structure in the dbauthl list contains information about a schema. For a description of the DBAUTH structure, see “The RHLI Structures” chapter. Schema names specified in the dbauthl list must be unique within the database. Security To modify a schema name, you must have one of the following permissions. DBA authority: the name for any schema in the database can be changed SCHEMA authority and schema access permission: the name for the current schema only can be changed 410 For information about granting DBA and SCHEMA authority, see the uaddprv function; for information about granting schema access permission, see the uaddprm function. If the database was opened with the DB_REMAP or DB_PARTMAP options, compileĆtime IDs are automatically mapped to their runtime equivalents for all database objects. Therefore, you do not need to map the object IDs by calling the mapping functions. If you do call a mapping function for an already mapped object, the UEISMAP error is returned. You can ignore this error. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the umodath function to change a schema name. if ( ubndath(txid, 1, &argv[1], USLCK, &aid, statusl, &status) != UFAILURE ) { dbauthl.authnm = argv[2]; /* Change schema name */ dbauthl.aid = aid; /* Authorization ID */ /* Modify the schema */ if ( umodath(txid, 1, &dbauthl, &statusl, &status) != USUCCESS) { fprintf(stderr, ”Unable to modify schema: %ld\n”,status); fprintf(stderr,”Schema name %s Auth. ID. %ld status: %ld\n”, dbauthl.authnm, dbauthl.aid, statusl); uexit(1); } } else fprintf(stderr, ”Schema %s not in database\n”, argv[1]); See Also uaddath, uallath, ubndath, udrpath, ufchath, uinfath, umapath, unumath For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about ID mapping, see page 42. 411 umodbt Modify a B-tree name #include <include/rhli.h> #include <include/rhlierr.h> Syntax int umodbt( txid, nbt, dbbtl, statusl, status ) UTXID txid; int nbt; DBBTREE *dbbtl; USTATUS statusl[]; USTATUS *status; Arguments Description txid The transaction ID nbt Number of B-trees in the dbbtl list dbbtl List of DBBTREE structures, each describing a separate B-tree to modify statusl Returned list of status values, one for each dbbtl entry status Returned status of the function Modifies the B-tree name for one or more B-trees listed in the dbbtl list. Each DBBTREE structure in the dbbtl list must be specified with the new B-tree name. Because the umodbt function can change only a B-tree name, initialize the DBBTREE structure members btname and btid only. For a description of the DBBTREE structure, see “The RHLI Structures” chapter. B-tree names specified in the dbbtl list must be unique within a table. Security To modify the name for a B-tree, the B-tree must be in the current schema and you must have one of the following permissions. DBA authority SCHEMA authority 412 To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. If the database was opened with the DB_REMAP or DB_PARTMAP options, compileĆtime IDs are automatically mapped to their runtime equivalents for all database objects. Therefore, you do not need to map the object IDs by calling the mapping functions. If you do call a mapping function for an already mapped object, the UEISMAP error is returned. You can ignore this error. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the umodbt function to change the name of a B-tree index. if ((dbbtl = (DBBTREE *)calloc(1, sizeof(DBBTREE))) == NULL) uexit(1); dbbtl.btid = $parts.*part_num_btree$; dbbtl.btname = ”PARTS”; if ((dbbtl.statusl = (USTATUS *)calloc(1, sizeof(USTATUS))) == NULL) { free (dbbtl); uexit(1); } if ( umodbt(txid, 1, dbbtl, statusl, &status) != USUCCESS ) { fprintf(stderr,”Failure modifying B-tree name: status = %ld\n”,status); free (dbbtl); uexit(1); } See Also uaddbt, ubndbt, udrpbt For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about ID mapping, see page 42. 413 umoddb Modify a database #include <include/rhli.h> #include <include/rhlierr.h> Syntax int umoddb( txid, dbinfo, status ) UTXID txid; DBINFO *dbinfo; USTATUS *status; Arguments Description txid The transaction ID dbinfo Database information descriptor status Returned status of the function. Modifies the following attributes of a database: file access mode (dbmode) file owner ID (dbusrid) file group ID (dbgrpid) database description (descprt) database name (dbname) Database parameters are modified by initializing a DBINFO structure (pointed to by dbinfo). For a description of the DBINFO structure, see “The RHLI Structures” chapter. The variable name following each parameter in the list above is the corresponding DBINFO structure member to initialize. Security You must have DBA authority to modify a database. For information about granting DBA authority, see uaddprv function. 414 Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the umoddb function to modify database attributes. dbinfo.dbmode dbinfo.dbusrid dbinfo.dbgrpid dbinfo.descrpt database”; dbinfo.dbname = = = = 0666; 0; /* owned by root (see /etc/passwd) */ 0; /* group root (see /etc/group) */ ”Inventory, capacity, & production planning = ”parts_data_base”; if ( umoddb(txid, &dbinfo, &status) != USUCCESS ) fprintf(stderr,”Failure modifying database: status = %ld\n”,status); See Also uadddb, udrpdb chmod(2) and chown(2) in your UNIX system manual For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 415 umodhsh Modify a hash table name #include <include/rhli.h> #include <include/rhlierr.h> Syntax int umodhsh( txid, nmods, dbhashl, statusl, status ) UTXID txid; int nmods; DBHASH *dbhashl; USTATUS statusl[]; USTATUS *status; Arguments Description txid The transaction ID nmods Number hash tables in the dbhashl list dbhashl List of DBHASH structures, each describing a separate hash table to modify statusl Returned list of status values, one for each dbhashl entry status Returned status of the function Modifies the hash table name for one or more hash tables listed in the dbhashl list. Each DBHASH structure in the dbhashl list must be specified with the new hash table name. Hash table names specified in the dbhashl list must be unique within a table. Because the umodhsh function can change only a hash table name, initialize the DBHASH structure members htname and hid only. For a description of the DBHASH structure, see “The RHLI Structures” chapter. Security To modify the name for a hash table, the hash table must be in the current schema and you must have one of the following permissions. DBA authority SCHEMA authority 416 To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the umodhsh function to change the hash table name. dbhash.hid dbhash.htname = $manf.#mano_ht$; = ”mano_hshtbl”; if ( umodhsh(txid, 1, &dbhash, statusl, &status) != USUCCESS ) { fprintf(stderr,”Failure modifying a hash index: status = %ld\n”,status); fprintf(stderr, ”index status = %ld\n”, *statusl); } See Also uaddhsh, ubndhsh, udrphsh For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 417 umodlnk Modify a link name #include <include/rhli.h> #include <include/rhlierr.h> Syntax int umodlnk( txid, nlnks, dblinkl, statusl, status ) UTXID txid; int nlnks; DBLINK *dblinkl; USTATUS statusl[]; USTATUS *status; Arguments Description txid The transaction ID nlnks Number of links in the dblinkl list dblinkl List of DBLINK structures, each describing a separate link to modify statusl Returned list of status values, one for each dblinkl entry status Returned status of the function. Modifies the link name for one or more links listed in the dblinkl list. Each DBLINK structure in the dblinkl list must be specified with the new link name. Because the umodlnk function can change a link name only, initialize only the DBLINK structure members lnkname and lid. For a description of the DBLINK structure, see “The RHLI Structures” chapter. The new link name must be unique within both the parent and child tables. Security To modify a link, the link must be in the current schema and you must have one of the following permissions. DBA authority SCHEMA authority 418 Also, the two tables comprising the link definition must be made accessible by one of these methods: Make the schema that contains both tables current. If the tables are located in two separate schemas, make the schema that contains the first table current and have schema access to the schema that contains the second table. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. To grant schema access, use the uaddprm function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the umodlnk function to modify the link index name. dblink.lid = $parts.@part_num_link$; dblink.lnkname = ”part_number_link”; if ( umodlnk(txid, 1, &dblink, statusl, &status) != USUCCESS ) { fprintf(stderr, ”Error modifying link: status = %ld\n”, status); fprintf(stderr, ”link status = %ld\n”, *statusl); } See Also uaddlnk, ubndlnk, udrplnk For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 419 umodusr Modify a user name and default schema #include <include/rhli.h> #include <include/rhlierr.h> Syntax int umodusr( txid, nuser, dbuserl, statusl, status ) UTXID txid; int nuser; DBUSER *dbuserl; USTATUS statusl[]; USTATUS *status; Arguments Description txid The transaction ID nuser Number of users in the dbuserl list dbuserl List of DBUSER structures, each describing a separate user’s information to modify statusl Returned list of status values, one for each dbuserl entry status Returned status of the function Modifies the name and default schema for each user listed in the dbuserl list. Each DBUSER structure in the dbuserl list must be specified with the new user or default schema name information; one of these items may be omitted by specifying a null value. For a description of the DBUSER structure, see “The RHLI Structures” chapter. Security A user’s name and default schema can be modified only under the following conditions: If you have DBA authority, any user name or default schema can be modified. If you don’t have DBA authority, only your user name can be modified; the default schema can be changed only if you have schema access permission to the new schema. To grant DBA authority, use the uaddprv function. To grant schema access permission, use the uaddprm function. 420 Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example calls the umodusr function to change a user name. /* define required parameters */ UTXID txid; DBUSER * dbuserl; USTATUS statusl; USTATUS status; / * Retrieve user ID (uoid) */ ... /* Initialize DBUSER structure */ dbuserl.uoid = uoid; /* User ID to change */ dbuserl.usernm = argv[2]; /* New user name */ dbuserl.aid = UNULLAID; /* Keep same default schema */ /* Modify the user information */ if (umodusr(txid, 1, &dbuserl, &statusl, &status) != USUCCESS){ fprintf(stderr, ”Unable to modify user: %ld\n”, status); fprintf(stderr, ”Users %s auth. id. %ld status: %ld\n”, dbuserl.usernm, dbuserl.aid, statusl); uexit(1); } else fprintf(stderr, ”User %s not registered in database\n”,argv[1]); See Also uaddusr, udrpusr For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 421 umodvol Modify a volume name #include <include/rhli.h> #include <include/rhlierr.h> Syntax int umodvol( txid, nmods, dbvolml, statusl, status ) UTXID txid; int nmods; DBVOLM *dbvolml; USTATUS statusl[]; USTATUS *status; Arguments Description txid The transaction ID nmods Number of volumes in the dbvolml list dbvolml List of DBVOLM structures, each describing a separate volume to modify statusl Returned list of status values, one for each dbvolml entry status Returned status of the function. Modifies the volume name for each volume listed in the dbvolml list. Each DBVOLM structure in the dbvolml list contains information about a volume. Because the umodvol function can change a volume name only, initialize only the DBVOLM structure members volname and vid. For a description of the DBVOLM structure, see “The RHLI Structures” chapter. The new volume name must be unique within the current database. Security You must have DBA authority to modify a volume. For information about granting DBA authority, see the uaddprv function. 422 Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the umodvol function to change a volume name. dbvolml.volname = ”master parts database”; if ( umodvol(txid, 1, &dbvolml, statusl, &status) != USUCCESS ) { fprintf(stderr, ”Failure modifying a volume: status = %ld\n”,status); fprintf(stderr, ”volume status = %ld\n”, *statusl); } See Also uaddvol, udrpvol For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 423 unumath Retrieve the number of schemas in the database #include <include/rhli.h> #include <include/rhlierr.h> Syntax int unumath( txid, count, status ) UTXID txid; long *count; USTATUS *status; Arguments txid The transaction ID count Pointer to the number of schemas currently in the database status Returned status of the function Description Returns the number of schemas registered in the database. Security The number of schemas is returned only for schemas to which you have access. You always have access to the current schema. Additionally, you can access other schemas if you have been granted schema access permission to the schema or if a schema you can access has schema privileges to another schema (or a table or column in the schema). If you have DBA authority, you can access any schema in the database. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. To grant schema access, use the uaddprm function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 424 Example This example calls the unumath function to retrieve the number of accessible schemas. /* define required parameters */ UTXID txid; long count; USTATUS status; /* open the database and begin a transaction */ ... /* Retrieve number of accessible schemas */ if (unumath(txid, &count, &status) != USUCCESS) { printf(”Cannot get the auth. count: status = %ld\n”, status); } See Also uinfath, ubndath For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 425 unumcgp Retrieve the number of columns in a column group #include <include/rhli.h> #include <include/rhlierr.h> Syntax int unumcgp( txid, cid, ngid, status ) UTXID txid; UCID cid; long *ngid; USTATUS *status; Arguments Description txid The transaction ID cid Column ID of a column ngid Pointer to a buffer which contains the number of columns in the column group, or, if the specified column is not a column group, the number of column groups to which this column belongs status Returned status of the function. If the column ID (specified by cid) is member of a column group, the unumcgp function returns the total number of members in that column group that are available. Otherwise, the number of column groups to which the ordinary column belongs is returned. There is no indication as to which count is returned; for further information on how to identify the type of column, see the uinfcol function. 426 Security The unumcgp function counts and returns the number of columns in column groups of tables that you have access to only. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. To grant schema access, use the uaddprm function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the unumcgp function to retrieve the number of column groups in the parts table. /* define RHLI interface required parameters */ UTXID txid; long count; USTATUS status; if (unumcgp(txid, $parts.$, &count, &status) != USUCCESS) { printf(”Could not get the number of column groups:”); printf(” status = %ld\n”, status); uexit(101); } See Also unumtbl, unumrow, ubndtbl, uinfcol For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 427 unumchd Retrieve the number of child rows for a link #include <include/rhli.h> #include <include/rhlierr.h> Syntax int unumchd( txid, ptid, prid, lid, nchild, status ) UTXID txid; UTID ptid; URID prid; ULID lid; long *nchild; USTATUS *status; Arguments txid The transaction ID ptid Table ID of the parent row of the link prid Row ID of the parent row of the link lid Link ID of the link nchild Pointer to the number of rows that are children of the known parent row status Returned status of the function. Description Given a parent row ID, returns the number of rows that are children of the specified parent row of a link. Because multiple links can exist on a given parent table, the Link ID is used to determine the child table to query. Security Link information is returned only for links in tables that you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. 428 To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. To grant schema access, use the uaddprm function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the unumchd function to retrieve the number of child rows in the link index. /* define RHLI interface required parameters */ UTXID txid; URID prid; long nchild; USTATUS status; /* this assumes a database is open & a transaction begun */ if ( unumchd(txid, $parts.$, prid, $@parts_avail$, &nchild, &status) != USUCCESS ) { printf(”Cannot get the child row count: status = %ld\n”, status); uexit(99); } See Also uaddlnk, ubndtbl, udrplnk, unumcol, unumprn For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about using links, see “Choosing an Access Method” in Unify DataServer: Managing a Database. 429 unumcol Retrieve the number of columns in a table #include <include/rhli.h> #include <include/rhlierr.h> Syntax int unumcol( txid, tid, ncol, status ) UTXID txid; UTID tid; long *ncol; USTATUS *status; Arguments txid The transaction ID tid Table ID of the table containing the columns ncol Pointer to the number of columns in the table specified status Returned status of the function. Description Returns the number of available columns for the specified table. Security Column information can be retrieved only for columns in tables that you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 430 Example This example uses the unumcol function to retrieve the number of accessible columns in the parts table. /* define RHLI interface required parameters */ UTXID txid; long ncol; USTATUS status; if ( unumcol(txid, $parts.$, &ncol, &status) != USUCCESS) { printf(”Could not get the number of columns:”); printf(” status = %ld\n”, status); uexit(99); } See Also unumtbl, unumrow, ubndtbl, uinftbl For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 431 unumprn Retrieve the number of parent rows for a link #include <include/rhli.h> #include <include/rhlierr.h> Syntax int unumprn( txid, ctid, crid, lid, nparent, status ) UTXID txid; UTID ctid; URID crid; ULID lid; int *nparent; USTATUS *status; Arguments Description txid The transaction ID ctid Table ID of the child row of the link crid Row ID of the child row of the link lid Link ID of the link nparent Pointer to the number of parent rows status Returned status of the function. Given a child row ID, returns the number of rows (0 or 1) that are parents of the specified child row of the link. A return value of zero for nparent implies the child row link column has a null value. Because multiple links can exist on a given parent table, the Link ID must be specified to determine the parent table to query. A child row can have only zero or one parent. 432 Security Link information is returned only for links in tables that you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the unumprn function to determine if a row is a parent row in a link index. /* define required parameters */ UTXID txid; URID crid; long nparent; USTATUS status; if ( unumprn(txid, $parts.$, crid, $@parts_used$, &nparent, &status) != USUCCESS) { printf(”Cannot get the parent row count: status = %ld\n”, status); uexit(99); } See Also uaddlnk, udrplnk, unumcol, unumchd For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about using links, see “Choosing an Access Method” in Unify DataServer: Managing a Database. 433 unumrow Retrieve the number of rows in a table #include <include/rhli.h> #include <include/rhlierr.h> Syntax int unumrow( txid, tid, count, status ) UTXID txid; UTID tid; long *count; USTATUS *status; Arguments txid The transaction ID tid Table ID of the table count Pointer to the number of rows currently in the table status Returned status of the function. Description Returns the number of available rows in the specified table. Security Row information is returned only for tables you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 434 Example This example uses the unumrow function to return the number of accessible rows. /* define required parameters */ UTXID txid; long count; USTATUS status; /* this assumes a database is open & a transaction begun */ if ( unumrow(txid, $parts.$, &count, &status) != USUCCESS) { printf(”Cannot get the row count: status = %ld\n”, status); uexit(99); } See Also unumcol, unumtbl, ubndtbl For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 435 unumtbl Retrieve the number of tables in the database #include <include/rhli.h> #include <include/rhlierr.h> Syntax int unumtbl( txid, aid, ntbl, status ) UTXID txid; UAID aid; long *ntbl; USTATUS *status; Arguments Description txid The transaction ID aid A valid authorization ID for a schema or UNULLAID ntbl Pointer to the number of tables currently in the database status Returned status of the function. Returns the number of available tables in the database. If aid is set to UNULLAID, the unumtbl function returns the total number of tables the current schema has access to; otherwise, if aid is set to an authorization ID for a schema, the number of tables that are accessible to the specified schema is returned. Security The count is returned for tables that you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. To make a schema current, use the ufchath function. To grant authority or schema privileges, use the uaddprv function. 436 Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the unumtbl function to return the number of accessible tables. /* define required parameters */ UTXID txid; long ntbl; USTATUS status; if ( unumtbl(txid, aid, &ntbl, &status) != USUCCESS) { printf(”Cannot get the table count: status = %ld\n”, status); uexit(99); } printf(”There are currently %ld tables in the database.\n”, ntbl); See Also unumcol, unumrow For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 437 unumusr Retrieve the number of users in the database #include <include/rhli.h> #include <include/rhlierr.h> Syntax int unumusr( txid, nuser, status ) UTXID txid; long *nuser; USTATUS *status; Arguments txid The transaction ID nuser Pointer to the number of users currently in the database status Returned status of the function Description Returns the number of users registered in the database. Security The number of users are returned as follows: If you have DBA authority, the count of all users currently accessing the database is returned. If you don’t have DBA authority, only your user ID is counted. For information about granting DBA authority, see the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 438 Example This example calls the unumusr function to return the number of database users. /* define required parameters */ UTXID txid; long nuser; USTATUS status; /* open the database and begin a transaction */ ... /* retrieve number of users */ if ( unumusr(txid, &nuser, &status) != USUCCESS ) { printf(”Cannot get the user count: status = %ld\n”, status); uexit(99); } See Also unumrow, unumcol, unumtbl, unumvol For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 439 unumvol Retrieve the number of volumes in the database #include <include/rhli.h> #include <include/rhlierr.h> Syntax int unumvol( txid, nvol, status ) UTXID txid; int *nvol; USTATUS *status; Arguments txid The transaction ID nvol Number of volumes currently in the database status Returned status of the function. Description Counts the number of volumes in the database. Security You must have DBA authority to retrieve the number of volumes in the database. For information about granting DBA authority, see the uaddprv function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the unumvol function to retrieve the number of active volumes in the database: /* define RHLI interface required parameters */ UTXID txid; long nvol; USTATUS status; 440 /* this assumes a database is open & a transaction begun */ if ( unumvol(txid, &nvol, &status) != USUCCESS ) { printf(”Cannot get the volume count: status = %ld\n”, status); uexit(99); } See Also unumrow, unumcol, unumtbl, unumusr For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about managing database volumes, see “Managing the Database Files” in Unify DataServer: Managing a Database. 441 uopndb Open a database #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uopndb( db_name, opnopts, dbid, status ) char *db_name; UOPTS opnopts; UDBID *dbid; USTATUS *status; Arguments Description db_name A fully-qualified database name. If parts of the fully-qualified name are omitted (or have the value (char *)0), the values in the DBHOST, DBUSER, DBPATH, and DBNAME configuration variables determine the missing part. opnopts Database open options; see the following sections dbid Pointer to a returned database ID that uniquely describes this database status Returned status value of the operation Opens the database specified by database name. The values of the DBNAME, DBPATH, DBHOST, and DBUSER configuration variables affect the default database as described in Unify DataServer: Configuration Variable and Utility Reference. When a database is opened, a database ID is returned (pointed to by dbid). This database ID should be used for all future references to the database. A database opened during the execution of a program must be closed (by calling the uclsdb function) before the program terminates. Only one database can be opened at a time. Be sure to close an open database (by calling the uclsdb function) before opening a new one. 442 When the uopndb function is called for the first time, several daemons that Unify DataServer uses to manage database operations are started up. These daemons stay running until the database is shut down (by using the shutdb utility). Once the daemons have been started, all subsequent calls to the uopndb function (by any transaction) do not have to re-start the daemons, causing the database to be opened faster than the first time. 1. Database Valid opnopts values are described below. To specify more than one option, Open Options separate options in the argument list with the bitwise logical OR operator (|). DB_NORMAL Normal mode; default actions are taken for all categories of options described below. If DB_NORMAL is combined with any other option, that option overrides the default behavior in its category. Definition Lock Options DB_LCKSCH Lock all table definitions in the database regardless of whether or not the object is referenced. If you specify this option, no data definition operations can be performed in the database. That is, no adding or dropping of tables, columns, or indexes is allowed. DB_NOLOCK Do not lock any table definitions. If you specify this option, your transaction can perform data definition operations in the database. You can acquire and release definition locks explicitly at runtime by calling the ulcksch and uulksch functions. You can use the name binding functions to perform all name binding at runtime. If no definition lock options are specified, the DB_NORMAL option places shared locks only on object definitions that are referenced at compile time by your application. These locks prevent all transactions (including your own) from redefining any of these database objects while this application has the database open. Unify DataServer uses an object ID list created by the Unify DataServer linker, uld, to lock objects referenced at compile time when the DB_NORMAL option is chosen. If you do not use ucc and uld to compile and link your application programs, no definition locks are acquired. Definition locks acquired by the DB_NORMAL option are released only when the database is closed. 443 If you specify DB_NOLOCK, you cannot also specify DB_REMAP. You cannot specify both DB_NOLOCK and DB_LCKSCH. Signal Handling Options DB_NOEXIT Do not exit when an interrupt occurs; interrupt signals should be handled by Unify DataServer. When an interrupt signal is received, all RHLI functions after the signal fail with a status value of UEINTRPT until the signal is cleared with the uclritr function. The effect of DB_NOEXIT is ignored in these cases: The DB_NOSIGS option is also specified. The DB_EXIT option was specified on a previous usetitr call. DB_NOSIGS Do not install signal handlers when you open the database. Possible reasons for wanting to use the DB_NOSIGS option are: You are certain that stray interrupt signals from other users will not occur. You want to handle interrupt signals yourself. You want increased performance when opening the database since the signal handlers do not need to be installed. The signal handlers monitor and handle interrupt signals while the database is opening. For example, if a “kill” signal is received (such as a kill –3), Unify DataServer makes every attempt to bring the database to a consistent state (by ending transactions, releasing locks, writing updates to the disk, etc.) before aborting the current operation. Warning Some signals, if not handled by your application, can corrupt your database. Use DB_NOSIGS with caution. Compile-time ID Mapping Options DB_REMAP Remap compile-time objects produced during compile-time name-binding, to runtime object IDs. This runtime feature is useful for cases where your application is run on machines where the database may have been redefined, causing the original object IDs to have changed. When specifying the DB_REMAP option, your application must be loaded by using the Unify DataServer loader, uld. 444 If you specify DB_REMAP, you cannot also specify DB_NOLOCK. DB_PARTMAP Used with the DB_REMAP option to perform partial ID-mapping. You must also specify the DB_REMAP option. You must use the special syntax compiler directives in place of each object ID referenced in the application program. DB_CONDMAP Used with the DB_PARTMAP option to indicate conditional ID-mapping. You must also specify the DB_REMAP option. Performance Specifying the DB_REMAP option can impact performance when opening a database since remapping is performed on all object IDs defined for the database, whether or not the IDs are accessed. Database Update Options The DB_RDONLY option means that only non-update, insert, or delete operations are allowed. The database can be opened with DB_RDONLY mode even if other processes have opened the database for update operations. If DB_RDONLY is not specified, DB_NORMAL opens the database in read/write mode. Security To open a database, one of the following conditions must be true: You are the creator of the database. You have access to the database. For information about creating a database, see the uadddb function; for information about granting database access, see the uaddusr function. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 445 Example This example uses the uopndb function to open the database. /* define required parameters */ UDBID dbid; char * dbname = ”parts.db”; USTATUS status; ... /* open ’parts.db’ database */ if ( uopndb(dbname, DB_NORMAL, &dbid, &status) != USUCCESS) { printf(”Unable to open parts database %s\n”, dbname); uexit(1); } See Appendix A for more examples. See Also ubegtx, uclsdb, uinfdb, ulcksch fflush(3) in your UNIX system manual AUTOSTART configuration variable in Unify DataServer: Configuration Variable and Utility Reference For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 446 uopnjrn Open a journal file Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uopnjrn(dbname, jrnname, jrnfd, sequence, status) char *dbname; char *jrnname; int *jrnfd; long *sequence; USTATUS *status; Arguments Description dbname Pointer to the path name of the database associated with the journal jrnname Pointer to the path name of the journal file to open jrnfd Pointer to the returned file descriptor for the opened journal file sequence Pointer to the returned journal sequence number of the journal header status Returned status of the function Opens the specified journal file. The uopnjrn function call returns a file descriptor of the opened journal file and the journal sequence number contained in the journal header. The sequence number is a unique number assigned to each journal file when it is created. The first journal created following a backup of the database is assigned a sequence number of 1. The sequence number is incremented for each subsequent journal file. Information in the journal can then be read using the ufchjrn function call. After processing the journal file, the uclsjrn function closes the journal and frees up resources. 447 The uopnjrn function implicitly opens and closes the database in order to extract schema information needed to process the journal file. Therefore the database cannot be opened at the time that the uopnjrn function call is called. Only one journal file can be open at a time. Security To open and process the journal, one of the following conditions must be true: You have DBA authority. You have access to the tables and columns that are referenced in the journal. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example See Appendix A. See Also ufchjrn, uclsjrn For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about transaction management and using the transaction journal, see “Managing Transactions and Locks”and “Preventing Data Loss” in Unify DataServer: Managing a Database. 448 upkfrow Fetch a row by the value of its primary key Syntax #include <include/rhli.h> #include <include/rhlierr.h> int upkfrow(txid, tid, lcktype, rid, keycoll, unipcoll, statusl, status) UTXID txid; UTID tid; int lcktype; URID *rid; UNIPCOLL *keycoll; UNIPCOLL *unipcoll; USTATUS statusl[]; USTATUS *status; Arguments txid The transaction ID tid Table ID of the table that contains the row to fetch lcktype Lock type to place on the fetched row. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock rid Returned row ID of the retrieved row keycoll A pointer to a UNIPCOLL structure that specifies the primary key column list unipcoll A pointer to a UNIPCOLL structure that specifies the column retrieval list statusl Returned list of status values, one for each keycoll/unipcoll list entry status Returned status of the function 449 Description Retrieves columns of the indicated row from the specified table by using the supplied primary key. For the keycoll argument, specify any multiple key column values in any order. A primary key column cannot be a null value. UNIPCOLL contains several lists that describe both the column and the column’s value. All of the columns specified in the column list structure must belong to the table ID specified. For a description of how to initialize the UNIPCOLL structure, see “The RHLI Structures” chapter. Security A row can be returned only for tables you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also ufchrow, ufchtbl, ufchlnk For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 450 uposord Position by values to a row within an ordered access Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uposord (txid, oaccid, rop, keyvall, lcktype, ridp, statusl, status) UTXID txid; UOACCID oaccid; int rop; UDBVAL *keyvall; int lcktype; URID *ridp; USTATUS statusl[]; USTATUS *status; Arguments txid Current transaction ID oaccid An ordered access ID as returned from the ubegord function rop A value indicating the relational operator (in comparison with the beginning of the order) for the comparison. Options are: ULTTST Greatest value less than. ULETST Greatest value less than or equal to. UEQTST Equal to. UGETST Least value greater than or equal to. UGTTST Least value greater than. keyvall A pointer to a list of UDBVAL structures, one for each key column whose value is to be positioned to. lcktype Lock type to place on row positioned to. Options are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock 451 Description ridp Returned as the row ID of the new current row statusl Returned list of status values, one for each column value specified in the keyvall list status Returned status of the operation Alters the current position within the ordered access specified by oaccid. If the function returns USUCCESS, the current position is set to the row whose column values in conjunction with the relational operator match the column values specified by the keyvall list. The row is locked with a lock no lower than the requested lock. The relational operator is in terms of the order specified by the UOACOLS structure list passed to the ubegord function, not the physical order of the underlying access method. For example, for an ascending order with values 1, 2, 3, 3, 3, 4, and 5, UGTTST on value 3 will position to value 4. For the same values with a descending order (5, 4, 3, 3, 3, 2, 1) UGTTST on value 3 will position to value 2. A call to the ufchord function with a direction parameter of USAME, UNEXT, or UPREV immediately following a call to the uposord function returns the row positioned to by the uposord function. If the function returns USUCCESS, the row ID of the new current row is returned with a lock no lower than the requested lock. If the function fails due to the inability to obtain the specified row lock, the current position does not change. The position remains set to the position before the uposord function was called. Security No privileges required. Example See Appendix A. See Also ubegord, uendord, ufchord, uskrord 452 usetitr Set an interrupt; disallow database operations Syntax #include <include/rhli.h> #include <include/rhlierr.h> int usetitr( itropts, status ) UOPTS itropts; USTATUS status; Arguments itropts Process interrupt option. Options are: DB_EXIT Gracefully exit the application after an interrupt has occurred DB_NORMAL Return to the application after an interrupt has occurred DB_NOEXIT status Description Do not exit when an interrupt occurs Returned status of the function Causes the current process to disallow database operations both in RHLI and in Embedded SQL/A. The usetitr function should be called only when an interrupt signal has been caught by a signal handler routine in your application program. All database operations are disallowed while the interrupt is set, with the exception of certain operations providing for process termination and closing of the database. The database interrupts are not implemented through the UNIX signal function interfaces. Once an interrupt is set, all Unify DataServer-managed child processes are terminated at a point where the database is in a known consistent state. Process interruption by any other mechanism may require database recovery and is not recommended. Before setting an interrupt with usetitr, be sure that a database is currently open. 453 If a database is opened by specifying the DB_NOEXIT option to the uopndb function, or if the database is opened by using the Embedded SQL/A CONNECT or OPEN DATABASE statements, the current application does not exit when an interrupt occurs. However, if you specify the DB_EXIT option (for itropts) when calling the usetitr function, the uopndb function DB_NOEXIT option is overridden, and the application exits on this interrupt. Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example calls the usetitr function when the SIGQUIT signal occurs. /* add include files */ #include <stdio.h> #include <signal.h> #include <include/rhli.h> /* a signal handler routine */ void syncsig(syssig) int syssig; /* signal value received */ { USTATUS status; /* function status */ ... (void)signal(syssig, syncsig); /* determine the signal type */ switch ( syssig ) { case SIGHUP: { syncprt(); /* print statistics */ break; } case SIGQUIT: { usetitr(DB_EXIT, &status); /* exit application */ break; } case SIGUSR1: { sleeptm = 1; /* reset global sleep time */ break; 454 } default: { printf(”unknown signal caught: %d\n”, syssig); break; } } return; } See Also uclritr signal(2) in your UNIX system manual 455 uskrord Alter the current position within the ordered access #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uskrord (txid, oaccid, lcktype, rid, status) UTXID txid; UOACCID oaccid; int lcktype; URID rid; USTATUS *status; Arguments Description txid Current transaction ID (from the ubegtx or ucbgtx functions) oaccid An ordered access ID as returned from the ubegord function lcktype Lock type to place on row positioned to. Valid lock type values are: UNULLLCK No lock USLCK Shared lock UXLCK Exclusive lock rid Row ID of the row to seek to status Returned status of the operation Given a valid Row ID, the uskrord function alters the current position within the ordered access specified by oaccid. The current position is set to the row specified by rid. If the function fails due to the inability to obtain the specified row lock, the current position does not change. The position remains set to the position before the uposord function was called. 456 A call to the ufchord function with a direction parameter of USAME, UNEXT, or UPREV immediately following a call to the uskrord function returns the row positioned to by the uskrord function. If the function returns USUCCESS, the row is locked with a lock no lower than the requested lock. Security No privileges required. Example See Appendix A. See Also ubegord, uendord, ufchord, uposord 457 utrnctbl Truncate a table #include <include/rhli.h> #include <include/rhlierr.h> Syntax int utrnctbl( txid, tid, status) UTXID txid; UTID tid; USTATUS *status; Arguments txid Transaction ID tid Table ID of the table to be truncated. The table cannot be a child in a link index and it can only be a parent in a link index if the child tables are empty. The table cannot have any B-trees stored in external files (.idx) status Description Returned status of the function The utrnctbl function deallocates the allocated segments for the specified table and returns the segments to the volume free space. All data in the table and in its indexes is permanently dropped. The table definition and any index definitions that the table has, however, are retained. Any triggers on the table are not executed and any DIS definitions for the table are ignored. Use the utrnctbl function when you want to quickly free up the segments being used by a table but retain the table definition for future use. (The udrptbl function also frees the allocated segments but does not retain the table definition.) Although this is a DML operation, it is not recoverable. The table truncation is logged however, and therefore replayable by the irma utility. The utrnctbl function must be performed within its own transaction and the transaction cannot be rolled back. 458 If the table being truncated contains variable-length data, the space held in the variable–length data volumes must be released. For information on how to do this, see the entry titled “Reclaim utility is no longer available” in the Unify DataServer README. The utrnctbl() function is not available with remote access. Security To use this statement, one of the following conditions must be true: you have DBA authority the table is in the current schema the table is in another schema and the current schema has DELETE privilege on the table or the schema containing the table Example The following example uses the utrnctbl() function to truncate a table. /* define required parameters */ UTXID txid; USTATUS status; ... /* start a new transaction */ ... /* xlock the table to be truncated */ ... /* truncate table scratchTbl */ if ( utrnctbl(txid, $scratchTbl.$, &status) != SUCCESS ) { /* handle error */ fprintf(stderr, “Error truncating scratchTbl: status = %ld\n”, status); ... } /* commit the transaction */ ... 459 uulkdb Unlock a database #include <include/rhli.h> #include <include/rhlierr.h> Syntax int uulkdb( txid, status ) UTXID txid; USTATUS *status; Arguments Description txid The transaction ID. status Returned status of the function Unlocks the database that acquired a lock during the specified transaction. Unlocking a database releases all locks on the tables and rows in the database. An exclusive lock is not released until the transaction that acquired the exclusive lock is committed or aborted (by calling the ucmttx, ucbgtx, or uabttx function). Security No privileges required. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Example This example uses the uulkdb function to unlock the database. if ( uulkdb(txid, &status) != USUCCESS ) { printf(”Unable to unlock database: status = %ld\n”, status); uexit(99); } See Also ulckrow, uulkrow, ulcktbl, uulktbl, ulckdb, ubegtx For information about transaction and lock management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 460 uulkrow Unlock a row Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uulkrow( txid, tid, rid, status ) UTXID txid; UTID tid; URID rid; USTATUS *status; Arguments Description txid The transaction ID. tid Table ID of the table containing the row to unlock rid Row ID of the row to unlock status Returned status of the function Unlocks the row ID for the specified table so that other transactions can change the row’s column data. An exclusive lock is not released until the transaction that acquired the exclusive lock is committed or aborted (by calling the ucmttx, ucbgtx, or uabttx function). Security Rows can be unlocked only for tables you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 461 Example This example uses the uulkrow function to unlock a row (with a rowid of 3) in the company table. /* define RHLI interface required parameters UTXID txid; URID rid; USTATUS status; /* open the database and begin a transaction ... /* unlock the row */ rid = 3; if ( uulkrow(txid, $company.$, rid, &status) { printf(”Unable to unlock row %ld: status status); uexit(99); } See Also */ */ != USUCCESS ) = %ld\n”, rid, ulckdb, ulckrow, ulcktbl, uulktbl, uulkdb For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about transaction and lock management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 462 uulksch Unlock a table definition Syntax #include <include/rhli.h> #include <include/rhlierr.h> bool uulksch( txid, tid, status ) UTXID txid; UTID tid; USTATUS *status; Arguments Description txid Current transaction ID tid Table ID of the table to unlock status Returned status of the function Unlocks (or releases) a table definition lock placed by the ulcksch function. A table definition lock prevents any other transactions from changing or deleting the definition for a table while table data is being accessed. A table definition lock is separate from and does not interfere with a data lock. To release a data lock, see the uulktbl or uulkrow function. Security Definition locks can only be released on tables you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 463 Example This example uses the uulksch function to release a table definition lock for the company table. /* define RHLI interface required parameters */ UTXID txid; USTATUS status; /* open the database and begin a transaction */ ... /* Unlock the table definition lock */ if (uulksch(txid, $company.$, &status) != USUCCESS) { printf(”definition unlock of company table failed—%ld\n”, status); uexit(99); } See Also ulcksch, uinfsch For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about transaction and lock management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 464 uulktbl Unlock a table Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uulktbl( txid, tid, status ) UTXID txid; UTID tid; USTATUS *status; Arguments Description txid Current transaction ID tid Table ID of the table to unlock status Returned status of the function Unlocks all rows in that table specified by the table ID. An exclusive lock is not released until the transaction that acquired the exclusive lock is committed or aborted (by calling the ucmttx, ucbgtx or uabttx function). Security Only tables that you have access to can be unlocked. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. 465 Example This example uses the uulktbl function to unlock the company table. /* define RHLI interface required parameters */ UTXID txid; USTATUS status; /* open the database and begin a transaction */ ... /* lock the table */ if ( uulktbl(txid, $company.$, &status) != USUCCESS ) { printf(”Unable to unlock table: status = %ld\n”, status); uexit(99); } See Also ulckrow, uulkrow, ulcktbl, ulckdb, uulkdb For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. For information about transaction and lock management, see “Managing Transactions and Locks” in Unify DataServer: Managing a Database. 466 uupdrow Update a row Syntax #include <include/rhli.h> #include <include/rhlierr.h> int uupdrow( txid, tid, rid, unipcoll, statusl, status ) UTXID txid; UTID tid; URID rid; UNIPCOLL *unipcoll; USTATUS statusl[]; USTATUS *status; Arguments Description txid Current transaction ID tid Table ID of the table containing the row to update rid Row ID of the row to update unipcoll A pointer to a UNIPCOLL structure specifying the column update list statusl Returned list of status values, one for each of the columns listed in unipcoll status Returned status of the function Updates a row according to the column values specified by the structure UNIPCOLL. Columns specified in UNIPCOLL must belong to the table ID specified by tid. UNIPCOLL contains several lists which describe both the column and the column value. For a description of how to initialize the UNIPCOLL structure, see “The RHLI Structures” chapter. If any column has legal values defined by the Data Integrity Subsystem (DIS), the valid DIS value is checked before any update is allowed. If any column value is rejected, the entire row remains unmodified. For more information about DIS, see Unify DataServer: Managing a Database. 467 When inserting into or updating columns of type U_STR (string), Unify DataServer adds trailing blanks to values less than the total size defined for the column. Security Rows can only be updated for tables you have access to. You can always access all tables in the current schema. Additionally, you can access tables (or just columns within a table) in other schemas if your current schema has been granted schema privileges to the other schema. If you have DBA authority, you can access any table in the database. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. See Also uinsrow, udelrow For information about database security, see “Setting Up and Managing Security” in Unify DataServer: Managing a Database. 468 uXMLAddAttribute() Adds a new attribute to an element Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLAddAttribute ( xmlObj, xmlStatus, attributeName, attributeValue ) xml_obj* xmlObj; USTATUS xmlStatus; const char* attributeName; const char* attributeValue; Arguments xmlObj DOM object reference xmlStatus Returned status of the function attributeName Name of the new attribute to be added attributeValue Value of the new attribute Description Adds a new attribute and its value to the current element. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. 469 Example The following lines in the “person.xml” file define a person element. <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; const char* attributeName= “mothername”; const char* attributeValue= “Maria”; USTATUS xmlStatus; .... The following lines in an RHLI program check the “person.xml” for wellformedness, parse it, and print the result. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following line in an RHLI program adds a new attribute to the current element. If the current element pointer for element is at “<name>” , and attributeName=”mothername” and attributeValue=”Maria”, then the XML becomes: <_> <name surname=”anderson” fathername=”mike” mothername=”Maria”>Jack</name> <_> result = uXMLAddAttribute(xmlObj,&xmlStatus,attributeName,attributeValue); See Also uXMLCreateElement 470 uXMLCreateElement() Creates a new element Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLCreateElement ( xmlObj, xmlStatus, elementName, elementValue ) xml_obj* xmlObj; USTATUS xmlStatus; const char* elementName; const char* elementValue; Arguments xmlObj DOM object reference xmlStatus Returned status of the function elementName Name of the element elementValue Value of the element Description Creates a new element and appends it as child element to the current element. The value of the child element is set to the specified elementValue. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. 471 Example The following lines in the “person.xml” file define a person element. <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; const char* elementName= “country”; const char* elementValue= “USA”; USTATUS xmlStatus; .... The following lines in an RHLI program check the “person.xml” for wellformedness, parse it, and print the result. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following line in an RHLI program adds a new element to the current document as a child element of the current element. If the current element pointer is at <person>, elementName=”country” and elementValue=”USA”, then XML document after the function is called becomes: <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> <country>USA</country> </person> result =uXMLCreateElement(xmlObj,&xmlStatus,elementName,elementValue); See Also uXMLAddAttribute 472 uXMLDeleteAttribute() Deletes an attribute Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLDeleteAttribute( xmlObj, xmlStatus, attributeName ) xml_obj* xmlObj; USTATUS xmlStatus; const char* attributeName; Arguments xmlObj DOM object reference xmlStatus Returned status of the function attribute Name Name of the attribute to be deleted Description Deletes the specified attribute from the current element. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information can be obtained by calling getErrorInfo(). Security No privileges needed. The following lines in the person.xml file define a person element. 473 <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> const char* xmlfile=”person.xml”; xml_obj xmlObj; int result = 0; const char* attributeName= fathername ; USTATUS xmlStatus; .... The following lines in an RHLI program checks the person.xml file for wellformedness, parses it, and prints the result. result = uXMLParse( &xmlObj, &xmlStatus, xmlfile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following line in an RHLI program deletes the attribute of the current element. If the current element pointer is set to <name> and the function is called with attributeName= fathername”, the XML becomes: <&> <name surname=”anderson” >Jack</name> <&. > result =uXMLDeleteAttribute(xmlObj,&xmlStatus,attributeName); See Also uXMLChangeAttributeValue 474 uXMLDeleteCurrentElement() Deletes the current element Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLDeleteCurrentElement ( xmlObj, xmlStatus ) xml_obj* xmlObj; USTATUS xmlStatus; Arguments xmlObj DOM object reference xmlStatus Returned status of the function. Description Deletes the current element from the document and sets the current element pointer to the parent element of the deleted element. If this function is called with the current element set to the root element of the document, then the entire underlying document structure is removed. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. Example The following lines in the “person.xml” file define a person element. <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; USTATUS xmlStatus; .... 475 The following lines in an RHLI program check the “person.xml” file for wellformedness, parse it, and print the result. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following line in an RHLI program deletes the current element from the document. If the current element pointer is at element <name>, and the function uXMLDeleteCurrentElement() is called, it deletes the <name> element from the document. The XML document becomes: <person> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> result =uXMLDeleteCurrentElement(xmlObj,&xmlStatus); See Also uXMLDeleteElementValue, uXMLModifyElementValue, uXMLDeleteAttribute 476 uXMLDeleteCurrentElementValue() Deletes the value of an element Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLDeleteCurrentElementValue ( xmlObj, xmlStatu s) xml_obj* xmlObj; USTATUS xmlStatus; Arguments xmlObj DOM object reference xmlStatus Returned status of the function. Description Deletes the value of the current element. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. Example The following lines in the “person.xml” file define a person element. <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; const* char elementName= “name”; USTATUS xmlStatus; .... 477 The following lines in an RHLI program check the “person.xml” file for wellformedness, parse it, and print the results. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following lines in an RHLI program delete the value of the current element. If the current element pointer is at element <age> then the uXMLDeleteCurrentElementValue() function deletes the value for the element <age>. The document becomes <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”></age> <me>John</me> </person> result =uXMLDeleteCurrentElementValue(xmlObj, &xmlStatus); 478 uXMLDestroy Destroys an XML object Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <uXML.h> int uXMLDestroy (xmlObj, xmlStatus) xml_obj* xmlObj; USTATUS xmlStatus; Arguments xmlObj Pointer to Object reference xmlStatus Returned status of the function. Description Destroys the given xmlObj Reference. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Security No privileges needed. Example const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; USTATUS xmlStatus; .... The following lines in an RHLI program check the “person.xml” file for wellformedness, parse it, and print the results. 479 result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... result =uXMLDestroy(&xmlObj, &xmlStatus); 480 uXMLGetAttributeCount() Returns the number of attributes Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLGetAttributeCount ( xmlObj, xmlStatus, count ) xml_obj* xmlObj; USTATUS xmlStatus; int* count; Arguments xmlObj DOM object reference xmlStatus Returned status of the function count Returned count of attributes of an element Description Returns the count of attributes associated with the current element. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. Example The following lines in the “person.xml” file define a person element. <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person>const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; int count; char* name; USTATUS xmlStatus; .... 481 The following lines in an RHLI program checks the “person.xml” file for wellformedness, parses it, and prints the result. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following line in an RHLI program returns the number (integer value) of attributes associated with current element. If the current element is <name> then the uXMLGetAttributeCount() function sets the count variable to 2. result = uXMLGetAttributeCount( xmlObj, &xmlStatus, &count ); See Also uXMLGetAttributeNames, uXMLGetAttributeValues. 482 uXMLGetAttributeNames() Returns the attribute names of an element Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLGetAttributeNames ( xmlObj, xmlStatus, attributeNames ) xml_obj* xmlObj; USTATUS xmlStatus; char*** attributeNames; Arguments xmlObj Object reference xmlStatus Returned status of the function attributeNames Pointer to an array of attribute names, which are allocated and returned by the function. The array is terminated by a NIL string. Description The function allocates an array of attribute names as strings. The calling program is responsible to free the memory occupied by the array. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. 483 Example The following lines in the “person.xml” file define a person element. <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; char * attributeNames; USTATUS xmlStatus; .... The following lines in an RHLI program check the “person.xml” for wellformedness, parse it, and print the result. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following line in an RHLI program assigns the list of the attribute names to “result”. If the current element pointer is at <name>,the function returns attributeNames as an array of character pointers terminated by NIL(char): surname, fathername. result = uXMLGetAttributeNames(xmlObj, &xmlStatus, &attributeNames); ... /* use attributeNames */ char ** name = attributeNames; printf(”\nThe attribute names are:\n”); for( ; *name != NIL(char); ++name ) { printf(”\t%s\n”, *name); free(*name); } free(attributeNames); attributeNames = 0; See Also uXMLGetAttributeValues, uXMLGetAttributeCount 484 uXMLGetAttributeValue() Returns an attribute value Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLGetAttributeValue ( xmlObj, xmlStatus, attributeName, attributeValue ) xml_obj* xmlObj; USTATUS xmlStatus; const char* attributeName; char** attributeValue; Arguments xmlObj DOM object reference xmlStatus Returned status of the function attributeName Name of the attribute whose value you want to retrieve attributeValue Returned value of attribute Description Returns the value of the specified attribute (attributeName) of the current element. The calling program is responsible to free the memory occupied by the attributeValue variable. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. For a list of the most commonly occurring function status values, see the status code at beginning of this document. For a complete list of status returns and the associated error messages with their explanations and corrections, see Unify DataServer: Error messages. 485 Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. Example The following lines in the “person.xml” file define a person element. <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; const char* attributeName= “fathername”; char* attributeValue; USTATUS xmlStatus; .... The following lines in an RHLI program check the “person.xml” for wellformedness, parse it, and print the result. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following line in an RHLI program returns the value of an attribute of the current element. If the current element pointer is set to <name> and the uXMLGetAttributeValue() function is called with attributeName=”fathername”, the returned attribute value is ”mike”. result = uXMLGetAttributeValue(xmlObj, &xmlStatus, attributeName, &attributeValue) See Also uXMLGetAttributeValues, uXMLGetAttributeNames 486 uXMLGetAttributeValues() Returns the attribute values of an element Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLGetAttributeValues ( xmlObj, xmlStatus, attributeValues ) xml_obj* xmlObj; USTATUS xmlStatus; char** attributeValues; Arguments xmlObj DOM object reference xmlStatus Returned status of the function attributeValues Pointer to an array of attribute values, which are allocated and returned by the function. the array is terminated by a NIL string. Description This function allocates an array of attribute names as strings. The calling program is responsible to free the memory occupied by the attributeValues variable. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. 487 Example The following lines in the “person.xml” file define a person element. <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; char* attributeValues; USTATUS xmlStatus; .... The following lines in an RHLI program check the “person.xml” for wellformedness, parse it, and print the result. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following line in an RHLI program assigns the list of the attribute values to “result”. If the current element pointer is at <name>, the function returns the values of all attributes as an array of character pointers terminated by NIL(char), such as: Sanderson, mike result=uXMLSetAttributeValues(xmlObj, &xmlStatus, &attributeValues); ... /* use attributeNames */ char ** name = attributeValues; printf(”\n the attribute values are:\n”); for( ; *name != NIL(char); ++name ) { printf(”\t%s\n”, *name); free(*name); } free(attributeValues); attributeValues = 0; } See Also uXMLGetAttributeNames, uXMLGetAttributeCount 488 uXMLGetCurrentElementName() Returns the tag name of the current element Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLGetCurrentElementName ( xmlObj, xmlStatus, name ) xml_obj* xmlObj; USTATUS xmlStatus, char** name; Arguments Description xmlObj DOM object reference xmlStatus Returned status of the function name Returned tag name (allocated) Returns the tag name of the current element in the name argument. The calling program is responsible to free the memory occupied by name. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. Example The following lines in the “person.xml” file define a person element. <person> <name surname=”anderson” fathername=”mike”>Jack</name> 489 <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; char* name; USTATUS xmlStatus; .... The following lines in an RHLI program check the “person.xml” file for wellformedness and parse it. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following lines in an RHLI program print the current element name. If the current element pointer is at element <person>, the uXMLGetCurrentElementName function sets the name to “person”. result = uXMLGetCurrentElementName( xmlObj, &xmlStatus, &name ); printf( ”\n name of element = %s \n”, name ); free (name); 490 uXMLGetErrorInfo() Returns an error description Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLGetErrorInfo ( xmlObj, xmlStatus, errorInfo ) xml_obj* xmlObj; USTATUS xmlStatus; char** errorInfo; Arguments Description xmlObj DOM object reference xmlStatus Returned status of the function. errorInfo Returned error information Returns a description (errorInfo) about the last error that occurred during XML processing. The calling program is responsible to free the memory occupied by errorInfo. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. 491 Example The following lines in the “person.xml” file define a person element. <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; char* errorInfo; USTATUS xmlStatus; .... The following lines in an RHLI program check the “person.xml” file for wellformedness, parse it, and print the results. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following line in an RHLI program provide the information about the last error; the errorInfo variable has the information on the error. result =uXMLGetErrorInfo(xmlObj,&xmlStatus,&errorInfo); 492 uXMLGetValueOfCurrentElement() Returns the value of the current element Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLGetValueOfCurrentElement( xmlObj, xmlStatus, elementValue ) xml_obj* xmlObj; USTATUS xmlStatus; char** elementValue; Arguments Description xmlObj DOM object reference xmlStatus Returned status of the function elementValue Returned value of the current element Returns the value of the current element in the elementValue argument. The caller is responsible to free the memory occupied by the elementValue variable. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. 493 Example The following lines in the “person.xml” file define a person element. <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; char* elementValue; USTATUS xmlStatus; .... The following lines in an RHLI program check the “person.xml” file for wellformedness and parses it. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following lines in an RHLI program print the current element name. If the current element pointer is at element <me>, the uXMLGetCurrentElementName() function sets the name variable to “John”. result = uXMLGetValueOfCurrentElement( xmlObj, &xmlStatus, &elementValue ); printf( ”\n name of element = %s \n”, elementValue ); free (elementValue); See Also uXMLGetCurrentElementName 494 uXMLModifyElementValue() Modifies the value of an element Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLModifyElementValue ( xmlObj, xmlStatus, elementValue ) xml_obj* xmlObj; USTATUS xmlStatus; const char* elementValue; Arguments xmlObj DOM object reference xmlStatus Returned status of the function elementValue New value of the element Description Modifies the current value of an element. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. Example The following lines in the “person.xml” file define a person element. <person> 495 <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me></person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; const char* elementValue=”Joseph”; USTATUS xmlStatus; .... The following lines in an RHLI program check the “person.xml” file for wellformedness, parse it, and then print the result. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following line in an RHLI program sets the new value to the current element. If the current element pointer is set to <me>, and the function is called with a new element value of “Joseph” (elementValue=”Joseph”), the function modifies the document as: <_> <me>Joseph</me> <_> result =uXMLModifyElementValue(xmlObj,&xmlStatus,elementValue); See Also uXMLDeleteCurrentElement, uXMLCreateElement, uXMLDeleteElementValue 496 uXMLMoveToFirstChild() Moves the current element pointer to the first child Syntax Arguments #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLMoveToFirstChild ( xmlObj, xmlStatus ) xml_obj* xmlObj; USTATUS xmlStatus; xmlObj DOM object reference xmlStatus Returned status of the function Description Moves the current element pointer to the first child of the current element. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. Example The following lines in the “person.xml” file define a person element. <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; char* name; USTATUS xmlStatus; .... 497 The following lines in an RHLI program check the “person.xml” file or wellformedness and parse it. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following lines in an RHLI program move the current element pointer to the first child of current element. If the current element is <person>, the uXMLMoveToFirstChild() function sets the current element pointer to <name>. result = uXMLMoveToFirstChild( xmlObj, &xmlStatus ); See Also uXMLMoveToLastChild, uXMLMoveToNextSibling, uXMLMoveToPreviousSibling, uXMLMoveToParentElement 498 uXMLMoveToLastChild() Moves the current element pointer to the last child Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLMoveToLastChild ( xmlObj, xmlStatus ) xml_obj* xmlObj; USTATUS xmlStatus; Arguments xmlObj DOM object reference xmlStatus Returned status of the function Description Moves the current element pointer to the last child of the current element. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. Example The following lines in the “person.xml” file define a person element. <person> 499 <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; char* name; USTATUS xmlStatus; .... The following lines in an RHLI program check the “person.xml” file for wellformedness and parse it. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following line in an RHLI program moves the current element pointer to the last child of current element. If the current element is <person>, the uXMLMoveToLastChild() function sets the current element pointer to <me>. result = uXMLMoveToLastChild( xmlObj, &xmlStatus ); See Also uXMLMoveToNextSibling,uXMLMoveToPreviousSibling, uXMLMoveToFirstChild, uXMLMoveToParentElement 500 uXMLMoveToNextSibling() Moves the current element pointer to next sibling Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLMoveToNextSibling ( xmlObj, xmlStatus ) xml_obj* xmlObj; USTATUS xmlStatus; Arguments xmlObj DOM object reference xmlStatus Returned status of the function Description Moves the current element pointer to the next sibling of the current element. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. Example The following lines in the “person.xml” file define a person element. <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; char* name; USTATUS xmlStatus; .... 501 The following lines in an RHLI program checks the “person.xml” file for wellformedness, parses it, and prints the result status. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following line in an RHLI program moves the current element pointer to the next sibling of the current element. If the current element is <name>, the uXMLMoveToNextSibling() function sets the current element pointer to <age>. result = uXMLMoveToNextSibling( xmlObj, &xmlStatus ); See Also uXMLMoveToLastChild, uXMLMoveToPreviousSibling, uXMLMoveToFirstChild, uXMLMoveToParentElement 502 uXMLMoveToParentElement() Moves the current element pointer to the parent Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLMoveToParentElement ( xmlObj, xmlStatus ) xml_obj* xmlObj; USTATUS xmlStatus; Arguments xmlObj DOM object reference xmlStatus Returned status of the function Description Moves the current element pointer to the parent of the current element. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. For a list of the most commonly occurring function status values, see the status code at beginning of this document. For a complete list of status returns and the associated error messages with their explanations and corrections, see Unify DataServer: Error messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. 503 Example The following lines in the “person.xml” file define a person element. <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; char* name; USTATUS xmlStatus; .... The following lines in an RHLI program check the “person.xml” file for wellformedness, parse it, and print the result. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following line in an RHLI program moves the current element pointer to the parent element of current element. If the current element is <me> the uXMLMoveToParentElement() function sets the current element pointer to <person>. result = uXMLMoveToParentElement( xmlObj, &xmlStatus ); See Also uXMLMoveToLastChild, uXMLMoveToNextSibling, uXMLMoveToFirstChild, uXMLMoveToPreviousSibling 504 uXMLMoveToPreviousSibling() Moves the current element pointer to the previous sibling Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLMoveToPreviousSibling ( xmlObj, xmlStatus ) xml_obj* xmlObj; USTATUS xmlStatus; Arguments xmlObj DOM object reference xmlStatus Returned status of the function Description Moves the current element pointer to the pervious sibling of the current element. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. Example The following lines in the “person.xml” file define a person element. <person> 505 <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; char* name; USTATUS xmlStatus; .... The following lines in an RHLI program check the “person.xml” file for wellformedness and parse it. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following line in an RHLI program moves the current element pointer to the previous sibling of the current element. If the current element is <me>, the uXMLMoveToPreviousSibling() function sets the current element pointer to <age>. result = uXMLMoveToPreviousSibling( xmlObj, &xmlStatus ); See Also uXMLMoveToLastChild, uXMLMoveToNextSibling, uXMLMoveToFirstChild, uXMLMoveToParentElement 506 uXMLMoveToRootElement() Moves the current element pointer to the root element Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLMoveToRootElement ( xmlObj, xmlStatus ) xml_obj* xmlObj; USTATUS xmlStatus; Arguments xmlObj DOM object reference xmlStatus Returned status of the function Description Moves the current element cursor to the root element of the XML. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. Example The following lines in the “person.xml” file define a person element. <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> me>John</me> </person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; char* name; USTATUS xmlStatus; .... 507 The following lines in an RHLI program check the “person.xml” file for wellformedness and parse it. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following lines in an RHLI program move the current element pointer to the root element of the XML file. If the root element is <person>, the uXMLMoveToRootElement() function sets the current element pointer to <person>. result = uXMLMoveToRootElement( xmlObj, &xmlStatus ); 508 uXMLParse() Parses an XML file Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLParse ( xmlObj, xmlStatus, gXmlFile ) xml_obj* xmlObj; USTATUS xmlStatus; const char* gXmlfile; Arguments Description xmlObj DOM object reference xmlStatus Returned status of the function gXmlfile XML file name and absolute path Parses the specified XML file (gXmlfile) and checks it for wellformedness. The return value is 1 for success and 0 for failure. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. Example The following lines in the “person.xml” file define a person element. <person> <name surname=”anderson” fathername=” >Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> 509 The following lines in an RHLI program check the “person.xml” file for wellformedness and parse it: const char* gXmlfile=”person.xml”; xml_obj* xmlObj; int result = 0; USTATUS xmlStatus; .... result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); 510 uXMLSetAttributeValue() Sets an attribute value Syntax #include <include/rhli.h> #include <include/rhlierr.h> #include <include/uXML.h> int uXMLSetAttributeValue ( xmlObj, xmlStatus, attributeName, attributeValue) xml_obj* xmlObj; USTATUS xmlStatus; const char* attributeName; const char* attributeValue; Arguments xmlObj DOM object reference xmlStatus Returned status of the function attributeName Name of the attribute whose value you want to change attributeValue New value of the attribute Description Sets the attribute value (attributeValue) of the specified attribute (attributeName) of the current element. Status Values For a list of the most commonly occurring function status values, see page 54. For a complete list of status returns and the associated error messages with their explanations and corrections, Unify DataServer: Error Messages. Additional information, if any, can be obtained by calling uXMLGetErrorInfo(). Security No privileges needed. 511 Example The following lines in the “person.xml” file define a person element. <person> <name surname=”anderson” fathername=”mike”>Jack</name> <age dob=”12–12–1980” birth_place=”Sacramento”>20</age> <me>John</me> </person> const char* gXmlFile=”person.xml”; xml_obj* xmlObj; int result = 0; const char* attributeName= “fathername”; const char* attributeValue= “Benjamin Martin”; USTATUS xmlStatus; .... The following lines in an RHLI program check the “person.xml” file for wellformedness, parse it, and print the result. result = uXMLParse( &xmlObj, &xmlStatus, gXmlFile ); printf( ”\n Parsed file result = %d:\n”, result ); .... The following line in an RHLI program modifies the value of the attribute of the current element. If the current element pointer is set to <name> and the uXMLSetAttributeValue() function is called with attributeName=”surname” and attribute value=”Sanderson”, the XML becomes: <_> <name surname=”Sanderson” fathername=”mike”>Joseph</name> <_> result =uXMLSetAttributeValue(xmlObj,&xmlStatus,attributeName, attributeValue ); See Also uXMLDeleteAttribute 512 Appendixes 513 514 Appendix A: Examples This appendix contains complete and partial examples related to the following areas: Transaction handling Scanning An ordered access Name binding ID mapping Transaction journalling DDL Text conversion XML All the RHLI functions used are listed before the examples. Additional Help For information about the structures used in the examples, see “The RHLI Structures” on page 75. 515 Transaction Handling This example performs various database transactions by using the following RHLI functions: ualltx, ubegtx, uclsdb, ucmttx, ufchmsg, uinftx, uinimsg, ulogmsg, and uopndb. #include <rhli.h> #include <rhlierr.h> #define n_tx 10 main(argc, argv) int argc; char **argv; { int ntx = 2; USTATUS status; USTATUS fstatus; USTATUS statusl[n_tx]; USTATUS lstats; UDBID dbid; char *dbname = ”file.db”; UTXID txid, *txl; UTXOPT txopt; UTXID *txidl; UTXINF txinfl[n_tx]; char *errmsg; /* initialize the logging facilities */ if ( uinimsg(argv[0], &status) != USUCCESS) { printf (”error logging messages; ”); if ((errmsg = ufchmsg(status, &fstatus)) == (char *) 0) { printf(”error %ld fetching message \n”, fstatus); uexit(1); } else { printf(”error is %s \n”, errmsg); uexit(1); }; } 516 Appendix A: Examples /* open the database */ if ( uopndb ( dbname, DB_NORMAL, &dbid, &status) != USUCCESS ) { printf (”unable to open the database; ”); if ((errmsg = ufchmsg(status, &fstatus)) == (char *) 0) { printf(”error %ld fetching message \n”, fstatus); uexit(1); } else { printf(”error is %s \n”, errmsg); }; if (ulogmsg(”uopndb”, status, ”Opening database DB_NORMAL”, &lstats) != USUCCESS) { printf (”error logging message; ”); if ((errmsg = ufchmsg(lstats, &fstatus)) == (char *) 0) { printf(”error %ld fetching message \n”, fstatus); uexit(1); } else { printf(”error is %s \n”, errmsg); uexit(1); }; } } else printf (”Database is open.\n”); /* begin a transaction */ txopt.txknd = UTXTYPE2; txopt.waitflg = UWAIT; if ( ubegtx (dbid, UROOTTXID, &txid, &txopt, &status) != USUCCESS ) { printf (”unable to begin a transaction; ”); if ((errmsg = ufchmsg(status, &fstatus)) == (char *) 0) { printf(”error %ld fetching message \n”, fstatus); uexit(1); } else { printf(”error is %s \n”, errmsg); uexit(1); }; uexit(1); } else printf (”Transaction has begun. The transaction ID is: %d\n”, txid); Appendix A: Examples 517 /* retrieve information about the current transaction */ if (ualltx(dbid, UNULLPID, &ntx, &txidl, &status) != USUCCESS) { printf (”unable to get tx list; ”); if ((errmsg = ufchmsg(status, &fstatus)) == (char *) 0) { printf(”error %ld fetching message \n”, fstatus); uexit(1); } else { printf(”error is %s \n”, errmsg); uexit(1); }; uexit(1); } printf (”%d\n”, ntx); printf (”%d\n”, txidl[1]); printf (”%d\n”, txidl[0]); if (uinftx(dbid, ntx, txidl, UCLASS1, txinfl, statusl, &status) != USUCCESS) { printf (”unable to get tx information; ”); if ((errmsg = ufchmsg(status, &fstatus)) == (char *) 0) { printf(”error %ld fetching message \n”, fstatus); free((UTXID *) txidl); uexit(1); } else { printf(”error is %s \n”, errmsg); free((UTXID *) txidl); uexit(1); }; } else printf (”The state of the first transaction is : %d\n”, txinfl[0].state); free((UTXID *) txidl); 518 Appendix A: Examples /* end the transaction */ if ( ucmttx (txid, &status) != USUCCESS ) { printf (”unable to end transaction, ”); if ((errmsg = ufchmsg(status, &fstatus)) == (char *) 0) { printf(”error %ld fetching message \n”, fstatus); uexit(1); } else { printf(”error is %s \n”, errmsg); uexit(1); }; } else printf (”Transaction is committed.\n”); /* close the database */ if ( uclsdb(dbid, DB_NORMAL, &status) != USUCCESS ) { printf (”unable to close the database; ”); if ((errmsg = ufchmsg(status, &fstatus)) == (char *) 0) { printf(”error %ld fetching message \n”, fstatus); uexit(1); } else { printf(”error is %s \n”, errmsg); uexit(1); }; } else printf (”Database is closed.\n”); } Appendix A: Examples 519 Scanning This example uses the following RHLI functions: ualcscn, ubegscn, ubegtx, uclsdb, ucmttx, uendscn, ufchmsg, ufchscn, ufrescn, uinfacs, uinfscn, uinsand, uinsor, uinsprj, uinssrt, uinstbl, and uopndb. This example performs the following scan: select nv_number,nv_description where nv_number between 5000 and 5100 and (nv_description=”Bookcase, Steel” or nv_description=”Bookends, oak”); where a1 = 5000, a2 = 5100, b1 = ”Bookcase, Steel”, and b2 = ”Bookends, oak” #include <rhli.h> #include <rhlierr.h> #include <stdio.h> #define N_TABLE1_COL 6 UDBID dbid; UTXID txid; USTATUS status; USTATUS statusl[N_TABLE1_COL]; UTXOPT txopt; URID rid; char *scaninf; int scanid; int prevlck; UDBVAL search_dbv[5]; UDBVAL *prjvall; UDBVAL *nulldbv; UACSINF *acsinf; UTID tid = $SQL_books.nvntry.$; UCID cidl[] = {$SQL_books.nvntry.nv_number$, $SQL_books.nvntry.nv_description$}; long a1,a2; char b1[15],b2[15]; main () { /* initialization */ init_dbvall(); txopt.txknd = UTXTYPE2; txopt.waitflg = UWAIT; 520 Appendix A: Examples /* open the database */ if (uopndb (””,DB_NORMAL,&dbid,&status) != USUCCESS) fatal (”uopndb”); /* begin a transaction */ if (ubegtx (dbid,UROOTTXID,&txid,&txopt,&status) != USUCCESS) fatal (”ubegtx”); /* perform scan */ if (ualcscn (txid,0,0,0,&scaninf,&status) != USUCCESS) fatal (”ualcscn”); if (uinstbl (txid,scaninf,tid,&status) != USUCCESS) fatal (”uinstbl”); if (uinsprj (txid,scaninf,2,cidl,statusl,&status) != USUCCESS) fatal (”uinsprj”); init_criteria(); if (uinsand (txid,scaninf,4,&status) != USUCCESS) fatal (”uinsand”); if (uinsor (txid,scaninf,$SQL_books.nvntry.nv_number$,URNGCC, &search_dbv[0],&search_dbv[1], &status) != USUCCESS) fatal (”uinsor”); if (uinsor(txid,scaninf,$SQL_books.nvntry.nv_description$,ULIKE, &search_dbv[2],nulldbv,&status) != USUCCESS) fatal (”uinsor”); if (uinsor(txid,scaninf,$SQL_books.nvntry.nv_description$,ULIKE, &search_dbv[3], nulldbv,&status) != USUCCESS) fatal (”uinsor”); if (uinfscn (txid,scaninf,&prjvall,&status) != USUCCESS) fatal (”uinfscn”); /* define sort criteria */ if (uinssrt (txid, scaninf, $SQL_books.nvntry.nv_description$, DB_ASCEND, &status) != USUCCESS) fatal (”uinssrt”); /* begin scan */ if (ubegscn (txid,USLCK,scaninf,&scanid,&status) != USUCCESS) fatal (”ubegscn”); Appendix A: Examples 521 /* retrieve access method type */ if (uinfacs (scanid, &acsinf, &status) != USUCCESS) fatal (”uinfacs”); else printf (”Scan using : ”); switch(acsinf.acsknd) { case USQSCN: printf (”sequential access.\n”); break; case UBTSCN: printf (”btree access.\n”); break; case UDKSCN: printf (”direct access.\n”); break; case UHKSCN: printf (”hash–key access.\n”); break; case ULKSCN: printf (”link access.\n”); break; case UNULLSCN: printf (”scan not initialized.\n”); break; default: printf (”undefined return from infacs.\n”); } while (ufchscn(scanid,&rid,prjvall,&prevlck,statusl,&status) == USUCCESS) printf (”%d | %s \n”, *(prjvall[0].unip.hintp), prjvall[1].unip.strp); if (status != UEEOSCN) fatal (”ufchscn”); if (uendscn (scanid,&status) != USUCCESS) fatal (”uendscn”); if (ufrescn (txid,&scaninf,DB_RESET,&status) != USUCCESS) fatal (”ufrescn”); if (ucmttx (txid,&status) != USUCCESS) fatal (”ucmttx”); if (uclsdb (dbid,DB_NORMAL,&status) != USUCCESS) fatal (”uclsdb”); uexit (0); } 522 Appendix A: Examples /* Since these buffers probably only need to be set up once, no matter how often you do the query, it is best to initialize them in some function once, and if possible call it only once. This will save runtime CPU time. */ init_dbvall() { search_dbv[0].unip.hintp = &a1; search_dbv[0].valopts = UNORMAL; search_dbv[1].unip.hintp = &a2; search_dbv[1].valopts = UNORMAL; search_dbv[2].unip.strp = b1; search_dbv[2].valopts = UNORMAL; search_dbv[3].unip.strp = b2; search_dbv[3].valopts = UNORMAL nulldbv = (UDBVAL *) 0L; } /* The actual query is set up in terms of variables, not constants.So the query reads: select nv_number,nv_description where nv_number between a1 and a2 and (nv_description = b1 or nv_description = b2) ; */ init_criteria() { a1 = 5000; a2 = 5100; strcpy (b1,”Bookcase, Steel”); strcpy (b2,”Bookends, oak”); } /* A simple error routine */ fatal (fn) char *fn; { char *errmsg; USTATUS fchstats; printf (”An error occurred in %s;”, fn); if ((errmsg = ufchmsg(status, &fchstats)) == (char *) 0) printf (”error fetching message: %ld\n”, fchstats); else printf (”error is %s \n”, errmsg); uexit(99); } Appendix A: Examples 523 An Ordered Access The following example uses all ordered access functions ufchord, uendord, uposord, and uskrord. ubegtx, uclsdb, ucmttx, and uopndb are also called. #include #include <rhli.h> <rhlierr.h> main(argc, argv) int argc; char **argv; { /* define required parameters */ UTXID txid; /* Transaction ID UTID tid; /* Table ID URID rid; /* Row ID UOACCID oaccid; /* Order Access ID UOACOLS UNIPEXT UNIPCOLL UCID UDBVAL UTP_HINT */ URID UDBVAL USTATUS USTATUS ord_list[2]; /* column sort order list unipext[3]; unipcoll; cidl[3]; /* list of Column IDs dbvall[3]; /* list of column values maxsalesrep = 0; /* maximum company salesrep */ svrid = 0; keyvall[2]; statusl[3]; status; */ /* Define char char ETP_HINT external and internal data types */ extaddress2[30]; extaddress1[30]; extsalesrep; char char UTP_HINT intaddress2[30]; intaddress1[30]; intsalesrep; char UTXOPT UDBID /* save area for Row ID */ */ /* status value list */ /* status returned by RHLI functions */ *dbname; txopt; dbid; /* Set up for ordered ord_list[0].oacid = ord_list[0].oacopts = ord_list[1].oacid = ord_list[1].oacopts = 524 */ */ */ */ access on the address */ $SQL_books.company.co_address_2$; DB_ASCEND; $SQL_books.company.co_address_1$; DB_ASCEND; Appendix A: Examples dbname = ”file.db”; if ( uopndb ( dbname, DB_NORMAL, &dbid, &status) != USUCCESS ) { printf (”unable to open the database; ”); uexit(1); } else printf (”Database is open.\n”); /* begin a transaction */ txopt.txknd = UTXTYPE2; txopt.waitflg = UWAIT; if (ubegtx (dbid, UROOTTXID, &txid, &txopt, &status) != USUCCESS ) { printf (”unable to begin a transaction; ”); uexit(1); } else printf (”Transaction has begun. The transaction ID is:%ld\n”, txid); /* Begin ordered access on table company */ if ( ubegord (txid, $SQL_books.company.$, 2, ord_list, &oaccid, statusl, &status) != USUCCESS ) { printf(”ubegord failed. \n”); uexit(1); } /* Position to second half of the alphabet */ keyvall[0].unip.strp = ”M”; keyvall[1].unip.strp = ”A”; if ( uposord (txid, oaccid, UGETST, keyvall, USLCK, &rid, statusl, &status) != USUCCESS ) { printf(”uposord failed. \n”); uexit(1); } /* Initialize UNIPCOLL structure components */ unipcoll.ncol = 3; unipcoll.cidl = cidl; unipcoll.dbvall = dbvall; /* Initialize (at compile–time) Column IDs list */ cidl[0] = $SQL_books.company.co_address_2$; cidl[1] = $SQL_books.company.co_address_1$; cidl[2] = $SQL_books.company.co_sales_rep$; Appendix A: Examples 525 /* Define the internal–format buffer addresses */ dbvall[0].unip.strp = intaddress2; dbvall[1].unip.strp = intaddress1; dbvall[2].unip.hintp = &intsalesrep; dbvall[0].valopts = UNORMAL; dbvall[1].valopts = UNORMAL; dbvall[2].valopts = UNORMAL; /* Read each row looking for highest salesrep */ while ( ufchord(txid, oaccid, UNEXT, USLCK, &rid, &unipcoll, statusl, &status) != UFAILURE ) { if ( intsalesrep > maxsalesrep ) { maxsalesrep = intsalesrep; svrid = rid; } /* Reset valopts for next fetch */ unipcoll.dbvall[0].valopts = UNORMAL; unipcoll.dbvall[1].valopts = UNORMAL; unipcoll.dbvall[2].valopts = UNORMAL; } /* Error if other than end of file was returned */ if ( status != UEEOSCN ) { printf(”ufchord failed. \n”); uexit(1); } /* Seek row with highest salesrep */ if ( uskrord (txid, oaccid, USLCK, svrid, &status) != USUCCESS ) { printf(”uskrord failed. \n”); uexit(1); } /* Now fetch the address of the company and its salesrep */ if ( ufchord(txid, oaccid, USAME, &rid, &unipcoll, USLCK, statusl, &status) != USUCCESS ) { printf(”ufchord failed. \n”); uexit(1); } /* Define the external–format buffer addresses */ unipext[0].univ.strp = extaddress2; unipext[1].univ.strp = extaddress1; unipext[2].univ.hintp = &extsalesrep; /* Display external–format data */ printf(”Highest salesrep in M through Z locates at:\n %s\n%s\n”, extaddress1, extaddress2); printf(” with a salesrep of %ld\n”, extsalesrep); 526 Appendix A: Examples /* End the ordered access and release resources */ if ( uendord (oaccid, &status) != USUCCESS ) { printf(”uendord failed \n”); uexit(1); } /* end the transaction */ if ( ucmttx (txid, &status) != USUCCESS ) { printf (”unable to end transaction \n ”); uexit(1); } else printf (”Transaction is committed.\n”); /* close the database */ if ( uclsdb(dbid, DB_NORMAL, &status) != USUCCESS ) { printf (”unable to close the database \n ”); uexit(1); } else printf (”Database is closed.\n”); } Appendix A: Examples 527 Name Binding This example performs name binding with the following function calls involved: ubegtx, ubndath, ubndtbl, uclsdb, ucmttx, ufchath, ufchmsg, ulogmsg, and uopndb. #include <rhli.h> #include <rhlierr.h> main(argc, argv) int argc; char **argv; { USTATUS status; USTATUS fstatus; USTATUS statusl[3]; USTATUS lstats; UDBID dbid; char *dbname = ”file.db”; UTXID txid; UTXID *txl; UTXOPT txopt; UTXID *txidl; UTXINF *txinfl; int nauth; char *authnml; int ntbl = 3; int i; char * tblnml[3]; UTID *tidl[3]; UAID aid; char * db_name; DBUSER *dbuser; int nuser; char *errmsg; 528 Appendix A: Examples /* open the database */ if ( uopndb ( dbname, DB_NORMAL, &dbid, &status) != USUCCESS ) { printf (”unable to open the database; ”); if ((errmsg = ufchmsg(status, &fstatus)) == (char *) 0) { printf(”error %ld fetching message \n”, fstatus); uexit(1); } else { printf(”error is %s \n”, errmsg); uexit(1); }; if (ulogmsg(”uopndb”, status, ”Opening database DB_NORMAL”, &lstats) != USUCCESS) { printf (”error logging message; ”); if ((errmsg = ufchmsg(lstats, &fstatus)) == (char *) 0) { printf(”error %ld fetching message \n”, fstatus); uexit(1); } else { printf(”error is %s \n”, errmsg); uexit(1); }; } } else printf (”Database is open.\n”); /* begin a transaction */ txopt.txknd = UTXTYPE2; txopt.waitflg = UWAIT; if ( ubegtx (dbid, UROOTTXID, &txid, &txopt, &status) != USUCCESS ) { printf (”unable to begin a transaction; ”); if ((errmsg = ufchmsg(status, &fstatus)) == (char *) 0) { printf(”error %ld fetching message \n”, fstatus); uexit(1); } else { printf(”error is %s \n”, errmsg); uexit(1); }; } else printf (”Transaction has begun. The transaction ID is:%d\n”, txid); Appendix A: Examples 529 authnml = ”SQL_books”; if ( ubndath(txid, 1, &authnml, USLCK, &aid, statusl, &status) != USUCCESS ) { printf(”Unable to get Auth.ID by name: status = %ld\n”, status); uexit(1); } /* aid = UNULLAID; */ /* set schema to SQL_books */ if (ufchath(txid, aid, &status) != USUCCESS) { printf(”Unable to change current schema.\n”); uexit(1); } /* bind the table names to IDs */ tblnml[0] = ”company”; tblnml[1] = ”orders”; tblnml[2] = ”nvntry”; if (ubndtbl(txid, aid, ntbl, tblnml, USLCK, tidl, statusl, &status) != USUCCESS) { printf (”Unable to get tables ids; ”); if ((errmsg = ufchmsg(status, &fstatus)) == (char *) 0) { printf(”error %ld fetching message \n”, fstatus); uexit(1); } else { printf(”error is %s \n”, errmsg); uexit(1); }; } for (i = 0; i < ntbl; ++i) printf (”The table ID is : %ld\n”, tidl[i]); /* end the transaction */ if ( ucmttx (txid, &status) != USUCCESS ) { printf (”unable to end transaction, ”); if ((errmsg = ufchmsg(status, &fstatus)) == (char *) 0) { printf(”error %ld fetching message \n”, fstatus); uexit(1); } else { printf(”error is %s \n”, errmsg); uexit(1); }; } else printf (”Transaction is committed.\n”); 530 Appendix A: Examples /* close the database */ if ( uclsdb(dbid, DB_NORMAL, &status) != USUCCESS ) { printf (”unable to close the database; ”); if ((errmsg = ufchmsg(status, &fstatus)) == (char *) 0) { printf(”error %ld fetching message \n”, fstatus); uexit(1); } else { printf(”error is %s \n”, errmsg); uexit(1); }; } else printf (”Database is closed.\n”); } Appendix A: Examples 531 ID Mapping The following example uses the functions umapath, umapcol, and umaptbl to map compile-time IDs to their runtime equivalents. #define NATH 4 /* # of schemas in the database */ /* define required parameters */ USTATUS statusl[NATH]; /* schema status list */ USTATUS status; /* function status */ UDBID dbid; /* database ID */ UATHMAP uathmap[NATH]; /* schema mapping structure */ ... /* re-map compile-time Authorization IDs –> runtime IDs */ INIATHMAP(manf_schema, aidl[1], &uathmap[0]); INIATHMAP(model_schema, aidl[2], &uathmap[1]); INIATHMAP(item_schema, aidl[3], &uathmap[2]); if (umapath(dbid, 3, uathmap, statusl, &status) != USUCCESS) { printf(”Auth. ID re-mapping error: status=%ld\n”, status); uexit(99); } The following example uses the umapcol function to map a compile-time ID to its runtime equivalent. #define NCOL 3 #define NCOLMAP 3 #define NTBLMAP 2 /* define required parameters */ UTXID txid; /* Transaction ID */ USTATUS status; /* function status */ ... /* define umapcok requirements */ UCOLMAP colmapl[NCOLMAP]; /* column mapping structure */ USTATUS cmapstl[NCOLMAP]; /* column status value list */ ... /* re-map compile-time Column IDs –> runtime IDs */ INICOLMAP(sales.revenues.cust_name, $(UCOL:CID)sales.revenues.cust_name$, &colmapl[0]); INICOLMAP(sales.revenues.cust_id, $(UCOL:CID)sales.revenues.cust_id$, &colmapl[1]); INICOLMAP(sales.revenues.sale_date, $(UCOL:CID)sales.revenues.sale_date$, &colmapl[2]); 532 Appendix A: Examples if (umapcol(dbid, NCOLMAP, colmapl, cmapstl, &status) != USUCCESS) { printf(“Unable to re-map Column ID: status = %ld\n”, status); uexit(101); } /* re-map compile-time Table IDs –> runtime IDs */ INITBLMAP(sales.commissions, $(UTBL:TID)sales.commissions$, &tblmapl[0]); INITBLMAP(sales.revenues, $(UTBL:TID)sales.revenues$, &tblmapl[1]); if (umaptbl(dbid, NTBLMAP, tblmapl, tmapstl, &status) != USUCCESS) { printf(“Unable to re-map Table ID: status = %ld\n”, status); uexit(101); } Appendix A: Examples 533 Transaction Journaling This example uses the journaling functions uclsjrn, ufchjrn, and uopnjrn to manipulate the journal file. /* define required parameters */ char *dbname = “”; /* use DBPATH */ int jrnfd; /* returned journal file descriptor */ long seqn; /* sequence number */ UJRNINF jrninfo; /* journal oper info structure */ USTATUS status; /* status returned by uxxxjrn */ USTATUS fchstat; /* status returned by ufchmsg */ /* open journal file */ if (uopnjrn(dbname, “/usr/user/linda/db/file.jn”, &jrnfd, &seqn, &status) != USUCCESS) { printf ( “uopnjrn failed: \n”); uexit( 1 ); } /* fetch journal information */ while ( ufchjrn ( jrnfd, &jrninfo, &status ) == USUCCESS) { switch ( jrninfo.opknd ) { case UINSOPR: printf (“Insert to table %ld by user %ld\n”,jrninfo.tid, jrninfo.ids.uid ); free ( jrninfo.txinfo.insert.newucol ); break; case UUPDOPR: printf (“Update to table %ld by user%ld\n”, jrninfo.tid,jrninfo.ids.uid ); free ( jrninfo.txinfo.update.newucol ); free ( jrninfo.txinfo.update.olducol ); break; case UDELOPR: printf (“Delete to table %ld by user,%ld\n”, jrninfo.tid, jrninfo.ids.uid ); free ( jrninfo.txinfo.delete.olducol ); break; } } 534 Appendix A: Examples if ( status != UENORM && status != UEJNEOTP ) { printf (”ufchjrn() failed: \n”); uexit( 1 ); } /* close journal files */ if ( uclsjrn(jrnfd, &status) == UFAILURE ) { printf (“uclsjrn() failed: \n”); uexit( 1 ); } Appendix A: Examples 535 DDL This example creates a table named member. The RHLI functions used are uaddtbl, ubegtx, ubndbt, uclsdb, ucmttx, udrpbt, and uopndb. #include <rhli.h> #include <rhlierr.h> UDBID dbid; UTXID txid; USTATUS status; USTATUS statusl[4]; UTXOPT txopt; DBCOL dbcoll[4]; DBTABLE dbtablel[1]; int i; main(argc, argv) int argc; char **argv; { /* open the database */ if (uopndb ((char *) 0,DB_NORMAL,&dbid,&status) != USUCCESS) fatal (”uopndb”, status); printf(”Database opened.\n”); /* begin a transaction */ txopt.txknd = UTXTYPE2; txopt.waitflg = UWAIT; if (ubegtx (dbid,UROOTTXID,&txid,&txopt,&status) != USUCCESS) fatal (”ubegtx”, status); printf(”Transaction begun.\n”); /* add a table */ dbcoll[0].coltyp = U_INT; dbcoll[0].collen = UIL_INT; dbcoll[0].colscl = 0; dbcoll[0].dsplen = UDL_INT; dbcoll[0].dspscl = 0; dbcoll[0].colopts = DB_NORMAL; dbcoll[0].colnm = ”mem_id”; dbcoll[0].coldesc = (char *) 0; dbcoll[0].dsppict = (char *) 0; dbcoll[1].coltyp = U_STR; dbcoll[1].collen = 20; 536 Appendix A: Examples dbcoll[1].colscl = 0; dbcoll[1].dsplen = 20; dbcoll[1].dspscl = 0; dbcoll[1].colopts = DB_NONULL; dbcoll[1].colnm = ”mem_name”; dbcoll[1].coldesc = (char *) 0; dbcoll[1].dsppict = (char *) 0; dbcoll[2].coltyp = U_AMT; dbcoll[2].collen = 9; dbcoll[2].colscl = 2; dbcoll[2].dsplen = 9; dbcoll[2].dspscl = 0; dbcoll[2].colopts = DB_NORMAL; dbcoll[2].colnm = ”mem_sal”; dbcoll[2].coldesc = (char *) 0; dbcoll[2].dsppict = (char *) 0; dbcoll[3].coltyp = U_HDATE; dbcoll[3].collen = 0; dbcoll[3].colscl = 0; dbcoll[3].dsplen = 10; dbcoll[3].dspscl = 0; dbcoll[3].colopts = DB_NONULL; dbcoll[3].colnm = ”mem_date”; dbcoll[3].coldesc = (char *) 0; dbcoll[3].dsppict = (char *) 0; dbtablel[0].dbcol = dbcoll; dbtablel[0].vid = UNULLVID; dbtablel[0].segsz = 0L; dbtablel[0].expnum = 1000; dbtablel[0].statusl = statusl; dbtablel[0].tblopts = DB_NORMAL; dbtablel[0].ncol = 4; dbtablel[0].tblnm = ”member”; dbtablel[0].tbldesc = (char *) 0; if (uaddtbl(txid, 1, dbtablel, statusl, &status) != USUCCESS) { for (i=0; i<4; i++) printf(”Column %d status: %ld\n”, i, statusl[i]); fatal (”uaddtbl”, status); } printf(”Table added. \n”); /* commit the transaction and close the database */ if (ucmttx (txid, &status) != USUCCESS) fatal (”ucmttx”, status); printf(”Transaction committed.\n”); if (uclsdb (dbid,DB_NORMAL, &status) != USUCCESS) fatal (”uclsdb”, status); printf(”Database closed.\n”); uexit (0); } Appendix A: Examples 537 /* A simple error routine that prints out the function name and the errorstatus of the last RHLI call that failed, and exit */ fatal (fctname, status) char * fctname; USTATUS status; { int i; printf (”Fatal error on %s;”, fctname); printf (”status is %d\n”, status); uexit(1); } The following example drops two B-tree indexes. /* allocate storage for the B-tree and status list */ bidl = (UBTID *)calloc(2, sizeof(UBTID)); statusl = (USTATUS *)calloc(2, sizeof(USTATUS)); ... /* bind the B-tree named ’part_number’ */ ubndbt(txid, $parts$, 1, ”part_number”, USLCK, &bidl[0], &statusl[0], &status); ... /* bind to the B-tree named ’ord_number’ */ ubndbt(txid, $parts$, 1, ”ord_number”, USLCK, &bidl[1], &statusl[1], &status); ... /* drop both B-trees */ if ( udrpbt(txid, 2, bidl, statusl, &status) != USUCCES) { fprintf(stderr, ”B-trees could not be dropped: status=%ld\n”, status); } 538 Appendix A: Examples Text Conversion This example uses the uextint function to convert data values before they are inserted into a row by calling uinsrow. #define NCOL 3 /* define external data types ETP_INT extpnum; /* ETP_STRP extpnam[80]; /* ETP_DATE extpdat; /* */ external–format part number */ external–format part name */ external–format part date */ /* define internal data types UTP_INT intpnum; /* char intpnam[80]; /* UTP_DATE intpdat; /* */ internal–format part number */ internal–format part name */ internal–format part date */ /* define RHLI interface required parameters */ URID rid; USTATUS status; USTATUS statusl[NCOL]; UTXID UNIPCOLL UNIPEXT UCID UDBVAL txid; unipcoll; unipext[NCOL]; cidl[NCOL]; dbvall[NCOL]; /* initialize UNIPCOLL structure components */ unipcoll.ncol = NCOL; unipcoll.cidl = cidl; unipcoll.dbvall = dbvall; /* initialize (at compile–time) Column IDs list */ cidl[0] = $parts.part_name$; cidl[1] = $parts.part_number$; cidl[2] = $parts.part_date$; /* define the internal–format buffer addresses */ dbvall[0].unip.strp = intpnam; dbvall[1].unip.intp = &intpnum; dbvall[2].unip.datep = &intpdat; /* define the external–format buffer address */ unipext[0].univ.strp = extpnam; unipext[0].valopts = UNORMAL; Appendix A: Examples 539 unipext[1].univ.intp = &extpnum; unipext[1].valopts = UNORMAL; unipext[2].univ.datep = &extpdat; unipext[2].valopts = UNORMAL; /* define the external data values to be converted to internal format */ sprintf(extpnam, ”Ronnieco %s”, ”Wedg–o–matic”); extpnum = 1995; extpdat.month = 12; /* manufacture month */ extpdat.day = 31; /* manufacture day */ extpdat.year = 1987; /* manufacture year */ /* this assumes the database is open & a transaction begun */ if (uextint(txid, $parts.$, unipext, &unipcoll, statusl, &status) != USUCCESS) { printf(”data conversion failed: %ld\n”, status); uexit(100); } /* insert the data into the database */ if (uinsrow(txid, $parts.$, &rid, &unipcoll, statusl, &status) != USUCCESS) { printf(”data insertion failed: %ld\n”, status); uexit(101); } The next example calls ufchrow to retrieve a row from a database table, and uses the uintext function to convert data from internal format to external format. #define NCOL 3 /* define external data types */ ETP_INT extpnum; /* external–format part number */ char extpnam[80]; /* external–format part name */ ETP_DATE extpdat; /* external–format part date */ ... /* define internal data types */ UTP_INT intpnum; /* internal–format part number */ char intpnam[80]; /* internal–format part name */ UTP_DATE intpdat; /* internal–format part date */ ... /* define required parameters */ UTXID txid; URID rid; USTATUS status; USTATUS statusl[NCOL]; UNIPCOLL unipcoll; UNIPEXT unipext[NCOL]; UCID cidl[NCOL]; UDBVAL dbvall[NCOL]; 540 Appendix A: Examples /* initialize UNIPCOLL structure components */ unipcoll.ncol = NCOL; unipcoll.cidl = cidl; unipcoll.dbvall = dbvall; ... /* initialize (at compile–time) Column IDs list */ cidl[0] = $parts.part_name$; cidl[1] = $parts.part_number$; cidl[2] = $parts.part_date$; ... /* define the internal–format buffer addresses */ dbvall[0].unip.strp = intpnam; dbvall[0].valopts = UNORMAL; dbvall[1].unip.intp = &intpnum; dbvall[1].valopts = UNORMAL; dbvall[2].unip.datep = &intpdat; dbvall[2].valopts = UNORMAL; ... /* define the external–format buffer address */ unipext[0].univ.strp = extpnam; unipext[1].univ.intp = &extpnum; unipext[2].univ.datep = &extpdat; /* retrieve the internal–format data from the database */ if (ufchrow(txid, $parts.$, USLCK, rid, &unipcoll, statusl, &status) != USUCCESS) { printf(”data retrieval failed: %ld\n”, status); uexit(101); } ... /* convert the internal format data to external format */ if (uintext(txid, $parts.$, &unipcoll, &unipext, statusl, &status) != USUCCESS) { printf(”data conversion failed: %ld\n”, status); uexit(100); } /* now display the external–format data */ printf(”part–number: %d\n”, extpnum); printf(”part–name: %s\n”, extpnam); printf(”part–date: %02d/%02d/%02d\n”, extpdat.month, extpdat.day, extpdat.year); Appendix A: Examples 541 XML example This example uses the XML processing RHLI functions to parse an XML file. The XML file could be the result a previous select operation that specified the AS XML clause or it could be from another source. #include <stdio.h> #include <def/globdefs.h> #include <def/rhli.h> #include <def/uXML.h> int main(int argc,char* argv[]) { const char *xmlFile = ”person.xml”; xml_obj xmlObj = 0; int count; char** attributeNames = 0; char** attributeValues = 0; int result = 0; char* attributeValue; char* elementValue; USTATUS xmlStatus; char* elementName = 0; if ( ! uXMLParse(&xmlObj, &xmlStatus, xmlFile) ) { printf(”Error opening or parsing file ’%s’ error code=%d\n”, xmlFile, xmlStatus); exit(1); } 542 Appendix A: Examples if( ! uXMLMoveToRootElement(xmlObj, &xmlStatus) ) { printf(”Error in %s error code=%d\n”, ”uXMLMoveToRootElement”, xmlStatus); uXMLDestroy(&xmlObj, &xmlStatus); exit(1); } /* * The current element cursor now points to root element. */ if ( ! uXMLGetCurrentElementName(xmlObj, &xmlStatus, &elementName) ) { printf(”Error in %s error code=%d\n”, ”uXMLGetCurrentElementName”, xmlStatus); } else { printf(”\n name of root element = %s \n”, elementName); free(elementName); } if ( ! uXMLMoveToFirstChild(xmlObj, &xmlStatus) ) { printf(”Error in %s error code=%d\n”, ”uXMLMoveToFirstChild”, xmlStatus); } /* * The current element cursor now points to first child of root element. */ if ( ! uXMLGetCurrentElementName(xmlObj, &xmlStatus, &elementName) ) { printf(”Error in %s error code=%d\n”, ”uXMLGetCurrentElementName”, xmlStatus); } else { printf(”\n name of first child =%s \n”, elementName); free(elementName); } Appendix A: Examples 543 if ( ! uXMLGetAttributeNames(xmlObj, &xmlStatus, &attributeNames) ) { printf(”Error in %s error code=%d\n”, ”uXMLGetAttributeNames”, xmlStatus); } else { /* * The variable ”attributeNames” contains attribute names of * current element i.e. ”name”, The variable attributeNames will * contain ”surname, fathername” */ char ** name = attributeNames; printf(”\nThe attribute names are:\n”); for( ; *name != NIL(char); ++name ) { printf(”\t%s\n”, *name); free(*name); } free(attributeNames); attributeNames = 0; } if ( ! uXMLGetAttributeValues(xmlObj, &xmlStatus, &attributeValues) ) { printf(”Error in %s error code=%d\n”, ”uXMLGetAttributeValues”, xmlStatus); } else { /* * The variable ”attributeValues” conatins attribute values of * current element i.e. ”name”, the varibale attributeValues * will contain ”anderson, mike”. */ char ** name = attributeValues; printf(”\n the attribute values are:\n”); for( ; *name != NIL(char); ++name ) { printf(”\t%s\n”, *name); free(*name); } free(attributeValues); attributeValues = 0; } 544 Appendix A: Examples /* Get attribute count of an element */ if ( ! uXMLGetAttributeCount(xmlObj, &xmlStatus, &count) ) { printf(”Error in %s error code=%d\n”, ”uXMLGetAttributeCount”, xmlStatus); } else { /* The variable ”count” should contain value 2 */ printf(”\n number of attributes are :%d ”, count); } if ( ! uXMLAddAttribute(xmlObj, &xmlStatus, ”mothername”, ”Maria”) ) { printf(”Error in %s error code=%d\n”, ”uXMLAddAttribute”, xmlStatus); } if ( ! uXMLGetAttributeNames(xmlObj, &xmlStatus, &attributeNames) ) { printf(”Error in %s error code=%d\n”, ”uXMLGetAttributeNames”, xmlStatus); } else { /* * THE VARIABLE ”attributeNames” CONATINS ATTRIBUTE NAMES OF * CURRENT ELEMENT i.e. ”name”, THE VARIBALE ”attributeNames” WILL * CONTAIN ”surname, fathername, mothername” */ char ** name = attributeNames; printf(”\nThe attribute names are:\n”); for( ; *name != NIL(char); ++name ) { printf(”\t%s\n”, *name); free(*name); } free(attributeNames); attributeNames = 0; } if ( ! uXMLMoveToNextSibling(xmlObj, &xmlStatus) ) { printf(”Error in %s error code=%d\n”, ”uXMLMoveToNextSibling”, xmlStatus); } Appendix A: Examples 545 /* THE CURRENT ELEMENT CURSOR NOW POINTS TO NEXT SIBLING i.e.”age” */ if ( ! uXMLGetCurrentElementName(xmlObj, &xmlStatus, &elementName) ) { printf(”Error in %s error code=%d\n”, ”uXMLGetCurrentElementName”, xmlStatus); } else { /* THE TAG NAME OF CURRENT ELEMENT i.e. ”age” */ printf(”\n the next sibling is= %s \n”, elementName); free(elementName); } /* DELETE THE VALUE OF CURRENT ELEMENT i.e. ”age” DELETES THE VALUE OF CURRENT ELEMENT i.e. 20 */ if ( ! uXMLDeleteCurrentElementValue(xmlObj, &xmlStatus) ) { printf(”Error in %s error code=%d\n”, ”uXMLDeleteCurrentElementValue”, xmlStatus); } if ( ! uXMLMoveToParentElement(xmlObj, &xmlStatus) ) { printf(”Error in %s error code=%d\n”, ”uXMLMoveToParentElement”, xmlStatus); } if ( ! uXMLMoveToLastChild(xmlObj, &xmlStatus) ) { printf(”Error in %s error code=%d\n”, ”uXMLMoveToLastChild”, xmlStatus); } if ( ! uXMLGetCurrentElementName(xmlObj, &xmlStatus, &elementName) ) { printf(”Error in %s error code=%d\n”, ”uXMLGetCurrentElementName”, xmlStatus); } else { printf(”\n name of last child = %s \n”, elementName); free(elementName); } /* THE VALUE OF CURRENT ELEMENT i.e. ”me” IS ”john” */ if ( ! uXMLGetValueOfCurrentElement(xmlObj, &xmlStatus, &elementValue) ) { printf(”Error in %s error code=%d\n”, ”uXMLGetValueOfCurrentElement”, xmlStatus); } else { printf(”\n the value of current element is = %s \n”, elementValue); free(elementValue); } 546 Appendix A: Examples /* MODIFY THE VALUE OF CURRENBT ELEMENT i.e. ”me” FROM ”john” TO ”Joseph” */ if ( ! uXMLModifyElementValue(xmlObj, &xmlStatus, ”Joseph”) ) { printf(”Error in %s error code=%d\n”, ”uXMLModifyElementValue”, xmlStatus); } if ( ! uXMLGetValueOfCurrentElement(xmlObj, &xmlStatus, &elementValue) ) { printf(”Error in %s error code=%d\n”, ”uXMLGetValueOfCurrentElement”, xmlStatus); } else { printf(”\n the new value is = %s \n”, elementValue); free(elementValue); } if ( ! uXMLMoveToPreviousSibling(xmlObj, &xmlStatus) ) { printf(”Error in %s error code=%d\n”, ”uXMLMoveToPreviousSibling”, xmlStatus); } if ( ! uXMLGetCurrentElementName(xmlObj, &xmlStatus, &elementName) ) { printf(”Error in %s error code=%d\n”, ”uXMLGetCurrentElementName”, xmlStatus); } else { /* THE CURRENT ELEMENT CURSOR NOW POINTS TO PreviousSIBLING i.e.”age” */ printf(”\n the previous sibling is = %s \n”, elementName); free(elementName); } if ( ! uXMLMoveToNextSibling(xmlObj, &xmlStatus) ) { printf(”Error in %s error code=%d\n”, ”uXMLMoveToNextSibling”, xmlStatus); } if ( ! uXMLMoveToParentElement(xmlObj, &xmlStatus) ) { printf(”Error in %s error code=%d\n”, ”uXMLMoveToParentElement”, xmlStatus); } Appendix A: Examples 547 /* * THE CURRENT ELEMENT CURSOR NOW POINTS TO PARENT ELEMENT * i.e. person */ if ( ! uXMLGetCurrentElementName(xmlObj, &xmlStatus, &elementName) ) { printf(”Error in %s error code=%d\n”, ”uXMLGetCurrentElementName”, xmlStatus); } else { printf(”\n The parent element is = %s \n”, elementName); free(elementName); } if ( ! uXMLMoveToFirstChild(xmlObj, &xmlStatus) ) { printf(”Error in %s error code=%d\n”, ”uXMLMoveToFirstChild”, xmlStatus); } /* THE CURRENT ELEMENT CURSOR IS AT <name> TAG, THE VARIABLE ”attributeValue” will contain the value OF ATTRIBUTE ”fathername” i.e. ”mike” */ if ( ! uXMLGetAttributeValue(xmlObj, &xmlStatus, ”fathername”, &attributeValue) ) { printf(”Error in %s error code=%d\n”, ”uXMLGetAttributeValue”, xmlStatus); } else { printf(”\n the attribute value is = %s \n”, attributeValue); free(attributeValue); } /* TO DELETE AN ATTRIBUTE */ if ( ! uXMLDeleteAttribute(xmlObj, &xmlStatus,”mothername”) ) { printf(”Error in %s error code=%d\n”, ”uXMLDeleteAttribute”, xmlStatus); } if ( ! uXMLDeleteAttribute(xmlObj, &xmlStatus, ”fathername”) ) { printf(”Error in %s error code=%d\n”, ”uXMLDeleteAttribute”, xmlStatus); } 548 Appendix A: Examples if ( ! uXMLGetAttributeNames(xmlObj, &xmlStatus, &attributeNames) ) { printf(”Error in %s error code=%d\n”, ”uXMLGetAttributeNames”, xmlStatus); } else { /* THE VARIABLE ”attributeNames” CONATINS ATTRIBUTE NAMES OF CURRENT ELEMENT i.e. ”name”, THE VARIBALE attributeNames WILL CONTAIN ”surname” */ char ** name = attributeNames; printf(”\nThe attribute names are:\n”); for( ; *name != NIL(char); ++name ) { printf(”\t%s\n”, *name); free(*name); } free(attributeNames); attributeNames = 0; } /* CHECKED TO SET ATTRIBUTE VALUE */ if ( ! uXMLSetAttributeValue(xmlObj, &xmlStatus, ”surname”, ”Sanderson”)) { printf(”Error in %s error code=%d\n”, ”uXMLSetAttributeValue”, xmlStatus); } /* THE NEW ATTRIBUTE VALUE */ if ( ! uXMLGetAttributeValue(xmlObj, &xmlStatus, ”surname”, &attributeValue)) { printf(”Error in %s error code=%d\n”, ”uXMLGetAttributeValue”, xmlStatus); } else { printf(”\n the attribute value is = %s \n”, attributeValue); free(attributeValue); } if ( ! uXMLMoveToParentElement(xmlObj, &xmlStatus) ) { printf(”Error in %s error code=%d\n”, ”uXMLMoveToParentElement”, xmlStatus); } /* CREATE ELEMENT country=’USA’ */ if ( ! uXMLCreateElement(xmlObj, &xmlStatus, ”country”, ”USA”) ) { printf(”Error in %s error code=%d\n”, ”uXMLCreateElement”, xmlStatus); } Appendix A: Examples 549 /* THE CURRENT ELEMENT CURSOR POINTS TO TAG ”country” */ if ( ! uXMLGetCurrentElementName(xmlObj, &xmlStatus, &elementName) ) { printf(”Error in %s error code=%d\n”, ”uXMLGetCurrentElementName”, xmlStatus); } else { /* THE TAG NAME OF CURRENT ELEMENT i.e. ”country” */ printf(”\n name of last child = %s \n”, elementName); free(elementName); } if ( ! uXMLMoveToParentElement(xmlObj, &xmlStatus) ) { printf(”Error in %s error code=%d\n”, ”uXMLMoveToParentElement”, xmlStatus); } if ( ! uXMLMoveToFirstChild(xmlObj, &xmlStatus) ) { printf(”Error in %s error code=%d\n”, ”uXMLMoveToFirstChild”, xmlStatus); } if ( ! uXMLDeleteCurrentElement(xmlObj, &xmlStatus) ) { printf(”Error in %s error code=%d\n”, ”uXMLDeleteCurrentElement”, xmlStatus); } /*DELETES THE CURRENT ELEMENT, POINTS TO ”person” */ THE CURRENT ELEMENT CURSOR NOW uXMLDestroy(&xmlObj, &xmlStatus); exit(0); } 550 Appendix A: Examples