XML User Guide

Informatica PowerCenter®
(Version 7.1.1)
Informatica PowerCenter XML User Guide
Version 7.1.1
August 2004

Copyright (c) 1998–2004 Informatica Corporation.

All rights reserved. Printed in the USA.

This software and documentation contain proprietary information of Informatica Corporation, they are provided under a license agreement
containing restrictions on use and disclosure and is also protected by copyright law. Reverse engineering of the software is prohibited. No
part of this document may be reproduced or transmitted in any form, by any means (electronic, photocopying, recording or otherwise)
without prior consent of Informatica Corporation.

Use, duplication, or disclosure of the Software by the U.S. Government is subject to the restrictions set forth in the applicable software
license agreement as provided in DFARS 227.7202-1(a) and 227.7702-3(a) (1995), DFARS 252.227-7013(c)(1)(ii) (OCT 1988), FAR
12.212(a) (1995), FAR 52.227-19, or FAR 52.227-14 (ALT III), as applicable.

The information in this document is subject to change without notice. If you find any problems in the documentation, please report them to
us in writing. Informatica Corporation does not warrant that this documentation is error free.
Informatica, PowerMart, PowerCenter, PowerChannel, PowerCenter Connect, MX, and SuperGlue are trademarks or registered trademarks
of Informatica Corporation in the United States and in jurisdictions throughout the world. All other company and product names may be
trade names or trademarks of their respective owners.

Portions of this software are copyrighted by DataDirect Technologies, 1999-2002.

Informatica PowerCenter products contain ACE (TM) software copyrighted by Douglas C. Schmidt and his research group at Washington
University and University of California, Irvine, Copyright (c) 1993-2002, all rights reserved.

Portions of this software contain copyrighted material from The JBoss Group, LLC. Your right to use such materials is set forth in the GNU
Lesser General Public License Agreement, which may be found at http://www.opensource.org/licenses/lgpl-license.php. The JBoss materials
are provided free of charge by Informatica, “as-is”, without warranty of any kind, either express or implied, including but not limited to the
implied warranties of merchantability and fitness for a particular purpose.

Portions of this software contain copyrighted material from Meta Integration Technology, Inc. Meta Integration® is a registered trademark
of Meta Integration Technology, Inc.

This product includes software developed by the Apache Software Foundation (http://www.apache.org/).
The Apache Software is Copyright (c) 1999-2004 The Apache Software Foundation. All rights reserved.

DISCLAIMER: Informatica Corporation provides this documentation “as is” without warranty of any kind, either express or implied,
including, but not limited to, the implied warranties of non-infringement, merchantability, or use for a particular purpose. The information
provided in this documentation may include technical inaccuracies or typographical errors. Informatica could make improvements and/or
changes in the products described in this documentation at any time without notice.
Table of Contents
List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
New Features and Enhancements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xiv
PowerCenter 7.1.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xiv
PowerCenter 7.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xvi
PowerCenter 7.0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xx
About Informatica Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxvi
About this Book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxvii
Document Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxvii
Other Informatica Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxviii
Visiting Informatica Customer Portal . . . . . . . . . . . . . . . . . . . . . . . . xxviii
Visiting the Informatica Webzine . . . . . . . . . . . . . . . . . . . . . . . . . . . xxviii
Visiting the Informatica Web Site . . . . . . . . . . . . . . . . . . . . . . . . . . xxviii
Visiting the Informatica Developer Network . . . . . . . . . . . . . . . . . . . xxviii
Obtaining Technical Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxix

Chapter 1: XML Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
XML Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
DTD Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
DTD Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
DTD Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
XML Schema Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Types of XML Metadata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Namespace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Name . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Datatype . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Hierarchy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Cardinality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Null Constraint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Simple and Complex XML Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

iii
 Simple Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Complex Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Component Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Element and Attribute Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Substitution Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
XML Path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Code Pages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Chapter 2: Using XML with PowerCenter . . . . . . . . . . . . . . . . . . . . . . 29

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Extracting XML Metadata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Metadata from an XML File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Metadata From DTD Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Metadata from an XML Schema . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Metadata from Relational Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Working with XML Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Rules for View Keys and Relationships . . . . . . . . . . . . . . . . . . . . . . . . . 36
Creating Custom XML Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Generating Hierarchical View Relationships . . . . . . . . . . . . . . . . . . . . . . . . 38
Generating Normalized Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Generating a Denormalized View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Generating Entity View Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Rules and Guidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Using Entity Relationships in an XML Definition . . . . . . . . . . . . . . . . . 42
Using Substitution Groups in an XML Definition . . . . . . . . . . . . . . . . . 45
Working with Circular References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Understanding View Rows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Pivoting Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Using Multiple-Level Pivots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

Chapter 3: Working with XML Sources . . . . . . . . . . . . . . . . . . . . . . . 55

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Importing an XML Source Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Multi-line Attributes Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Working with XML Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Generating Entity Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

iv Table of Contents
Generating Hierarchy Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Creating Custom XML Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Selecting Root Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Reducing Metadata Explosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Synchronizing XML Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Editing XML Source Definition Properties . . . . . . . . . . . . . . . . . . . . . . . . . 70
Creating XML Definitions from Repository Definitions . . . . . . . . . . . . . . . 72
Troubleshooting XML Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

Chapter 4: Using the XML Editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
XML Navigator Window . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Workspace Window . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Columns Window . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Creating and Editing Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Creating an XML View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Adding Columns to Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Adding Pivoted Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Applying Content to the Any Content Element . . . . . . . . . . . . . . . . . . . 85
Deleting Columns from Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Adding a Pass-Through Port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Adding a FileName Column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Maintaining View Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Creating Relationships Between Views . . . . . . . . . . . . . . . . . . . . . . . . . 89
Creating Type Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Recreating Entity Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Viewing Schema Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Updating the Namespace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Sorting Components in the Components Pane . . . . . . . . . . . . . . . . . . . 94
Navigating to Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Arranging Views in the Workspace . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Searching for Components in the Schema . . . . . . . . . . . . . . . . . . . . . . . 95
Searching for Components in XML Views . . . . . . . . . . . . . . . . . . . . . . 95
Viewing a Simple Type Hierarchy . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Viewing Complex Types in the Schema . . . . . . . . . . . . . . . . . . . . . . . . 96
Viewing XML Metadata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Previewing XML Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

Table of Contents v
 Validating XML Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Setting XML View Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
All Hierarchy Foreign Keys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Non-Recursive Row Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Hierarchy Relationship Row Option . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Force Row Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Type Relationship Row Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Troubleshooting XML Editing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

Chapter 5: Working with XML Targets . . . . . . . . . . . . . . . . . . . . . . . 103

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Importing an XML Target Definition from XML Files . . . . . . . . . . . . . . . . 105
Creating a Target from an XML Source Definition . . . . . . . . . . . . . . . . . . 106
Editing XML Target Definition Properties . . . . . . . . . . . . . . . . . . . . . . . . 107
Validating XML Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Hierarchy Relationship Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Type Relationship Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Inheritance Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Using an XML Target in a Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Active Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Selecting a Root Element . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Connecting Target Ports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Connecting Abstract Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Flushing XML Data to Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Naming XML Files Dynamically . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

Chapter 6: XML Source Qualifier Transformation . . . . . . . . . . . . . . 121

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
Adding an XML Source Qualifier to a Mapping . . . . . . . . . . . . . . . . . . . . . 123
Automatically Creating an XML Source Qualifier Transformation . . . . 123
Manually Creating an XML Source Qualifier Transformation . . . . . . . . 123
Editing an XML Source Qualifier Transformation . . . . . . . . . . . . . . . . . . . 125
Setting Sequence Numbers for Generated Keys . . . . . . . . . . . . . . . . . . 128
Using the XML Source Qualifier in a Mapping . . . . . . . . . . . . . . . . . . . . . 129
XML Source Qualifier Transformation Example . . . . . . . . . . . . . . . . . 131
Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

vi Table of Contents
Chapter 7: Midstream XML Transformations . . . . . . . . . . . . . . . . . . 137
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
XML Parser Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
XML Generator Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Creating a Midstream XML Transformation . . . . . . . . . . . . . . . . . . . . . . . 143
Editing Midstream XML Transformation Properties . . . . . . . . . . . . . . . . . 144
Midstream XML Parser Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
Midstream XML Generator Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Generating Pass-Through Ports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147

Chapter 8: Working with XML Sessions . . . . . . . . . . . . . . . . . . . . . . 151

Working with XML Sources in a Session . . . . . . . . . . . . . . . . . . . . . . . . . . 152
Server Handling for XML Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Working with XML Targets in a Session . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Server Handling for XML Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Character Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Special Characters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
Null and Empty String . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
Handling Duplicate Group Rows . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
DTD and Schema Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Flushing XML on Commits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
XML Caching Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Session Logs for XML Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Multiple XML Document Output . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Working with Midstream XML Transformations . . . . . . . . . . . . . . . . . . . . 165

Appendix A: XML Datatype Reference . . . . . . . . . . . . . . . . . . . . . . . 169

XML and Transformation Datatypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Unsupported Datatypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
XML Date Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175

Table of Contents vii

viii Table of Contents
List of Figures
Figure 1-1. Sample XML File: StoreInfo.xml . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Figure 1-2. Elements in the XML Hierarchy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Figure 1-3. Sample DTD: StoreInfo.dtd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Figure 1-4. Sample XML Schema File: StoreInfo.xsd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Figure 1-5. XML Cardinality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Figure 1-6. Relative Cardinality of Elements in StoreInfo.xml . . . . . . . . . . . . . . . . . . . . . . . . 16
Figure 1-7. Sample XSD File with Union Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Figure 1-8. Restricted and Extended Complex Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Figure 1-9. XPath of Elements and Attributes in the StoreInfo.xml . . . . . . . . . . . . . . . . . . . . 26
Figure 2-1. A Sample Employees XML File with Multiple-Occurring Elements . . . . . . . . . . . . 31
Figure 2-2. The Root Element and XML Views in an XML Definition . . . . . . . . . . . . . . . . . . 32
Figure 2-3. XML Definition From an XML File Referencing a DTD File . . . . . . . . . . . . . . . . 33
Figure 2-4. XML Schema with a Derived Complex Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Figure 2-5. XML Definition Containing a Derived Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Figure 2-6. XML Target Created from Two Relational Sources . . . . . . . . . . . . . . . . . . . . . . . 35
Figure 2-7. Sample DTD File to Create Normalized Views . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Figure 2-8. Source Definition for StoreInfo.dtd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Figure 2-9. Normalized Views Data Preview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Figure 2-10. Sample XML File to Create a Denormalized View . . . . . . . . . . . . . . . . . . . . . . . 40
Figure 2-11. Source Definition Containing a Denormalized View . . . . . . . . . . . . . . . . . . . . . 40
Figure 2-12. Data Preview for the ProdAndSales.xml Denormalized View . . . . . . . . . . . . . . . . 41
Figure 2-13. Complex Type View Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Figure 2-14. Complex Type XML Views Sample Schema . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Figure 2-15. Sample XML Schema Using Substitution Groups . . . . . . . . . . . . . . . . . . . . . . . . 46
Figure 2-16. Sample XML Definition Generated from Substitution Groups . . . . . . . . . . . . . . 46
Figure 2-17. Circular Reference View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Figure 2-18. Circular Reference Data Preview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Figure 2-19. ADDRESS Element Occurrences Pivoted into Columns . . . . . . . . . . . . . . . . . . . 52
Figure 3-1. The XML Wizard Options to Create Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Figure 3-2. Root Selection Page . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Figure 3-3. Reduce Metadata Explosions Page . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Figure 4-1. XML Editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Figure 4-2. Pivoted Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Figure 4-3. Specify XPath Pivot Column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Figure 4-4. Any Content Element in the Schema Navigator . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Figure 4-5. Edit Any or Any Attribute Dialog Box . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Figure 4-6. New Type Added to the Any Content Element . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Figure 4-7. Search XML Views and Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Figure 4-8. Preview XML Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Figure 4-9. Columns Window . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

List of Figures ix
 Figure 5-1. Filename Column in a Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .117
Figure 6-1. XML Source Qualifier Transformation Ports . . . . . . . . . . . . . . . . . . . . . . . . . . . .126
Figure 6-2. Linking XML Source Qualifier Transformations to One Input Group . . . . . . . . . .129
Figure 6-3. Linking XML Source Qualifier to Multiple Input Group Transformations . . . . . .130
Figure 6-4. Sample XML File StoreInfo.xml . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .131
Figure 6-5. Invalid use of XML Source Qualifier Transformation in Aggregator Mapping . . . .132
Figure 6-6. Using a Denormalized Group in a Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . .133
Figure 6-7. Using an XML Source Definition Twice in a Mapping . . . . . . . . . . . . . . . . . . . . .134
Figure 7-1. XML Parser Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .140
Figure 7-2. XML Generator Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .141
Figure 7-3. Sample XML Generator Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .142
Figure 7-4. Midstream XML Parser Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145
Figure 7-5. Midstream XML Generator Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .146
Figure 7-6. Pass-Through Ports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .149
Figure 8-1. Properties Settings for an XML Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .152
Figure 8-2. Properties Settings for an XML Writer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .155
Figure 8-3. Mapping Data to an XML Root . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .163
Figure 8-4. Properties Settings for an XML Generator Transformation . . . . . . . . . . . . . . . . . .165
Figure 8-5. Properties Settings for an XML Parser Transformation . . . . . . . . . . . . . . . . . . . . .167

x List of Figures
List of Tables
Table 1-1. Cardinality of Elements in XML . . . . . . . . . . . . . . . . . ... . . . . . . .. . . . . . . . . . . 14
Table 3-1. Create XML Views Options . . . . . . . . . . . . . . . . . . . . ... . . . . . . .. . . . . . . . . . . 59
Table 6-1. XML Source Qualifier Properties . . . . . . . . . . . . . . . . ... . . . . . . .. . . . . . . . . . 127
Table 7-1. Midstream XML Parser Settings . . . . . . . . . . . . . . . . . ... . . . . . . .. . . . . . . . . . 145
Table 7-2. Midstream XML Generator Settings . . . . . . . . . . . . . . ... . . . . . . .. . . . . . . . . . 146
Table 8-1. XML Reader Options . . . . . . . . . . . . . . . . . . . . . . . . . ... . . . . . . .. . . . . . . . . . 152
Table 8-2. XML Source Qualifier Options for a Session . . . . . . . . ... . . . . . . .. . . . . . . . . . 153
Table 8-3. XML Writer Options . . . . . . . . . . . . . . . . . . . . . . . . . ... . . . . . . .. . . . . . . . . . 156
Table 8-4. Null and Empty String Output for XML Targets . . . . . ... . . . . . . .. . . . . . . . . . 159
Table 8-5. XML Generator Transformation Session Options. . . . . ... . . . . . . .. . . . . . . . . . 165
Table 8-6. XML Parser Transformation Session Options . . . . . . . ... . . . . . . .. . . . . . . . . . 167
Table A-1. XML and Transformation Datatypes . . . . . . . . . . . . . . ... . . . . . . .. . . . . . . . . . 170

List of Tables xi
xii List of Tables
Preface

Welcome to PowerCenter, Informatica’s software product that delivers an open, scalable data
integration solution addressing the complete life cycle for all data integration projects
including data warehouses and data marts, data migration, data synchronization, and
information hubs. PowerCenter combines the latest technology enhancements for reliably
managing data repositories and delivering information resources in a timely, usable, and
efficient manner.
The PowerCenter metadata repository coordinates and drives a variety of core functions,
including extracting, transforming, loading, and managing data. The PowerCenter Server can
extract large volumes of data from multiple platforms, handle complex transformations on the
data, and support high-speed loads. PowerCenter can simplify and accelerate the process of
moving data warehouses from development to test to production.

xiii
New Features and Enhancements
This section describes new features and enhancements to PowerCenter 7.1.1, 7.1, and 7.0.

PowerCenter 7.1.1
This section describes new features and enhancements to PowerCenter 7.1.1.

Data Profiling
♦ Data sampling. You can create a data profile for a sample of source data instead of the
entire source. You can view a profile from a random sample of data, a specified percentage
of data, or for a specified number of rows starting with the first row.
♦ Verbose data enhancements. You can specify the type of verbose data you want the
PowerCenter Server to write to the Data Profiling warehouse. The PowerCenter Server can
write all rows, the rows that meet the business rule, or the rows that do not meet the
business rule.
♦ Session enhancement. You can save sessions that you create from the Profile Manager to
the repository.
♦ Domain Inference function tuning. You can configure the Data Profiling Wizard to filter
the Domain Inference function results. You can configure a maximum number of patterns
and a minimum pattern frequency. You may want to narrow the scope of patterns returned
to view only the primary domains, or you may want to widen the scope of patterns
returned to view exception data.
♦ Row Uniqueness function. You can determine unique rows for a source based on a
selection of columns for the specified source.
♦ Define mapping, session, and workflow prefixes. You can define default mapping,
session, and workflow prefixes for the mappings, sessions, and workflows generated when
you create a data profile.
♦ Profile mapping display in the Designer. The Designer displays profile mappings under a
profile mappings node in the Navigator.

PowerCenter Server
♦ Code page. PowerCenter supports additional Japanese language code pages, such as JIPSE-
kana, JEF-kana, and MELCOM-kana.
♦ Flat file partitioning. When you create multiple partitions for a flat file source session, you
can configure the session to create multiple threads to read the flat file source.
♦ pmcmd. You can use parameter files that reside on a local machine with the Startworkflow
command in the pmcmd program. When you use a local parameter file, pmcmd passes
variables and values in the file to the PowerCenter Server.

xiv Preface
♦ SuSE Linux support. The PowerCenter Server runs on SuSE Linux. On SuSE Linux, you
can connect to IBM, DB2, Oracle, and Sybase sources, targets, and repositories using
native drivers. Use ODBC drivers to access other sources and targets.
♦ Reserved word support. If any source, target, or lookup table name or column name
contains a database reserved word, you can create and maintain a file, reswords.txt,
containing reserved words. When the PowerCenter Server initializes a session, it searches
for reswords.txt in the PowerCenter Server installation directory. If the file exists, the
PowerCenter Server places quotes around matching reserved words when it executes SQL
against the database.
♦ Teradata external loader. When you load to Teradata using an external loader, you can
now override the control file. Depending on the loader you use, you can also override the
error, log, and work table names by specifying different tables on the same or different
Teradata database.

Repository
♦ Exchange metadata with other tools. You can exchange source and target metadata with
other BI or data modeling tools, such as Business Objects Designer. You can export or
import multiple objects at a time. When you export metadata, the PowerCenter Client
creates a file format recognized by the target tool.

Repository Server
♦ pmrep. You can use pmrep to perform the following functions:
− Remove repositories from the Repository Server cache entry list.
− Enable enhanced security when you create a relational source or target connection in the
repository.
− Update a connection attribute value when you update the connection.
♦ SuSE Linux support. The Repository Server runs on SuSE Linux. On SuSE Linux, you
can connect to IBM, DB2, Oracle, and Sybase repositories.

Security
♦ Oracle OS Authentication. You can now use Oracle OS Authentication to authenticate
database users. Oracle OS Authentication allows you to log on to an Oracle database if you
have a logon to the operating system. You do not need to know a database user name and
password. PowerCenter uses Oracle OS Authentication when the user name for an Oracle
connection is PmNullUser.

Web Services Provider

♦ Attachment support. When you import web service definitions with attachment groups,
you can pass attachments through the requests or responses in a service session. The
document type you can attach is based on the mime content of the WSDL file. You can
attach document types such as XML, JPEG, GIF, or PDF.

Preface xv
 ♦ Pipeline partitioning. You can create multiple partitions in a session containing web
service source and target definitions. The PowerCenter Server creates a connection to the
Web Services Hub based on the number of sources, targets, and partitions in the session.

XML
♦ Multi-level pivoting. You can now pivot more than one multiple-occurring element in an
XML view. You can also pivot the view row.

PowerCenter 7.1
This section describes new features and enhancements to PowerCenter 7.1.

Data Profiling
♦ Data Profiling for VSAM sources. You can now create a data profile for VSAM sources.
♦ Support for verbose mode for source-level functions. You can now create data profiles
with source-level functions and write data to the Data Profiling warehouse in verbose
mode.
♦ Aggregator function in auto profiles. Auto profiles now include the Aggregator function.
♦ Creating auto profile enhancements. You can now select the columns or groups you want
to include in an auto profile and enable verbose mode for the Distinct Value Count
function.
♦ Purging data from the Data Profiling warehouse. You can now purge data from the Data
Profiling warehouse.
♦ Source View in the Profile Manager. You can now view data profiles by source definition
in the Profile Manager.
♦ PowerCenter Data Profiling report enhancements. You can now view PowerCenter Data
Profiling reports in a separate browser window, resize columns in a report, and view
verbose data for Distinct Value Count functions.
♦ Prepackaged domains. Informatica provides a set of prepackaged domains that you can
include in a Domain Validation function in a data profile.

Documentation
♦ Web Services Provider Guide. This is a new book that describes the functionality of Real-time
Web Services. It also includes information from the version 7.0 Web Services Hub Guide.
♦ XML User Guide. This book consolidates XML information previously documented in the
Designer Guide, Workflow Administration Guide, and Transformation Guide.

Licensing
Informatica provides licenses for each CPU and each repository rather than for each
installation. Informatica provides licenses for product, connectivity, and options. You store

xvi Preface
the license keys in a license key file. You can manage the license files using the Repository
Server Administration Console, the PowerCenter Server Setup, and the command line
program, pmlic.

PowerCenter Server
♦ 64-bit support. You can now run 64-bit PowerCenter Servers on AIX and HP-UX
(Itanium).
♦ Partitioning enhancements. If you have the Partitioning option, you can define up to 64
partitions at any partition point in a pipeline that supports multiple partitions.
♦ PowerCenter Server processing enhancements. The PowerCenter Server now reads a
block of rows at a time. This improves processing performance for most sessions.
♦ CLOB/BLOB datatype support. You can now read and write CLOB/BLOB datatypes.

PowerCenter Metadata Reporter

PowerCenter Metadata Reporter modified some report names and uses the PowerCenter 7.1
MX views in its schema.

Repository Server
♦ Updating repository statistics. PowerCenter now identifies and updates statistics for all
repository tables and indexes when you copy, upgrade, and restore repositories. This
improves performance when PowerCenter accesses the repository.
♦ Increased repository performance. You can increase repository performance by skipping
information when you copy, back up, or restore a repository. You can choose to skip MX
data, workflow and session log history, and deploy group history.
♦ pmrep. You can use pmrep to back up, disable, or enable a repository, delete a relational
connection from a repository, delete repository details, truncate log files, and run multiple
pmrep commands sequentially. You can also use pmrep to create, modify, and delete a
folder.

Repository
♦ Exchange metadata with business intelligence tools. You can export metadata to and
import metadata from other business intelligence tools, such as Cognos Report Net and
Business Objects.
♦ Object import and export enhancements. You can compare objects in an XML file to
objects in the target repository when you import objects.
♦ MX views. MX views have been added to help you analyze metadata stored in the
repository. REP_SERVER_NET and REP_SERVER_NET_REF views allow you to see
information about server grids. REP_VERSION_PROPS allows you to see the version
history of all objects in a PowerCenter repository.

Preface xvii
 Transformations
♦ Flat file lookup. You can now perform lookups on flat files. When you create a Lookup
transformation using a flat file as a lookup source, the Designer invokes the Flat File
Wizard. You can also use a lookup file parameter if you want to change the name or
location of a lookup between session runs.
♦ Dynamic lookup cache enhancements. When you use a dynamic lookup cache, the
PowerCenter Server can ignore some ports when it compares values in lookup and input
ports before it updates a row in the cache. Also, you can choose whether the PowerCenter
Server outputs old or new values from the lookup/output ports when it updates a row. You
might want to output old values from lookup/output ports when you use the Lookup
transformation in a mapping that updates slowly changing dimension tables.
♦ Union transformation. You can use the Union transformation to merge multiple sources
into a single pipeline. The Union transformation is similar to using the UNION ALL SQL
statement to combine the results from two or more SQL statements.
♦ Custom transformation API enhancements. The Custom transformation API includes
new array-based functions that allow you to create procedure code that receives and
outputs a block of rows at a time. Use these functions to take advantage of the
PowerCenter Server processing enhancements.
♦ Midstream XML transformations. You can now create an XML Parser transformation or
an XML Generator transformation to parse or generate XML inside a pipeline. The XML
transformations enable you to extract XML data stored in relational tables, such as data
stored in a CLOB column. You can also extract data from messaging systems, such as
TIBCO or IBM MQSeries.

Usability
♦ Viewing active folders. The Designer and the Workflow Manager highlight the active
folder in the Navigator.
♦ Enhanced printing. The quality of printed workspace has improved.

Version Control
You can run object queries that return shortcut objects. You can also run object queries based
on the latest status of an object. The query can return local objects that are checked out, the
latest version of checked in objects, or a collection of all older versions of objects.

Web Services Provider

♦ Real-time Web Services. Real-time Web Services allows you to create services using the
Workflow Manager and make them available to web service clients through the Web
Services Hub. The PowerCenter Server can perform parallel processing of both request-
response and one-way services.
♦ Web Services Hub. The Web Services Hub now hosts Real-time Web Services in addition
to Metadata Web Services and Batch Web Services. You can install the Web Services Hub
on a JBoss application server.

xviii Preface
Note: PowerCenter Connect for Web Services allows you to create sources, targets, and
transformations to call web services hosted by other providers. For more informations, see
PowerCenter Connect for Web Services User and Administrator Guide.

Workflow Monitor
The Workflow Monitor includes the following performance and usability enhancements:
♦ When you connect to the PowerCenter Server, you no longer distinguish between online
or offline mode.
♦ You can open multiple instances of the Workflow Monitor on one machine.
♦ You can simultaneously monitor multiple PowerCenter Servers registered to the same
repository.
♦ The Workflow Monitor includes improved options for filtering tasks by start and end
time.
♦ The Workflow Monitor displays workflow runs in Task view chronologically with the most
recent run at the top. It displays folders alphabetically.
♦ You can remove the Navigator and Output window.

XML Support
PowerCenter XML support now includes the following features:
♦ Enhanced datatype support. You can use XML schemas that contain simple and complex
datatypes.
♦ Additional options for XML definitions. When you import XML definitions, you can
choose how you want the Designer to represent the metadata associated with the imported
files. You can choose to generate XML views using hierarchy or entity relationships. In a
view with hierarchy relationships, the Designer expands each element and reference under
its parent element. When you create views with entity relationships, the Designer creates
separate entities for references and multiple-occurring elements.
♦ Synchronizing XML definitions. You can synchronize one or more XML definition when
the underlying schema changes. You can synchronize an XML definition with any
repository definition or file used to create the XML definition, including relational sources
or targets, XML files, DTD files, or schema files.
♦ XML workspace. You can edit XML views and relationships between views in the
workspace. You can create views, add or delete columns from views, and define
relationships between views.
♦ Midstream XML transformations. You can now create an XML Parser transformation or
an XML Generator transformation to parse or generate XML inside a pipeline. The XML
transformations enable you to extract XML data stored in relational tables, such as data
stored in a CLOB column. You can also extract data from messaging systems, such as
TIBCO or IBM MQSeries.

Preface xix
 ♦ Support for circular references. Circular references occur when an element is a direct or
indirect child of itself. PowerCenter now supports XML files, DTD files, and XML
schemas that use circular definitions.
♦ Increased performance for large XML targets. You can create XML files of several
gigabytes in a PowerCenter 7.1 XML session by using the following enhancements:
− Spill to disk. You can specify the size of the cache used to store the XML tree. If the size
of the tree exceeds the cache size, the XML data spills to disk in order to free up
memory.
− User-defined commits. You can define commits to trigger flushes for XML target files.
− Support for multiple XML output files. You can output XML data to multiple XML
targets. You can also define the file names for XML output files in the mapping.

PowerCenter 7.0
This section describes new features and enhancements to PowerCenter 7.0.

Data Profiling
If you have the Data Profiling option, you can profile source data to evaluate source data and
detect patterns and exceptions. For example, you can determine implicit data type, suggest
candidate keys, detect data patterns, and evaluate join criteria. After you create a profiling
warehouse, you can create profiling mappings and run sessions. Then you can view reports
based on the profile data in the profiling warehouse.
The PowerCenter Client provides a Profile Manager and a Profile Wizard to complete these
tasks.

Data Integration Web Services

You can use Data Integration Web Services to write applications to communicate with the
PowerCenter Server. Data Integration Web Services is a web-enabled version of the
PowerCenter Server functionality available through Load Manager and Metadata Exchange. It
is comprised of two services for communication with the PowerCenter Server, Load Manager
and Metadata Exchange Web Services running on the Web Services Hub.

Documentation
♦ Glossary. The Installation and Configuration Guide contains a glossary of new PowerCenter
terms.
♦ Installation and Configuration Guide. The connectivity information in the Installation
and Configuration Guide is consolidated into two chapters. This book now contains
chapters titled “Connecting to Databases from Windows” and “Connecting to Databases
from UNIX.”
♦ Upgrading metadata. The Installation and Configuration Guide now contains a chapter
titled “Upgrading Repository Metadata.” This chapter describes changes to repository

xx Preface
 objects impacted by the upgrade process. The change in functionality for existing objects
depends on the version of the existing objects. Consult the upgrade information in this
chapter for each upgraded object to determine whether the upgrade applies to your current
version of PowerCenter.

Functions
♦ Soundex. The Soundex function encodes a string value into a four-character string.
SOUNDEX works for characters in the English alphabet (A-Z). It uses the first character
of the input string as the first character in the return value and encodes the remaining
three unique consonants as numbers.
♦ Metaphone. The Metaphone function encodes string values. You can specify the length of
the string that you want to encode. METAPHONE encodes characters of the English
language alphabet (A-Z). It encodes both uppercase and lowercase letters in uppercase.

Installation
♦ Remote PowerCenter Client installation. You can create a control file containing
installation information, and distribute it to other users to install the PowerCenter Client.
You access the Informatica installation CD from the command line to create the control
file and install the product.

PowerCenter Metadata Reporter

PowerCenter Metadata Reporter replaces Runtime Metadata Reporter and Informatica
Metadata Reporter. PowerCenter Metadata Reporter includes the following features:
♦ Metadata browsing. You can use PowerCenter Metadata Reporter to browse PowerCenter
7.0 metadata, such as workflows, worklets, mappings, source and target tables, and
transformations.
♦ Metadata analysis. You can use PowerCenter Metadata Reporter to analyze operational
metadata, including session load time, server load, session completion status, session
errors, and warehouse growth.

PowerCenter Server
♦ DB2 bulk loading. You can enable bulk loading when you load to IBM DB2 8.1.
♦ Distributed processing. If you purchase the Server Grid option, you can group
PowerCenter Servers registered to the same repository into a server grid. In a server grid,
PowerCenter Servers balance the workload among all the servers in the grid.
♦ Row error logging. The session configuration object has new properties that allow you to
define error logging. You can choose to log row errors in a central location to help
understand the cause and source of errors.
♦ External loading enhancements. When using external loaders on Windows, you can now
choose to load from a named pipe. When using external loaders on UNIX, you can now
choose to load from staged files.

Preface xxi
 ♦ External loading using Teradata Warehouse Builder. You can use Teradata Warehouse
Builder to load to Teradata. You can choose to insert, update, upsert, or delete data.
Additionally, Teradata Warehouse Builder can simultaneously read from multiple sources
and load data into one or more tables.
♦ Mixed mode processing for Teradata external loaders. You can now use data driven load
mode with Teradata external loaders. When you select data driven loading, the
PowerCenter Server flags rows for insert, delete, or update. It writes a column in the target
file or named pipe to indicate the update strategy. The control file uses these values to
determine how to load data to the target.
♦ Concurrent processing. The PowerCenter Server now reads data concurrently from
sources within a target load order group. This enables more efficient joins with minimal
usage of memory and disk cache.
♦ Real time processing enhancements. You can now use real-time processing in sessions that
also process active transformations, such as the Aggregator transformation. You can apply
the transformation logic to rows defined by transaction boundaries.

Repository Server
♦ Object export and import enhancements. You can now export and import objects using
the Repository Manager and pmrep. You can export and import multiple objects and
objects types. You can export and import objects with or without their dependent objects.
You can also export objects from a query result or objects history.
♦ pmrep commands. You can use pmrep to perform change management tasks, such as
maintaining deployment groups and labels, checking in, deploying, importing, exporting,
and listing objects. You can also use pmrep to run queries. The deployment and object
import commands require you to use a control file to define options and resolve conflicts.
♦ Trusted connections. You can now use a Microsoft SQL Server trusted connection to
connect to the repository.

Security
♦ LDAP user authentication. You can now use default repository user authentication or
Lightweight Directory Access Protocol (LDAP) to authenticate users. If you use LDAP, the
repository maintains an association between your repository user name and your external
login name. When you log in to the repository, the security module passes your login name
to the external directory for authentication. The repository maintains a status for each
user. You can now enable or disable users from accessing the repository by changing the
status. You do not have to delete user names from the repository.
♦ Use Repository Manager privilege. The Use Repository Manager privilege allows you to
perform tasks in the Repository Manager, such as copy object, maintain labels, and change
object status. You can perform the same tasks in the Designer and Workflow Manager if
you have the Use Designer and Use Workflow Manager privileges.
♦ Audit trail. You can track changes to repository users, groups, privileges, and permissions
through the Repository Server Administration Console. The Repository Agent logs
security changes to a log file stored in the Repository Server installation directory. The

xxii Preface
 audit trail log contains information, such as changes to folder properties, adding or
removing a user or group, and adding or removing privileges.

Transformations
♦ Custom transformation. Custom transformations operate in conjunction with procedures
you create outside of the Designer interface to extend PowerCenter functionality. The
Custom transformation replaces the Advanced External Procedure transformation. You can
create Custom transformations with multiple input and output groups, and you can
compile the procedure with any C compiler.
You can create templates that customize the appearance and available properties of a
Custom transformation you develop. You can specify the icons used for transformation,
the colors, and the properties a mapping developer can modify. When you create a Custom
transformation template, distribute the template with the DLL or shared library you
develop.
♦ Joiner transformation. You can use the Joiner transformation to join two data streams that
originate from the same source.

Version Control
The PowerCenter Client and repository introduce features that allow you to create and
manage multiple versions of objects in the repository. Version control allows you to maintain
multiple versions of an object, control development on the object, track changes, and use
deployment groups to copy specific groups of objects from one repository to another. Version
control in PowerCenter includes the following features:
♦ Object versioning. Individual objects in the repository are now versioned. This allows you
to store multiple copies of a given object during the development cycle. Each version is a
separate object with unique properties.
♦ Check out and check in versioned objects. You can check out and reserve an object you
want to edit, and check in the object when you are ready to create a new version of the
object in the repository.
♦ Compare objects. The Repository Manager and Workflow Manager allow you to compare
two repository objects of the same type to identify differences between them. You can
compare Designer objects and Workflow Manager objects in the Repository Manager. You
can compare tasks, sessions, worklets, and workflows in the Workflow Manager. The
PowerCenter Client tools allow you to compare objects across open folders and
repositories. You can also compare different versions of the same object.
♦ Delete or purge a version. You can delete an object from view and continue to store it in
the repository. You can recover or undelete deleted objects. If you want to permanently
remove an object version, you can purge it from the repository.
♦ Deployment. Unlike copying a folder, copying a deployment group allows you to copy a
select number of objects from multiple folders in the source repository to multiple folders
in the target repository. This gives you greater control over the specific objects copied from
one repository to another.

Preface xxiii
 ♦ Deployment groups. You can create a deployment group that contains references to
objects from multiple folders across the repository. You can create a static deployment
group that you manually add objects to, or create a dynamic deployment group that uses a
query to populate the group.
♦ Labels. A label is an object that you can apply to versioned objects in the repository. This
allows you to associate multiple objects in groups defined by the label. You can use labels
to track versioned objects during development, improve query results, and organize groups
of objects for deployment or export and import.
♦ Queries. You can create a query that specifies conditions to search for objects in the
repository. You can save queries for later use. You can make a private query, or you can
share it with all users in the repository.
♦ Track changes to an object. You can view a history that includes all versions of an object
and compare any version of the object in the history to any other version. This allows you
to see the changes made to an object over time.

XML Support
PowerCenter contains XML features that allow you to validate an XML file against an XML
schema, declare multiple namespaces, use XPath to locate XML nodes, increase performance
for large XML files, format your XML file output for increased readability, and parse or
generate XML data from various sources. XML support in PowerCenter includes the
following features:
♦ XML schema. You can use an XML schema to validate an XML file and to generate source
and target definitions. XML schemas allow you to declare multiple namespaces so you can
use prefixes for elements and attributes. XML schemas also allow you to define some
complex datatypes.
♦ XPath support. The XML wizard allows you to view the structure of XML schema. You
can use XPath to locate XML nodes.
♦ Increased performance for large XML files. When you process an XML file or stream, you
can set commits and periodically flush XML data to the target instead of writing all the
output at the end of the session. You can choose to append the data to the same target file
or create a new target file after each flush.
♦ XML target enhancements. You can format the XML target file so that you can easily view
the XML file in a text editor. You can also configure the PowerCenter Server to not output
empty elements to the XML target.

Usability
♦ Copying objects. You can now copy objects from all the PowerCenter Client tools using
the copy wizard to resolve conflicts. You can copy objects within folders, to other folders,
and to different repositories. Within the Designer, you can also copy segments of
mappings to a workspace in a new folder or repository.
♦ Comparing objects. You can compare workflows and tasks from the Workflow Manager.
You can also compare all objects from within the Repository Manager.

xxiv Preface
♦ Change propagation. When you edit a port in a mapping, you can choose to propagate
changed attributes throughout the mapping. The Designer propagates ports, expressions,
and conditions based on the direction that you propagate and the attributes you choose to
propagate.
♦ Enhanced partitioning interface. The Session Wizard is enhanced to provide a graphical
depiction of a mapping when you configure partitioning.
♦ Revert to saved. You can now revert to the last saved version of an object in the Workflow
Manager. When you do this, the Workflow Manager accesses the repository to retrieve the
last-saved version of the object.
♦ Enhanced validation messages. The PowerCenter Client writes messages in the Output
window that describe why it invalidates a mapping or workflow when you modify a
dependent object.
♦ Validate multiple objects. You can validate multiple objects in the repository without
fetching them into the workspace. You can save and optionally check in objects that
change from invalid to valid status as a result of the validation. You can validate sessions,
mappings, mapplets, workflows, and worklets.
♦ View dependencies. Before you edit or delete versioned objects, such as sources, targets,
mappings, or workflows, you can view dependencies to see the impact on other objects.
You can view parent and child dependencies and global shortcuts across repositories.
Viewing dependencies help you modify objects and composite objects without breaking
dependencies.
♦ Refresh session mappings. In the Workflow Manager, you can refresh a session mapping.

Preface xxv
About Informatica Documentation
The complete set of documentation for PowerCenter includes the following books:
♦ Data Profiling Guide. Provides information about how to profile PowerCenter sources to
evaluate source data and detect patterns and exceptions.
♦ Designer Guide. Provides information needed to use the Designer. Includes information to
help you create mappings, mapplets, and transformations. Also includes a description of
the transformation datatypes used to process and transform source data.
♦ Getting Started. Provides basic tutorials for getting started.
♦ Installation and Configuration Guide. Provides information needed to install and
configure the PowerCenter tools, including details on environment variables and database
connections.
♦ PowerCenter Connect® for JMS® User and Administrator Guide. Provides information
to install PowerCenter Connect for JMS, build mappings, extract data from JMS messages,
and load data into JMS messages.
♦ Repository Guide. Provides information needed to administer the repository using the
Repository Manager or the pmrep command line program. Includes details on
functionality available in the Repository Manager and Administration Console, such as
creating and maintaining repositories, folders, users, groups, and permissions and
privileges.
♦ Transformation Language Reference. Provides syntax descriptions and examples for each
transformation function provided with PowerCenter.
♦ Transformation Guide. Provides information on how to create and configure each type of
transformation in the Designer.
♦ Troubleshooting Guide. Lists error messages that you might encounter while using
PowerCenter. Each error message includes one or more possible causes and actions that
you can take to correct the condition.
♦ Web Services Provider Guide. Provides information you need to install and configure the Web
Services Hub. This guide also provides information about how to use the web services that the
Web Services Hub hosts. The Web Services Hub hosts Real-time Web Services, Batch Web
Services, and Metadata Web Services.
♦ Workflow Administration Guide. Provides information to help you create and run
workflows in the Workflow Manager, as well as monitor workflows in the Workflow
Monitor. Also contains information on administering the PowerCenter Server and
performance tuning.
♦ XML User Guide. Provides information you need to create XML definitions from XML,
XSD, or DTD files, and relational or other XML definitions. Includes information on
running sessions with XML data. Also includes details on using the midstream XML
transformations to parse or generate XML data within a pipeline.

xxvi Preface
About this Book
The XML User Guide is written for IS developers and software engineers responsible for
working with XML in a data warehouse environment. Before you use the XML User Guide,
ensure that you have a solid understanding of XML concepts, your operating systems, flat
files, or mainframe system in your environment. Also, ensure that you are familiar with the
interface requirements for your supporting applications.
The material in this book is available for online use.

Document Conventions
This guide uses the following formatting conventions:

If you see… It means…

italicized text The word or set of words are especially emphasized.

boldfaced text Emphasized subjects.

italicized monospaced text This is the variable name for a value you enter as part of an
operating system command. This is generic text that should be
replaced with user-supplied values.

Note: The following paragraph provides additional facts.

Tip: The following paragraph provides suggested uses.

Warning: The following paragraph notes situations where you can overwrite
or corrupt data, unless you follow the specified procedure.

monospaced text This is a code example.

bold monospaced text This is an operating system command you enter from a prompt to
run a task.

Preface xxvii
Other Informatica Resources
In addition to the product manuals, Informatica provides these other resources:
♦ Informatica Customer Portal
♦ Informatica Webzine
♦ Informatica web site
♦ Informatica Developer Network
♦ Informatica Technical Support

Visiting Informatica Customer Portal

As an Informatica customer, you can access the Informatica Customer Portal site at http://
my.informatica.com. The site contains product information, user group information,
newsletters, access to the Informatica customer support case management system (ATLAS),
the Informatica Knowledgebase, Informatica Webzine, and access to the Informatica user
community.

Visiting the Informatica Webzine

The Informatica Documentation team delivers an online journal, the Informatica Webzine.
This journal provides solutions to common tasks, detailed descriptions of specific features,
and tips and tricks to help you develop data warehouses.
The Informatica Webzine is a password-protected site that you can access through the
Customer Portal. The Customer Portal has an online registration form for login accounts to
its webzine and web support. To register for an account, go to http://my.informatica.com.
If you have any questions, please email [email protected].

Visiting the Informatica Web Site

You can access Informatica’s corporate web site at http://www.informatica.com. The site
contains information about Informatica, its background, upcoming events, and locating your
closest sales office. You will also find product information, as well as literature and partner
information. The services area of the site includes important information on technical
support, training and education, and implementation services.

Visiting the Informatica Developer Network

The Informatica Developer Network is a web-based forum for third-party software
developers. You can access the Informatica Developer Network at the following URL:
http://devnet.informatica.com

xxviii Preface
 The site contains information on how to create, market, and support customer-oriented add-
on solutions based on Informatica’s interoperability interfaces.

Obtaining Technical Support

There are many ways to access Informatica technical support. You can call or email your
nearest Technical Support Center listed below or you can use our WebSupport Service.
WebSupport requires a user name and password. You can request a user name and password at
http://my.informatica.com.

North America / South America Africa / Asia / Australia / Europe

Informatica Corporation Informatica Software Ltd.

2100 Seaport Blvd. 6 Waltham Park
Redwood City, CA 94063 Waltham Road, White Waltham
Phone: 866.563.6332 or 650.385.5800 Maidenhead, Berkshire
Fax: 650.213.9489 SL6 3TN
Hours: 6 a.m. - 6 p.m. (PST/PDT) Phone: 44 870 606 1525
email: [email protected] Fax: +44 1628 511 411
Hours: 9 a.m. - 5:30 p.m. (GMT)
email: [email protected]

Belgium
Phone: +32 15 281 702
Hours: 9 a.m. - 5:30 p.m. (local time)

France
Phone: +33 1 41 38 92 26
Hours: 9 a.m. - 5:30 p.m. (local time)

Germany
Phone: +49 1805 702 702
Hours: 9 a.m. - 5:30 p.m. (local time)

Netherlands
Phone: +31 306 082 089
Hours: 9 a.m. - 5:30 p.m. (local time)

Singapore
Phone: +65 322 8589
Hours: 9 a.m. - 5 p.m. (local time)

Switzerland
Phone: +41 800 81 80 70
Hours: 8 a.m. - 5 p.m. (local time)

Preface xxix
xxx Preface
 Chapter 1

XML Concepts

This chapter includes the following topics:

♦ Overview, 2
♦ XML Files, 3
♦ DTD Files, 7
♦ XML Schema Files, 9
♦ Types of XML Metadata, 11
♦ Simple and Complex XML Types, 17
♦ Component Groups, 23
♦ XML Path, 26
♦ Code Pages, 27

1
Overview
XML (Extensible Markup Language) is a flexible way to create common information formats
and share both the format and the data on the world wide web, intranets, and between
applications.
XML is similar to HTML because XML and HTML contain markup symbols to describe the
contents of a page or file. XML describes the content in terms of the data. HTML, however,
describes the content of a web page such as the text and graphics.
You can import XML definitions into PowerCenter from XML files, DTD files, and XML
schema files.
♦ An XML file contains data, and it can reference a Document Type Definition (DTD) or an
XML schema to describe the data.
♦ A Document Type Definition (DTD) file defines the element types, attributes, and
entities in an XML file. It provides some constraints on the XML document structure but
it does not contain any data.
♦ An XML schema file is called an XML Schema Definition (XSD). A schema file defines
elements and attributes, and it contains a description of the type of elements and attributes
in the associated XML file. Schemas contain simple and complex types. A simple type is an
XML element or attribute that can contains only text. A complex type is an XML element
that contains other elements and attributes.
In XML schemas, you can create element and attribute groups that you can reference
throughout a schema. You can also create substitution groups to substitute one element with
another element in the XML instance document.

2 Chapter 1: XML Concepts

XML Files
XML files contain tags that identify the data in a document, but not the format of the data.
The basic component of an XML file is an element. An XML element includes an element
start tag, content, and an element end tag. All XML documents must have a root element
defined by a single tag at the top and bottom of the document. The root element encloses all
the other elements in the file.
For example, the following XML file describes a book:
<book>
<title>Fun with XML</title>
<chapter>
<heading>Understanding XML</heading>
<heading>Using XML</heading>
</chapter>

<chapter>
<heading>Using DTD Files</heading>
<heading>Fun with Schemas</heading>
</chapter>
</book>

Book is the root element and it contains the title and chapter elements. Book is the parent
element of title and chapter, and chapter is the parent of heading. Title and chapter are sibling
elements because they have the same parent.
An element can have attributes that provide additional information about the element. In the
example below, the attribute graphic_type describes the content of file:
<file graphic_type="gif">computer.gif</file>

XML Files 3
 Figure 1-1 shows the structure, elements, and attributes in an XML file:

Figure 1-1. Sample XML File: StoreInfo.xml

Root Element Element Data

Enclosure
Element
Element Tags

Element Data

Attribute
Value

Attribute Tag

An XML file is a hierarchical structure. The XML hierarchy may contain the following
elements:
♦ Child element. An element contained within another element.
♦ Enclosure element. An element that contains other elements but does not contain data. It
can include other enclosure elements.
♦ Global element. An element that is a direct child of the root element. You can reference
global elements throughout an XML schema.
♦ Leaf element. An element that does not contain other elements. The lowest level element
in the XML hierarchy.
♦ Local element. An element that is nested in another element. You cannot reuse local
elements outside of the context of the parent element.
♦ Multiple-occurring element. An element that occurs more than once within its parent
element. Enclosure elements can be multiple-occurring elements.

4 Chapter 1: XML Concepts

♦ Parent chain. The succession of child-parent elements that traces the path from an
element to the root.
♦ Parent element. An element that contains other elements.
♦ Single-occurring element. An element that occurs once within its parent.
Figure 1-2 shows components in an XML hierarchy:

Figure 1-2. Elements in the XML Hierarchy

Enclosure Element:
Element Address encloses
elements StreetAddress, City,
State, and Zip. Element
Address is also a Parent
element.

Leaf Element:
Element Zip, along with all its
sibling elements, is the
lowest level element within
element Address.

Multiple-occurring Element:
Element Sales occurs more
than once within element
Product.
Single-occurring Element:
Element PName occurs only
once within element Product.

Child Element:
Element PName is a child of
product, which is a
descendant of Store.

Parent Chain:
Element YTDSales is a child of element Sales, which is a child of element
Product, which is a child of root element Store. All these elements belong in
the same parent chain.

XML Files 5
 A valid XML file conforms to the structure defined in an associated DTD or schema file. The
DOCTYPE declaration in an XML file references the location and name of a DTD file. It
also names the root element for the XML file. For example, the following DOCTYPE
declaration specifies the note.dtd file:
<?xml version="1.0"?>
<!DOCTYPE note SYSTEM
"http://www.w3schools.com/dtd/note.dtd">
<note>
<body>XML Data</body>
</note>

The schemaLocation declaration references the location and name of a schema. The following
XML file references an external schema:
<?xml version="1.0"?>
<note xsi:SchemaLocation="http://www.w3schools.com note.xsd">
<body>XML Data</body>
</note>

XML files contain an encoding declaration that indicates the code page in the file. The most
common code pages in XML are UTF-8 and UTF-16.

6 Chapter 1: XML Concepts

DTD Files
A Document Type Definition (DTD) file defines the element types, attributes, and entities in
an XML file. It also provides some constraints on the XML document structure. A DTD file
does not contain any data or element datatypes, but it can contain attribute datatypes.

DTD Elements
The following syntax describes a simple element in a DTD file:
<!ELEMENT product (#PCDATA)>

This DTD description defines the XML tag <product>. The description (#PCDATA) stands
for parsed character data. Parsed character data is plain text without child elements. You can
also use (#CDATA), which is character data. CDATA is not parsed or displayed.
An element that contains child elements has the following syntax in a DTD file:
<!ELEMENT boat (brand, type) >
<!ELEMENT brand (#PCDATA) >
<!ELEMENT type (#PCDATA) >

The boat element has two child elements: brand and type. Each child element can contain
characters. In this example, brand and type can occur once inside the element boat. To change
the number of possible occurrences use one of the following occurrence indicators:
♦ + must occur one or more times
♦ * may occur zero or more times
♦ ? may occur once or not at all
For example, to specify that color must occur one or more times for a boat:
<!ELEMENT boat (color+) >

DTD Attributes
The following syntax describes an attribute in a DTD file:
<!ATTLIST element_name attribute_name attribute_type “default_value”>

Use the following parameters to identify an element attribute in a DTD file:

♦ Element_name. The name of the element that has the attribute.
♦ Attribute_name. The name of the attribute.
♦ Attribute_type. The kind of attribute. The most common attribute type is CDATA.
♦ Default_value. The value of the attribute if no attribute value occurs in the XML file.
You can use the following options in place of a default value:
− #REQUIRED. The XML document must contain the attribute value.
− #IMPLIED. The attribute value is optional.

DTD Files 7
 − #FIXED. The XML document must use the default value from the DTD file. A valid
XML file can contain the same attribute value as the DTD, or it can have no attribute
value. You must specify a default value with this option. For example:
<!ATTLIST product product_name CDATA #FIXED “vacuum”>

Figure 1-3 shows elements and attributes in a DTD file:

Figure 1-3. Sample DTD: StoreInfo.dtd

Element
Element List

Element Occurrence

Attribute Value Option

Attribute

Attribute Name

8 Chapter 1: XML Concepts

XML Schema Files
The XML schema file, like the DTD file, contains only metadata. In addition to the
definition and structure of elements and attributes, an XML schema contains a description of
the type of elements and attributes in the associated XML file. An XML schema file is called
an XML Schema Definition (XSD).
Figure 1-4 shows XML schema components:

Figure 1-4. Sample XML Schema File: StoreInfo.xsd

Element Name

Attribute

Attribute Type
and Null
Constraint

Element
Datatype

Element Data

Element List
and
Occurrence

Element List and Datatype

When you use a schema to define an XML file, you can restrict data, define data formats, and
convert data between datatypes.

XML Schema Files 9

 XML schemas support inheritance in elements, complex types, substitution groups, attribute
groups, any content, and circular references.
For more information about XML schemas, see “Simple and Complex XML Types” on
page 17.
For information about element groups, see “Component Groups” on page 23.

10 Chapter 1: XML Concepts

Types of XML Metadata
You can create PowerCenter XML definitions from XML, DTD, or XML schema files. XML
files provide both data and metadata, while DTD and XML schema files provide only
metadata.
PowerCenter extracts the following types of metadata from XML, DTD, and XML schema
files:
♦ Namespace. A collection of element types and attribute names identified by a URI
reference in an XML document. Namespace differentiates between elements that come
from different sources. For more information about namespaces, see “Namespace” on
page 11.
♦ Name. A tag that contains the name of an element or attribute. For more information
about the name tag, see “Name” on page 13.
♦ Datatype. A classification of a data element, such as numeric, string, boolean, or time.
XML supports custom datatypes and inheritance. For more information about datatypes,
see “Datatype” on page 13.
♦ Hierarchy. The position of an element in relationship to other elements in an XML file.
For more information, see “Hierarchy” on page 13.
♦ Cardinality. The number of times an element occurs in an XML file. For more
information about cardinality, see “Cardinality” on page 14.
♦ Null constraint. Specification that determines if an element can contain null values. For
more information about the null constraint, see “Null Constraint” on page 16.

Namespace
An XML namespace identifies groups of elements. It can identify elements and attributes
combined from different XML documents into one file. For example, you can distinguish
meanings for the element table by declaring different namespaces, such as math:table and
furniture:table.
A namespace contains a Uniform Resource Identifier (URI) to identify schema location. A
URI is a string of characters that identifies an internet resource. It is an abstraction of a URL.
A URL locates a resource, but a URI identifies a resource.
The DTD or schema file does not have to exist at the location you specify in the URI. The
URI distinguishes between elements with the same name from different locations.
You can declare a namespace at the root level of an XML instance document, or you can
declare a namespace inside any element in an XML structure. A namespace declaration
appears in an instance document as an attribute that starts with xmlns. XML is case-sensitive,
so the namespace Math:table is different namespace from math:table.
When you declare multiple namespaces in the same instance document, you use prefixes to
associate an element to a namespace. Prefixes follow the namespace attribute and are declared
as xmlns:<prefix>. You can create a prefix name of any length, however, short prefixes are more

Types of XML Metadata 11

 efficient. The following example shows how to declare and use multiple namespaces in an
XML instance document:
<example>
xmlns:math = “http://www.mathtables.com”
xmlns:furniture = “http://www.home.com”>

<math:table> </math:table>
<furniture:table> </furniture:table>

</example>

If you associate a prefix with an attribute, the PowerCenter Server associates the attribute with
a namespace. If you do not associate a prefix, the PowerCenter Server use the default
namespace. An element cannot have two attributes with the same name unless the attributes
have different qualified names.
Qualified names are names that contain a namespace name. You create a qualified name by
using a prefix that is mapped to a namespace or by declaring a default namespace for an
element.
Each schema in the repository contains the default prefix for elements and attributes. For one
XML source or XML target, the list of prefixes and their namespaces should be unique. If a
duplicate prefix appears in the same instance document, the Designer makes the prefix unique
before storing it in the repository.
For example, the following is an example of a common schema declaration:
<?xml version="1.0"?>
<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"
targetNamespace="http://www.w3XML.com"
xmlns="http://www.w3XML.com"
elementFormDefault="qualified">...

...</xs:schema>

The fragment
targetNamespace="http://www.w3XML.com"

describes the namespace for the schema itself. The simple and complex datatypes that the
schema defines come from the "http://www.w3XML.com" namespace.
The fragment
xmlns:xs="http://www.w3.org/2001/XMLSchema"

indicates that the native XML schema elements and datatypes the schema uses come from the
“http://www.w3.org/2001/XMLSchema” namespace. The elements and data types that come
from the “http://www.w3.org/2001/XMLSchema” namespace have a prefix of xs:.
The fragment
xmlns="http://www.w3XML.com"

12 Chapter 1: XML Concepts

 indicates that the default namespace is "http://www.w3XML.com.” You can attach a default
namespace to an element and the child elements using an xmlns attribute with no prefix. The
element and its children are in the default namespace unless they have an explicit prefix.
The fragment
elementFormDefault="qualified"

indicates that any element that this schema declares must be namespace qualified in an XML
instance document.

Name
In an XML file, each tag is the name of an element or attribute. In a DTD file, the tag
<!ELEMENT> indicates the name of an element and the tag <!ATTLIST> indicates the set of
attributes for an element. In a schema file, <element name> indicates the name of an element
and <attribute name> indicates the name of an attribute.
When you import XML definitions into PowerCenter, the element tags become the column
names in the PowerCenter definition by default.

Datatype
The XML schema language has over 40 built-in datatypes, including numeric, string, time,
XML, and binary. These datatypes are called simple types. They contain text but no other
elements and attributes. You can derive new simple types from the basic XML simple types.
For more information about XML datatypes, see the W3C specifications for XML datatypes
at http://www.w3.org/TR/xmlschema-2.
You can create complex XML datatypes. A complex datatype is a datatype that can contain
more than one simple type. They can also contain other complex types and attributes. For
more information about simple and complex datatypes, see “Complex Types” on page 19.
XML files and DTD files do not store datatypes. When you import a source or target
definition from an XML file without an associated DTD or an XML schema, the Designer
assigns either a numeric or a string datatype to the elements. For information on how
PowerCenter uses XML datatypes, see “XML and Transformation Datatypes” on page 170.

Hierarchy
An XML document models a hierarchical database rather than a relational database. The
position of an element in an XML hierarchy represents its relationship to other elements. For
example, an element can contain child elements, and elements can inherit characteristics from
other elements.
When you import an XML definition in PowerCenter, the Designer creates a schema in the
repository. The schema models the hierarchy of the file you import. The Designer validates
the XML definition you create against the hierarchy of this schema.

Types of XML Metadata 13

 Cardinality
Cardinality is the number of members in a set. Element cardinality in a DTD or schema file is
the number of times an element might occur in an XML document. Element cardinality
affects how you can structure groups in an XML definition. Two types of cardinality affect an
XML definition: absolute cardinality and relative cardinality.

Absolute Cardinality
The absolute cardinality of an element is the number of times an element occurs within its
parent element in an XML hierarchy. DTD and XML schema files explicitly describe the
absolute cardinality of elements within the hierarchy. A DTD file uses symbols, and an XML
schema file uses the <minOccurs> and <maxOccurs> attributes to describe the absolute
cardinality of an element.
For example, an element has an absolute cardinality of Once(1) if it occurs once within its
parent element. However, it might occur many times within an XML hierarchy if its parent
element has a cardinality of one or more(+).
Table 1-1 describes the way DTD and XML schema files represent cardinality:

Table 1-1. Cardinality of Elements in XML

Absolute Cardinality DTD Schema

Zero or once ? minOccurs=0 maxOccurs=1

Zero or one or more times * minOccurs=0 maxOccurs=unbounded

minOccurs=0 maxOccurs=n

Once minOccurs=1 maxOccurs=1

One or more times + minOccurs=1 maxOccurs=unbounded

minOccurs=1 maxOccurs=n

14 Chapter 1: XML Concepts

Figure 1-5 shows the absolute cardinality of elements in a sample XML file:

Figure 1-5. XML Cardinality

Element Address occurs more

than once within Store. Its
absolute cardinality is One or
More(+).

Element City occurs only once

within its parent element
Address. Its absolute cardinality
is Once(1).

Element Sales occurs zero or

more times within its parent
element Product. Its absolute
cardinality is Zero or More(*).

Relative Cardinality
Relative cardinality is the relationship of an element to another element in the XML
hierarchy. An element can have a one-to-one, one-to-many, or many-to-many relationship to
another element in the hierarchy.
An element has a one-to-one relationship with another element if for every occurrence of one
element there is one occurrence of the other element. For example, an employee element can
have only one social security number element. Employee and social security number have a
one-to-one relationship.
An element has a one-to-many relationship with another element if for every occurrence of
one element, there can be multiple occurrences of the other element. For example, an
employee element might have multiple email elements.
An element has a many-to-many relationship with another element if an XML file can have
multiple occurrences of both elements. For example, an employee might have multiple email
addresses and multiple street addresses.

Types of XML Metadata 15

 Figure 1-6 shows the relative cardinality between elements in a sample XML file:

Figure 1-6. Relative Cardinality of Elements in StoreInfo.xml

One-to-Many Relationship
For every occurrence of
SNAME, there can be
many occurrences of
ADDRESS and, therefore,
many occurrences of
CITY.

Many-to-Many Relationship
There can be multiple
occurrences of STATE and
multiple occurrences of
YTDSALES.

One-to-One Relationship
For every occurrence of
PNAME, there is also
one occurrence of
PPRICE.

Null Constraint
The absolute cardinality of an element determines its null constraint. An element that has an
absolute cardinality of one or more(+) cannot have null values, but an element with a
cardinality of zero or more(*) can have null values. An attribute marked as fixed or required in
an XML schema or DTD file cannot have null values, but an implied attribute can have null
values. For more information about default, fixed, and implied attributes, see “DTD
Attributes” on page 7.
When you import an XML definition, the Designer sets the null constraints for the columns
depending on the absolute cardinality of the element or attribute the column points to. For
more information about absolute cardinality, see “Cardinality” on page 14.

16 Chapter 1: XML Concepts

Simple and Complex XML Types
Schema files contain simple and complex types. A simple type is a basic type that contains no
other elements or attributes. A complex type is an XML element that contains other elements
and attributes.

Simple Types
A simple type is an XML element or attribute that can contain only text. A simple type is
indivisible and it forms a column in an XML definition. Simple types cannot have attributes,
but attributes are simple types.
PowerCenter supports the following simple types:
♦ Atomic types
♦ Lists
♦ Unions

Atomic Types
An atomic datatype is a basic datatype in an XML schema definition such as a boolean, string,
integer, decimal, or date. You can define custom atomic datatypes by adding restrictions to an
atomic datatype in order to limit the content.
A facet is a specification of the values that are allowed or not allowed in a restriction. Facets
can specify minimum or maximum data values, a list of the legal values, or a data pattern. The
base attribute for the element specifies the datatype that the facet restricts.
The pattern facet restricts an element to an expression. The following example restricts a
string to one lowercase letter from a to z:
<xs:element name="letter">
<xs:simpleType>
<xs:restriction base="xs:string">
<xs:pattern value="[a-z]"/>
</xs:restriction>
</xs:simpleType></xs:element>

The enumeration facet lists all allowed values for an element. The following example restricts
a string to “a”, “b”, or “c”.
<xs:element name="letter">
<xs:simpleType>
<xs:restriction base="xs:string">
<xs:enumeration value="a"/>
<xs:enumeration value=”b”/>
<xs:enumeration value=”c”/>
</xs:restriction>
</xs:simpleType></xs:element>

Simple and Complex XML Types 17

 Lists
A list is an array collection of atomic types. For example, a list might include an array of sizes.
The following example shows a simple type called letter. It contains a list of strings. The list
itemType defines the datatype of the list components.
<xs:simpleType name="letter">
<xs:list itemType="xs:string" />
</xs:simpleType>

PowerCenter stores a list as one string that contains all array elements. It does not parse the
respective simple types from the string.

Unions
A union type is a combination of one or more atomic or list types that map to one simple type
in an instance document. When you define a union type, you specify what types to combine.
For example, you might create a type called Size. Size can include string data, such as S, M,
and L, or size might contain decimal sizes, such as 30, 32, and 34. If you define a union type
element, the XML file can include a sizename type for string sizes, and a sizenum type for
numeric sizes.

18 Chapter 1: XML Concepts

 Figure 1-7 shows a schema file containing a Shoesize union that contains sizenames and
sizenums lists:

Figure 1-7. Sample XSD File with Union Type

Sizename is a restricted
string type.

The sizenames type

accepts a list of strings.
The sizenums type
accepts a list decimals.
The ShoeSize union
accepts both the
decimal and string lists.

The union defines sizenames and sizenums as member types. Sizenames defines a list of string
values. Sizenums defines a list of decimal values. When you import an XML schema
containing this union, the Designer creates the port using a compatible datatype between the
types. In this case, the Designer creates the port as a string datatype because it can import
both strings and decimals as strings.

Complex Types
Complex types aggregate collections of simple types into a logical unit. For example, the
Customer type might include the customer number, name, street address, town, city, and zip
code. Complex types can also reference other complex types or element and attribute groups.
XML supports complex type inheritance. If you define a complex type, you can create other
complex types that inherit the components of the base type. In a type relationship, the base
type is the complex type from which you derive another type. A derived type is a complex
type that inherits elements from the base type.

Simple and Complex XML Types 19

 The derived type does not need to redefine the components it inherits from the base type. It
can extend or restrict the base type. An extended complex type inherits elements from a base
type and includes additional elements. For example, Customer Purchases might inherit its
definition from the Customer complex type, but it adds Item, Cost, and Date_Sold elements.
A restricted complex type restricts some elements from the base type. For example, Mail_List
might inherit elements from Customer, but restrict the Phone Number element by setting the
minoccurs and maxoccurs boundaries to zero.
Figure 1-8 shows an example of restricted and extended complex types:

Figure 1-8. Restricted and Extended Complex Types

Base Complex Type

Extended Complex
Type

Restricted Complex
Type

Element Reference

The base type is PublicationType. BookType extends the PublicationType and includes the
ISBN and Publisher elements.
Publication_Minimum restricts PublicationType. Publication_Minimun requires between 1
and 25 Authors, and restricts the date to just the year.

20 Chapter 1: XML Concepts

Element Type ANY
An element with a type of ANY can contain any type of data and any defined element. An
element that is ANY type is unrestricted. You can substitute it for any datatype you want. For
example, the following schema element describes a person. It includes a firstname, lastname,
and an ANY element:
<xs:element name="person">
<xs:complexType>
<xs:sequence>
<xs:element name="firstname" type="xs:string"/>
<xs:element name="lastname" type="xs:string"/>
<xs:any minOccurs="0"/>
</xs:sequence>
</xs:complexType>
</xs:element>

If you substitute the ANY type element for a complex type containing childnames, the XML
file could contain the following data:
<person>
<firstname>Danny</firstname>
<lastname>Russell</lastname>
<children>
<childname>Cissy</childname>
<childname>Cory</childname>
</children>
</person>

Abstract Elements
The abstract attribute specifies that an element cannot occur directly in an XML document. If
you set the abstract attribute to true, an XML instance document must contain a derived
element instead of the abstract type. The abstract attribute default value is false.
For example, the following schema specifies PublicationType as abstract. In an XML
document that references this schema, you must use the derived type, BookType. BookType
inherits the elements in PublicationType, but also includes ISBN and Publisher elements.
<xsd:complexType name="PublicationType" abstract="true">
<xsd:sequence>
<xsd:element name="Title" type="xsd:string"/>
<xsd:element name="Author" type="xsd:string" minOccurs="0"
maxOccurs="unbounded"/>
<xsd:element name="Date" type="xsd:gYear"/>
</xsd:sequence>
</xsd:complexType>

<xsd:complexType name="BookType">
<xsd:complexContent>
<xsd:extension base="PublicationType" >
<xsd:sequence>
<xsd:element name="ISBN" type="xsd:string"/>
<xsd:element name="Publisher" type="xsd:string"/>
</xsd:sequence>
</xsd:extension>

Simple and Complex XML Types 21

 </xsd:complexContent>
</xsd:complexType>

22 Chapter 1: XML Concepts

Component Groups
You can create the following types of groups in an XML schema:
♦ Element and attribute groups. Groups of elements and attributes that you can reference
throughout a schema.
♦ Substitution groups. Groups that you can use to replace one member element with
another member in an XML instance document. The substituted element must be in the
same complex or simple type derivation chain.

Element and Attribute Groups

You can assemble elements and attributes into groups that you reference throughout the
schema. You declare a group of elements or attributes and reference that group throughout
the document.
The following example shows the schema syntax for an element group:
<xs:group name="Songs">
<xs:element name="songTitle" type="xs:string" />
<xs:element name="artist" type="xs:string" />
<xs:element name="publisher" type="xs:string" />
</xs:group>

The following example shows the schema syntax for an attribute group:
<xs:attributeGroup name="Songs">
<xs:attribute name="songTitle" type="xs:string" />
<xs:attribute name="artist" type="xs:string" />
<xs:attribute name="publisher" type="xs:string" />
</xs:attributeGroup>

You can create the following element and attribute groups that have constraints:
♦ Sequence group. All elements in an XML file must occur in the order that the schema lists
them. For example, OrderHeader requires the customerName first, then orderNumber,
and then orderDate:
<xs:group name="OrderHeader">
<xs:sequence>
<xs:element name="customerName" type="xs:string" />
<xs:element name="orderNumber" type="xs:number" />
<xs:element name="orderDate" type="xs:date" />
</xs:sequence>
</xs:group>

♦ Choice group. Only one element in the group can occur in an XML document. For
example, only one element in the CustomerInfo group can occur in an XML document:
<xs:group name="CustomerInfo">
<xs:choice>
<xs:element name="customerName" type="xs:string" />
<xs:element name="customerID" type="xs:number" />
<xs:element name="customerNumber" type="xs:integer" />

Component Groups 23
 </xs:choice>
</xs:group>

♦ All group. All elements must occur in the XML document or none at all. The elements can
occur in any order. For example, CustomerInfo requires all or none of the three elements:
<xs:group name="CustomerInfo">
<xs:all>
<xs:element name="customerName" type="xs:string" />
<xs:element name="customerAddress" type="xs:string" />
<xs:element name="customerPhone" type="xs:string" />
</xs:all>
</xs:group>

Substitution Groups
Substitution groups allow you to replace one element with another in an XML instance
document. For example, if you have addresses from Canada and the United States, you might
want to create an address type for Canada and another type for the United States. You can
then create a substitution group that accepts either type of address.
The following schema section shows an Address base type, and the derived types,
CAN_Address, and USA_Address. CAN_Address has Province and PostalCode, and the
USA_Address has State and Zip. The MailAddress substitution group includes both address
types.
<xs:complexType name="Address">
<xs:sequence>
<xs:element name="Name" type="xs:string" />
<xs:element name="Street" type="xs:string"
minOccurs="1" maxOccurs="3" />
<xs:element name="City" type="xs:string" />
</xs:sequence>
</xs:complexType>

<xs:element name="MailAddress" type="Address" />

<xs:complexType name="CAN_Address">
<xs:complexContent>
<xs:extension base="Address">
<xs:sequence>
<xs:element name="Province" type="xs:string" />
<xs:element name="PostalCode" type="CAN_PostalCode"/>
</xs:sequence>
</xs:extension>
</xs:complexContent>
</xs:complexType>

<xs:complexType name="USA_Address">
<xs:complexContent>
<xs:extension base="Address">
<xs:sequence>
<xs:element name="State" type="USPS_StateCode" />
<xs:element name="ZIP" type="USPS_ZIP"/>
</xs:sequence>
</xs:extension>

24 Chapter 1: XML Concepts

 </xs:complexContent>
</xs:complexType>

<xs:element name="AddrCAN" type="CAN_Address"

substitutionGroup="MailAddress"/>

<xs:element name="AddrUSA" type="USA_Address"

substitutionGroup="MailAddress"/>

For more information about using Substitution Groups with PowerCenter, see “Using
Substitution Groups in an XML Definition” on page 45.

Component Groups 25
XML Path
The XML Path (XPath) represents the position of an element or attribute in the XML
hierarchy. In an XML definition, the XPath lists the path from an element or attribute to the
root. Except for generated key columns, each column in a group has an XPath and refers to an
attribute or an element in an XML hierarchy.
The XPath ensures integrity in the group column definition. PowerCenter uses a slash (/) to
depict the XPath of a column.
Figure 1-9 shows the XPath for different elements and attributes in a hierarchy:

Figure 1-9. XPath of Elements and Attributes in the StoreInfo.xml

STORE
STORE/SNAME

STORE/ADDRESS/STREETADDRESS

STORE/PRODUCT/PNAME

STORE/PRODUCT/SALES/YTDSALES

STORE/PRODUCT/@PID

STORE/PRODUCT/SALES/@REGION

26 Chapter 1: XML Concepts

Code Pages
XML files contain an encoding declaration that indicates the code page used in the file. The
most commonly used code pages in XML are UTF-8 and UTF-16. All XML parsers support
these two code pages. For information on the XML character encoding specification, go to the
W3C website at http://www.w3c.org.
PowerCenter supports the same set of code pages for XML files that it supports for relational
databases and other flat files. You can use any code page supported by both Informatica and
the XML specification. For a list of code pages that Informatica supports, see “Code Pages” in
the Installation and Configuration Guide. Informatica does not support any user-defined code
page.
For XML source definitions, PowerCenter uses the repository code page. When you import a
source definition from an XML file, the Designer displays the code page declared in the file
for verification only. It does not use the code page declared in the XML file.
For XML target definitions, PowerCenter uses the code page declared in the XML file. If
Informatica does not support the declared code page, the Designer returns an error. You
cannot import the target definition.

Code Pages 27
28 Chapter 1: XML Concepts
 Chapter 2

Using XML with PowerCenter

This chapter includes the following topics:

♦ Overview, 30
♦ Extracting XML Metadata, 31
♦ Working with XML Views, 36
♦ Generating Hierarchical View Relationships, 38
♦ Generating Entity View Relationships, 42
♦ Working with Circular References, 47
♦ Understanding View Rows, 49
♦ Pivoting Columns, 51
♦ Limitations, 54

29
Overview
You can create an XML definition in PowerCenter from an XML file, DTD file, XML
schema, flat file definition, or relational table definition. When you create an XML
definition, the Designer extracts XML metadata from the file or definition you choose and it
creates a schema in the repository. The schema provides the structure from which you create
and validate the XML definition.
An XML definition can contain multiple groups. In an XML definition, groups are called
views. The relationship between elements in the XML hierarchy defines the relationship
between the views. When you generate an XML definition, the Designer creates views for
multiple-occurring elements and complex types in a schema. The relative cardinality of
elements in an XML hierarchy affects how PowerCenter creates views in an XML definition.
Relative cardinality determines if elements can be part of the same view.
The Designer relates the views in an XML definition by keys. Source definitions do not
require keys, but target views must have them. Each view has a primary key, which can
contain an XML element or a generated key. The Designer defines relationships between
views with foreign keys. A foreign key in a view points to the primary key of the other view in
the relationship.
When you generate an XML definition, you can create a hierarchical model or an entity
relationship model. If you create a hierarchical model, you create normalized or denormalized
views. A normalized hierarchy contains separate views for multiple-occurring elements. A
denormalized hierarchy has one group with duplicate data for multiple-occurring elements.
If you create an entity model, the Designer creates views for multiple-occurring and complex
types. You can use simple types such as lists, unions, and substitution groups. You can also
model inheritance and circular relationships in PowerCenter definitions.

30 Chapter 2: Using XML with PowerCenter

Extracting XML Metadata
The Designer creates XML metadata from XML files, DTD files, and XML schemas. It also
creates XML metadata from flat file and relational definitions in the repository. When you
import an XML definition, the Designer creates a schema in the repository for the definition.
The repository schema provides the structure from which you edit and validate the XML
definition.

Metadata from an XML File

In an XML file, a pair of tags marks the beginning and end of each data element. These tags
are the basis for the metadata that PowerCenter extracts from the XML file. If you import an
XML file without an associated DTD or an XML schema, the Designer reads the XML tags to
determine the elements, their possible occurrences, and their position in the hierarchy. The
Designer checks the data within the element tags and assigns a string datatype or a numeric
datatype depending on the data representation. You can change the datatypes for these
elements in the XML definition.
Figure 2-1 shows a sample XML file. The root element is Employees. The Employee element
contains the LastName, FirstName, and Address. It also contains two multiple-occurring
elements: Phone and Email.

Figure 2-1. A Sample Employees XML File with Multiple-Occurring Elements

Employee, Phone, and

Email are multiple-occurring
elements.

Extracting XML Metadata 31

 The Designer determines a schema structure from the data. This example shows separate
views for the root element, and the multiple-occurring elements, Employee, Email, and
Phone.
Figure 2-2 shows a default XML source definition from the Employees XML file:

Figure 2-2. The Root Element and XML Views in an XML Definition

Root Element

Multiple-Occurring
Elements

When you import an XML file, you do not need all of the XML data to create an XML
definition. You need enough data to accurately show the hierarchy of the XML file.
Informatica recommends that you import an XML file that is under 100K. You can reduce the
size of a large XML file by deleting duplicate data elements.
The Designer can create an XML definition from an XML file that references an internal or
external DTD or an XML schema. The XML file uses a universal resource identifier (URI) to
refer to the address of an external DTD or an XML schema. If an XML file has a reference to
a DTD or an XML schema on another machine, the machine that hosts the PowerCenter
Client must have access to the machine where the schema resides so that the Designer can
read it.

Metadata From DTD Files

A DTD file provides constraints on a XML document structure. It lists elements, attributes,
entities, and notations that you can use in the XML document. It specifies relationships
between components. A DTD specifies cardinality and null constraint. However, a DTD file
does not contain any data and it does not specify datatypes.
When you import a DTD file you can change the datatypes for the elements in the XML
definition. You can change the null constraint, but you cannot change element cardinality.
If you import an XML file with an associated DTD, the Designer extracts the metadata based
on the DTD structure. However, you can import an XML file with elements and attributes
that include only a portion of the hierarchy defined in the DTD file.

32 Chapter 2: Using XML with PowerCenter

 Figure 2-3 shows an example of an XML file with a portion of the hierarchy from the
associated DTD and the source definition that the Designer creates:

Figure 2-3. XML Definition From an XML File Referencing a DTD File

StoreInfo.dtd contains the Store

element. Product is one of the child
elements of Store.

ProductInfo.xml uses only the Product

element from StoreInfo.dtd. Product
includes the multiple-occurring Sales
element.

The ProductInfo definition contains only the

Product and Sales groups. The XML file
determines what elements to include in the
definition. The DTD file determines the structure
of the XML definition.

Metadata from an XML Schema

A schema file defines the structure of elements and attributes in an XML file, but it also
contains a description of the type of elements and attributes in the file. When you import an
XML schema, the Designer determines the datatype, precision, and cardinality of the
elements. You cannot change them.
XML schemas can include complex types. Every attribute or simple type in a complex type
can become a column in an XML definition. Derived complex types inherit all the elements
from a base type. If you want to reduce metadata, you can create separate entities for the

Extracting XML Metadata 33

 complex types and the derived types do not contain the inherited components in the XML
view.
Figure 2-4 shows a schema containing a base type Address and a derived type CAN_address:

Figure 2-4. XML Schema with a Derived Complex Type

Address is a
Base Type.

CAN_Postal_
Code restricts a
string to a
pattern.

CAN_address
extends
Address.

The MailAddress element is an Address type. A derived type, Can_Address, inherits the
Name, City, and Street from the Address type, and it extends Address by adding a Province
and PostalCode. PostalCode is a simple type called CAN_Postal_Code.
Figure 2-5 shows an example XML definition that you might create from the schema if you
choose to import the schema with the default options:

Figure 2-5. XML Definition Containing a Derived Type

The Root element is Mail Address.

The Address type contains name, street, and

city.

The Canadian Address has a foreign key to

Address. Canadian Address includes Province
and PostalCode.

The Can_Address group contains only the elements that are unique for its type. The group
does not contain the Name, Street, and City that it inherits from MailAddress.

34 Chapter 2: Using XML with PowerCenter

Metadata from Relational Definitions
You can create an XML definition from relational source definitions. When you create an
XML definition from relational source definitions, you can select multiple definitions and
create relationships between them. The Designer creates an XML view for each relational
definition you import. It converts every column in the relational definition, and generates
primary key-foreign key relationships. It also creates a root element that encloses the views.
Figure 2-6 shows a sample XML definition from the relational definitions Orders and
Order_Items. The root is XRoot. It encloses Orders and Order Items. Order Items has a
foreign key that points to Orders.

Figure 2-6. XML Target Created from Two Relational Sources

Extracting XML Metadata 35

Working with XML Views
The relationship between elements in the XML hierarchy defines the relationship between
XML views. In a source definition, a view does not have to be related to any other view. A
denormalized view can be independent of any other view. Therefore, views in a source
definition do not require primary or foreign keys. However, if a view is related to another
view based on the XML hierarchy, and you do not designate any column as a key, the
Designer generates a key.
In a target definition, each view must be related to at least one other group. Therefore, each
view needs at least one key to establish its relationship with another view. If you do not
designate the keys, the Designer generates primary and foreign keys in the target views. You
can define primary and foreign keys for views if you create the views and relationships in the
XML Editor instead of allowing the Designer to create them for you.
When the Designer creates a primary or foreign key column, it assigns a column name with a
prefix. In an XML definition, the prefixes are XPK_ for a generated primary key column and
XFK_ for a generated foreign key column. The Designer uses the prefix FK_ for a foreign key
that points to a primary key.
For example, when the Designer creates a primary key column for the Sales group, it names
the column XPK_Sales. When it creates a foreign key column connecting a sales group to
another group, it names the column XFK_Sales. You can rename any column name that the
Designer creates. For an example of primary and foreign keys in a source group, see Figure 2-
8 on page 39.
If a mapping contains an XML source, the PowerCenter Server creates the values for the
generated primary key columns in the source definition when you run the session. You can
specify start values for the generated keys. For more information about setting generated key
values, see “Setting Sequence Numbers for Generated Keys” on page 128.

Rules for View Keys and Relationships

Consider the following rules and guidelines when you work with view keys and relationships:
♦ A group can have only one primary key.
♦ An enclosure element cannot be a key.
♦ A group can be related to several other groups, and can have multiple foreign keys.
♦ A column cannot be both a primary key and a foreign key.
♦ A source group does not require a key.
♦ A target group requires at least one key.
− The target root group requires a primary key. It does not require a foreign key.
− A target leaf group requires a foreign key. It does not require a primary key.
♦ A foreign key always refers to a primary key in another group. You cannot use self-
referencing keys.

36 Chapter 2: Using XML with PowerCenter

 ♦ A generated foreign key column always refers to a generated primary key column.
♦ The relative cardinality of elements in an XML hierarchy affects how PowerCenter creates
views in an XML definition. Relative cardinality determines if elements can be part of the
same view.
− Elements in a one-to-one relationship can be part of the same view.
− Elements in a one-to-many relationship can be part of the same normalized or
denormalized view.
− Elements in a many-to-many relationship cannot be part of the same view.

Creating Custom XML Views

Custom views are groups that you create with the XML Wizard or the XML Editor. If you use
the XML Wizard to create custom views, the wizard creates views containing all the
components in the schema. If you use the XML Editor, you can define each view and choose
the components.
The elements in the views and the relationship between views are dependent on the schema
the Designer creates in the repository when you import the definition. The XML Editor
validates XML definitions using the rules for valid views. For more information about rules
for view relationships, see “Rules for View Keys and Relationships” on page 36.
For more information about creating custom XML groups using the XML Editor, see
“Creating and Editing Views” on page 81.

Working with XML Views 37

Generating Hierarchical View Relationships
When you generate a hierarchical view relationship, the Designer expands each reference to a
component under its parent element. You can generate normalized or denormalized
hierarchical relationships.
The Designer uses the following guidelines when it generates hierarchical views:
♦ A normalized relationship reduces redundancy by separating multiple-occurring data into
separate views. The views are related by primary and foreign keys.
♦ A denormalized view contains all elements of the hierarchy. A source or target definition
can contain one denormalized view.

Generating Normalized Views

When the Designer generates normalized views for an XML definition, it establishes the root
element and the multiple-occurring elements that become views in an XML definition.
Figure 2-7 shows an example DTD file and the elements that become views in a normalized
XML definition:

Figure 2-7. Sample DTD File to Create Normalized Views

STORE: Root

ADDRESS+: Multiple-occurring
PRODUCT*: Multiple-occurring
EMPLOYEE+: Multiple-occurring

SALES*: Multiple-occurring

38 Chapter 2: Using XML with PowerCenter

Figure 2-8 shows a source definition based on the DTD file in Figure 2-7 with normalized
views. The root view is Store. The Address, Product, and Sales views have foreign keys to
Store. The Sales view has a foreign key to the Product view.

Figure 2-8. Source Definition for StoreInfo.dtd

STORE View

STORE Primary Key

Foreign Key to STORE

ADDRESS View

PRODUCT View

SALES View

EMPLOYEE View

Figure 2-9 shows a data preview for each view in the source definition:

Figure 2-9. Normalized Views Data Preview

STORE View

ADDRESS View

PRODUCT View

SALES View

EMPLOYEE View

Generating Hierarchical View Relationships 39

 Generating a Denormalized View
When the Designer generates a denormalized view, it creates one group and puts all elements
of the hierarchy into the group. All the elements in a denormalized group belong to the same
parent chain. Denormalized groups, like denormalized tables, generate duplicate data.
The Designer can generate denormalized views for XML definitions that contain more than
one multiple-occurring element if the multiple-occurring elements have a one-to-many
relationship and are all part of the same parent chain.
Figure 2-10 shows the ProdAndSales.dtd file, a sample XML source file that the Designer
converts to a denormalized view. The ProdAndSales.dtd file contains two multiple occurring
elements.

Figure 2-10. Sample XML File to Create a Denormalized View

STORE: Root

PRODUCT*: Multiple-occurring

SALES*: Multiple-occurring

Because the multiple-occurring elements have a one-to-many relationship, the Designer can
create a single denormalized group that includes all elements
Figure 2-11 shows the denormalized group for ProdAndSales.dtd in a source definition. The
group does not need a primary or foreign key.

Figure 2-11. Source Definition Containing a Denormalized View

The Designer creates a single group for all the elements in the
ProdAndSales hierarchy. Because a DTD file does not define
datatypes, the Designer assigns a datatype of string to all columns.

40 Chapter 2: Using XML with PowerCenter

Figure 2-12 shows a data preview for the denormalized view Store if the Designer imports the
source definition from ProdAndSales.xml with the associated ProdAndSales.dtd:

Figure 2-12. Data Preview for the ProdAndSales.xml Denormalized View

For an example of a denormalized group in a mapping, see “Adding an XML Source Qualifier
to a Mapping” on page 123.
For more information about generating hierarchy relationships using the XML Wizard, see
“Generating Hierarchy Relationships” on page 64.

Generating Hierarchical View Relationships 41

Generating Entity View Relationships
You can generate entity relationships from an XML schema. When you generate an XML
definition containing entity relationships, the Designer generates views for multiple-
occurring elements, element groups, and complex types in an XML schema. It creates links
and keys between the views based on type and hierarchy relationships.
When you work with XML schemas, you can use references to other parts of the schema
rather than repeat the same information in schema components. A component can inherit the
elements and attributes of another component and restrict or extend the elements from it. For
example, you might use a complex type as a base for creating a new complex type. You can
add more elements to the new type to create an extended complex type. Or, you might create
a restricted complex type, which is a subset of another complex type.
If you create groups manually or recreate entity relationships in the XML Editor, you choose
how you want to structure the metadata. If you create an XML definition based on an XML
schema that uses inheritance you can generate separate views for the base type and the derived
type. You might want to reduce metadata if you plan to map the XML data to normalized
relational tables.

Rules and Guidelines

The Designer generates entities based on the following guidelines:
♦ An entity represents a portion of an XML, DTD, or XML schema hierarchy. This
hierarchy does not need to start at the root of the instance document.
♦ The Designer uses entities defined in a DTD file to create entity relationships from a
DTD source.
♦ The Designer uses type structures defined in an XML schema to generate entity
relationships from an XSD source.
♦ The Designer creates a new entity when it encounters a multiple-occurring element under
a parent element.
♦ The Designer generates primary keys and foreign keys to relate separate entities.

Using Entity Relationships in an XML Definition

The following schema contains a PublicationType, BookType, and MagazineType.
PublicationType is the base type. It includes Title, Author, and Date. BookType and
MagazineType are derived types that extend the PublicationType. A Book has an ISBN and
Publisher, and a Magazine has a Volume and Edition.
<xsd:complexType name="PublicationType">
<xsd:sequence>
<xsd:element name="Title" type="xsd:string"/>
<xsd:element name="Author" type="xsd:string" maxOccurs="unbounded"/>
<xsd:element name="Date" type="xsd:string"/>

42 Chapter 2: Using XML with PowerCenter

 </xsd:sequence>
</xsd:complexType>

<xsd:element name="Publication" type="PublicationType"/>

<xsd:complexType name="BookType">
<xsd:complexContent>
<xsd:extension base="PublicationType">
<xsd:sequence>
<xsd:element name="ISBN" type="xsd:string"/>
<xsd:element name="Publisher" type="xsd:string
</xsd:sequence>
</xsd:extension>
</xsd:complexContent>
</xsd:complexType>

<xsd:complexType name="MagazineType">
<xsd:complexContent>
<xsd:extension base="PublicationType">
<xsd:sequence>
<xsd:element name="Volume" type="xsd:string"/>
<xsd:element name="Edition" type="xsd:string"/>
</xsd:sequence>
</xsd:extension>
</xsd:complexContent>
</xsd:complexType>
</xsd:schema>

When you generate XML views as entities in an XML definition, the Title and Date metadata
from PublicationType does not repeat in BookType or MagazineType by default. Instead,
these views contain only the metadata that distinguishes them from the PublicationType.
They have foreign keys that link them to the base type.
Author is a multiple-occurring element in a Publication. It becomes an XML view. This
example uses reduced metadata explosion. None of the elements in the base type repeat in the
derived types.

Generating Entity View Relationships 43

 Figure 2-13 shows the views that you might generate from the schema:

Figure 2-13. Complex Type View Relationships

Figure 2-14 shows a sample XML file containing a Publication, a Magazine, and two Books:

Figure 2-14. Complex Type XML Views Sample Schema

44 Chapter 2: Using XML with PowerCenter

 If you process this sample XML file using the XML definition in Figure 2-13 on page 44, you
create data in the following views:
♦ PublicationType view contains the Title and Date for each publication:

♦ BookType view contains the ISBN and Publisher. It contains a foreign key to
PublicationType:

♦ MagazineType view contains Volume and Edition. It also contains a foreign key to the
PublicationType:

♦ The Author view contains authors for all the publications. The Designer generates a
separate view for Author, because Author is a multiple-occurring element. Each
publication can contain multiple authors:

For more information about generating entity relationships using the XML Wizard, see
“Generating Entity Relationships” on page 63.
For more information about using the XML Editor to create relationships between XML
views, see “Creating Relationships Between Views” on page 89.

Using Substitution Groups in an XML Definition

When you import an XML schema that uses substitution groups, the Designer imports each
member of the substitution group as a separate view.

Generating Entity View Relationships 45

 Figure 2-15 shows a sample portion of an XML schema containing substitution groups:

Figure 2-15. Sample XML Schema Using Substitution Groups

Substitution Group
Members

Figure 2-16 shows an XML definition generated from substitution groups:

Figure 2-16. Sample XML Definition Generated from Substitution Groups

The Designer creates a new view for

each member of the substitution
group.

46 Chapter 2: Using XML with PowerCenter

Working with Circular References
A circular reference relationship is a circular hierarchy relationship between two views in an
XML definition. A view can also refer to itself recursively. For example, an complex element
called Part might contain an ID, part name, and a reference to another Part.
The following example shows the components for the Part element:
<xs:element name="Part">
<xs:complexType>
<xs:sequence>
<xs:element name="ID" type="xs:string"/>
<xs:element name="Name" type="xs:string"/>
<xs:element name="Type" type="xs:string"/>
<xs:element ref="Part" minOccurs="0" maxOccurs="unbounded"/>
/xs:sequence>
</xs:complexType>
</xs:element>

Figure 2-17 shows a circular reference in the XML Editor workspace:

Figure 2-17. Circular Reference View

You might use the Part XML definition to read the following sample XML file in a session:
<Part>
<ID>1</ID>
<Name>Big Part</Name>
<Type>L</Type>
<Part>
<ID>1.A</ID>
<Name>Middle Part</Name>
<Type>M</Type>
<Part>
<ID>1.A.B</ID>
<Name>Small Part</Name>
<Type>S</Type>
</Part>
</Part>
</Part>

In this sample file, Part 1 contains Part 1.A, and Part 1.A contains Part 1.A.B.

Working with Circular References 47

 Figure 2-18 shows the data and the keys that a session might generate from the XML source.

Figure 2-18. Circular Reference Data Preview

48 Chapter 2: Using XML with PowerCenter

Understanding View Rows
To extract data from an XML document, you must specify what rows to generate, what
columns of data to include, and when to generate the row. When you define a view in the
XML Editor, you create the view row, an element or a global complex type that the
PowerCenter Server requires to generate a row of data.
The PowerCenter Server uses a view row to determine when to read and write XML. You can
set a view row at any single or multiple-occurring element. Once you set the view row, every
element you add to the view has a one-to-one correspondence with the view row.
For example, in the view Employees, which contains elements Employee, Name, Firstname,
and Lastname, if you set the view row to Employee, the PowerCenter Server extracts XML
using an algorithm similar to this for-loop:
For every (Employees/Employee)

extract ./Name/Firstname/Lastname

An Employees XML schema might contain the following elements:

EMPLOYEES
EMPLOYEE+
ADDRESS+
NAME
FIRSTNAME
LASTNAME
EMAIL+

Employee, Address, and Email are multiple-occurring elements. You could create a view that
contains the following elements:
EMPLOYEE
ADDRESS
NAME

If you set the view row as Address, the PowerCenter Server extracts a Name for every
Employee/Address in the XML data. You cannot add Email to this view because you would
create a many-to-many relationship between Address and Email.
You can add a pivoted multiple-occurring column to the view. For example, you can add one
instance of Email as a pivoted column to the Employee view. The view would contain the
following elements:
EMPLOYEE
ADDRESS
NAME
EMAIL[1]

A view row can contain a pivoted column. For example, a view might have the view row,
EMPLOYEE/ADDRESS[1]. The PowerCenter Server extracts data for just the first instance of
Employee/Address. For more information about pivoting, see “Pivoting Columns” on
page 51.

Understanding View Rows 49

 The following rules apply to view rows in an XML definition:
♦ A view row must be a type or an element. It cannot be an attribute.
♦ Every view must have a view row, which must be an element or complex type.
♦ The view root is the top level element in a view. It is the parent to all the other elements in
the view.
♦ The view row can be the same as the view root unless the view is denormalized.
♦ Two views can have same view row in an XML source or XML Parser transformation.
♦ The view row element must be the lowest multiple-occurring element in the view. A view
cannot contain many-to-many relationships.
♦ If you add a multiple-occurring element to a view with no other multiple-occurring
elements, you change the view row to the new element by default. If the view already has a
multiple-occurring element, you cannot add another multiple-occurring element.
♦ You do not need to specify a view row when you create an empty view. However, as soon as
you add a column to the view, the Designer creates the view row. This is true even if you
are only adding a primary key.
♦ You can change a view row at a later time, but you cannot change a view root unless there
are no schema components in the view.
♦ You can specify a view row that consists of a pivoted element, such as:
Product/Order[2]/Customer

♦ An effective view row for a view is the path of view rows from the top of a hierarchy
relationship down to the view row in the view. A view can have multiple effective view
rows because it can have multiple hierarchy relationships in the XML definition.
You can specify options in the XML Editor that affect how view rows and effective view rows
affect data output. For more information about setting row options, see “Setting XML View
Options” on page 100.

50 Chapter 2: Using XML with PowerCenter

Pivoting Columns
Sometimes an element that occurs multiple times is a set of the same elements containing
different values. For example, an element called Sales that occurs 12 times might contain the
sales figures for each month of the year. Or, an element called Address that occurs twice might
be a home address and an office address.
If you have this type of element in your XML source and you treat the occurrences as separate
columns in the source definition, you can use a method called pivoting. Pivoting enables you
to treat occurrences of elements as columns in a group.
To pivot occurrences of an element in an XML view, create a column for each occurrence you
want to represent in the definition. In the monthly sales example, if you want to represent all
12 occurrences as columns, create 12 sales columns in the view. If you want to represent the
sales of one quarter, create three columns. When you run a session, the PowerCenter Server
ignores any XML data for the occurrences that you do not include in the definition.
You can pivot columns when you add or edit a view in the XML source definition. For more
information, see “Adding Columns to Views” on page 82.
You can pivot simple types and complex types. You cannot pivot a primary key column.
When you pivot columns in a view, the resulting group structure must follow the rules for a
valid normalized or denormalized view. The Designer displays warnings and errors if the
pivoted columns invalidates a view.
Pivoting affects an element only in the view where you pivot the element. It does not affect
the same element in another view.
Note: You cannot pivot columns in an XML target.

Pivoting Columns 51
 Figure 2-19 shows the ADDRESS element of the StoreInfo XML file pivoted into two sets of
address columns:

Figure 2-19. ADDRESS Element Occurrences Pivoted into Columns

First occurrence of Address

pivoted to home address
columns with prefix HOM_.

Second occurrence of
Address pivoted to office
address columns with
prefix OFC_.

XPath shows the two sets of columns

that come from the same elements.

The first and second occurrences of Address represented as columns in the group:

Using Multiple-Level Pivots

You can pivot more than one level of elements in a view by specifying fixed offsets for a
multiple-occurring elements in the XPath for a column. For example, you might have the
following elements in a view:

52 Chapter 2: Using XML with PowerCenter

 STORE
PRODUCT+
PNAME
ORDER+
ORDERNAME
CUSTOMER+
CUSTNAME

The XPath STORE/PRODUCT[2]/ORDER[1]/ORDERNAME refers to the ordername for the first

order for the second product in the store. The XPath STORE/PRODUCT[2]/ORDER/
CUSTOMER[1]refers to the first customer for all orders of the second product.

If you pivot a view row, any column in the XML view that occurs below the view row must
have an XPath that matches XPath of the view row.
For example, a view might have the following view row:
Transaction/Trade[1]

The following columns have the same occurrence of Trade in the XPath:
Transaction/Trade[1]/Date
Transaction/Trade[1]/Price
Transaction/Trade[1]/Person[1]/Firstname

You cannot create a column with the following XPath in the view:
Transaction/Trade[2]/Date

Pivoting Columns 53
Limitations
PowerCenter does not support the following functions:
♦ Concatenated columns. A column cannot be a concatenation of two elements. For
example, you cannot create a column FULLNAME that refers to a concatenation of two
elements FIRSTNAME and LASTNAME.
♦ Composite keys. A key cannot be a concatenation of two elements. For example, you
cannot create a key CUSTOMERID that refers to a concatenation of two elements
LASTNAME and PHONENUMBER.
♦ Parsed Lists. PowerCenter stores a list type as one string that contains all array elements. It
does not parse the respective simple types from the string.

54 Chapter 2: Using XML with PowerCenter

 Chapter 3

Working with XML

Sources
This chapter includes the following topics:
♦ Overview, 56
♦ Importing an XML Source Definition, 57
♦ Working with XML Views, 61
♦ Generating Entity Relationships, 63
♦ Generating Hierarchy Relationships, 64
♦ Creating Custom XML Views, 65
♦ Synchronizing XML Definitions, 68
♦ Editing XML Source Definition Properties, 70
♦ Creating XML Definitions from Repository Definitions, 72
♦ Troubleshooting XML Sources, 74

55
Overview
The Designer provides an XML Wizard that enables you to create XML definitions in the
repository. You can import files from a URL or a local machine. You can also import relational
or flat file definitions from a PowerCenter repository. You can create XML definitions from
the following types of files:
♦ XML files
♦ XML schema files
♦ DTD files
♦ Relational definitions
♦ Flat file definitions
Note: Informatica recommends you use an XML schema file to create XML source definitions.

When you create XML definitions, you import files with the XML Wizard and organize
metadata into XML views. XML views are groups of columns containing the elements and
attributes in the XML file. The wizard can generate views for you, or you can create custom
views. For more information see “Working with XML Views” on page 61.
You use the XML Wizard to create relationships between views. You can create hierarchy
relationships or entity relationships. For more information about creating hierarchy
relationships, see “Generating Hierarchy Relationships” on page 64. For more information
about creating entity relationships, see “Generating Entity Relationships” on page 63.
You can use the XML Wizard to synchronize an XML definitions against an XML schema file
if the structure of the schema changes. For more information, see “Synchronizing XML
Definitions” on page 68.

56 Chapter 3: Working with XML Sources

Importing an XML Source Definition
When you import a source definition from a DTD or an XML schema file, the Designer can
provide an accurate definition of the data based on the description provided in the DTD or
XML schema file.
When you import a source definition based on an XML file without an associated DTD or
XML schema, the XML Wizard determines the types and occurrences of the data based on
data represented in the XML file. To ensure that the wizard can provide an accurate definition
of the data, import from an XML source file that represents the occurrence and type of data as
accurately as possible. Once you create an XML source definition, you cannot change it to
any other source type. Conversely, you cannot change other types of source definitions to
XML definitions.
The XML Wizard uses keys to relate the XML views and reconstruct the XML hierarchy. You
can choose to generate views and primary keys, or you can create views and specify keys.
When you create custom views, you can select roots and choose how to handle metadata
expansion.
The XML Wizard saves the XML hierarchy and the view information as an XML schema in
the Repository. When you import an XML definition, your ability to change the cardinality
and datatype of the elements in the hierarchy depends on the type of file you are importing.
For example, DTD and XML files do not store datatype information. When you import these
files to create an XML definition, you can specify datatype, precision, and scale. If you import
an XML schema, you can change the precision and scale.
You cannot create an XML source definition from an XML file of exported repository objects.
For more information about exporting and importing repository objects, see “Exporting and
Importing Objects” in the Repository Guide.
When you import an XML source definition, the Designer uses the encoding declaration in
the XML file. For information on PowerCenter support for XML code pages, see “Code
Pages” on page 27.

Importing an XML Source Definition 57

 Use the XML Wizard to import XML source definitions.

To import an XML file:

1. Click Sources-Import XML Definition.

The Import XML Definition dialog box displays.

Select options to create XML views.

2. Click Advanced Options.

58 Chapter 3: Working with XML Sources

 The XML Views Creation dialog box opens:

3. Use Advanced Options to specify the way the Designer creates and names XML views.
You can change the following options:

Table 3-1. Create XML Views Options

Option Description

Override all infinite lengths You can specify a default length for components with undefined lengths,
such as strings. If you do not set a default length, the precision for these
components sets to infinite. This can cause DTM buffer size errors when
you run a session with large files.

Analyze elements/attributes in Choose this option to create global declarations of standalone XML
standalone XML as global elements or attributes. Otherwise, they become local declarations.
declarations

Create an XML view for an enclosure If the schema has an enclosure element that can occur more than once
element and the enclosure element has child elements that can occur more than
once, you can create a separate view for it. An enclosure element is an
element that has no text content or attributes but has child elements.

Pivot elements into columns You can pivot leaf elements if they have an occurrence limit. You can
pivot elements in source definitions only.

Ignore fixed element and attribute You can ignore fixed values in a schema and allow other element values
values in the data.

Importing an XML Source Definition 59

 Table 3-1. Create XML Views Options

Option Description

Ignore prohibited attributes In an XML schema or file, you can declare an attribute as prohibited.
This allows you to restrict a complex type by prohibiting existing
attributes. When you import the schema or file, you can choose to ignore
the prohibited attributes.

Generate names for XML columns You can choose to name XML columns by a sequence number or from
the element or attribute name in the schema. If you use names, you can
add the XML view as a prefix to each column, and you can add the
element name as a prefix to all the attributes.

4. Click OK to apply changes.

5. Choose the type of file to import. You can choose the following options:
♦ Import the definition from a local XML file or a URL. Create a source definition
from an XML, DTD, or XML schema file. If you import an XML file with an
associated DTD or schema, the XML Wizard uses the DTD or schema to generate the
XML document.
♦ Import the definition from a non-XML source or target. Use this option to create a
source definition from flat file or relational definitions. The new source definition
contains one group for each input definition plus a root element group. For more
information, see “Creating XML Definitions from Repository Definitions” on
page 72.
6. Click Next to complete the XML Wizard.

Multi-line Attributes Values

The XML Wizard does not allow attribute values that contain new line characters and span
more than one line. When you import a source or target definition from an XML file that
contains an attribute with new line characters, the XML Wizard displays an error and does
not import the file.

60 Chapter 3: Working with XML Sources

Working with XML Views
The Designer displays views as groups in the source definition. For more information about
how the Designer generates groups, see “Working with XML Views” on page 36.
Figure 3-1 shows the options for creating views in the XML Wizard:

Figure 3-1. The XML Wizard Options to Create Views

You can choose from the following options to create XML views:
♦ Generate entity relationships. If you create entity relationships, the XML Wizard
generates views for multiple-occurring or referenced elements and complex types. For
information about generating entity relationships, see “Generating Entity Relationships”
on page 63.
♦ Generate hierarchy relationships. When you create hierarchical relationships, each
reference to a component expands under its parent element. You can generate normalized
or denormalized XML views in a hierarchy relationship.
− Normalized XML views. When you generate a normalized XML view, elements and
attributes appear only once. Multiple-occurring elements, or elements in one-to-many
relationships appear in different views related by keys. For more information about how
the Designer generates normalized views, see “Generating Normalized Views” on
page 38.
− Denormalized XML views. When you generate a denormalized XML view, all elements
and attributes appear in one view. The Designer does not model many-to-many
relationships between elements and attributes in an XML definition. For more
information about how the Designer generates denormalized views, see “Generating a
Denormalized View” on page 40.

Working with XML Views 61

 ♦ Create a custom XML views. You can specify any global element as a root when creating a
custom XML view. You can choose to reduce metadata explosion for elements, complex
types, and inherited complex types. For information about creating custom views, see
“Creating Custom XML Views” on page 65.
♦ Synchronize XML definitions. You can update one or more XML definitions when their
underlying schemas change. For information about synchronizing XML definitions, see
“Synchronizing XML Definitions” on page 68.
♦ Do not create XML views. When you choose not to create XML views, you can create
them later in the XML Editor. For more information about creating views using the XML
Editor, see “Creating and Editing Views” on page 81.
Once you create an XML view, you cannot change the configuration options you set for the
view. For example, if you create a normalized XML view, you cannot change it to
denormalized. You must import a new XML source definition and select the denormalized
option.

62 Chapter 3: Working with XML Sources

Generating Entity Relationships
You can generate an XML hierarchy as an entity relationship model. When you generate XML
views as entity relationships, the Designer completes the following actions:
♦ Generates views for multiple-occurring and referenced elements and complex types.
♦ Creates relationships between the views.
When the Designer generates entity relationships, it generates different entities for complex
types, global elements, and multiple-occurring elements based on the relationships in the
schema. For a description of how the Designer generates Entity Relationships, see
“Generating Entity View Relationships” on page 42.
If you want to create different groups from the default groups, or if you want to combine
elements from different complex types, you can create custom XML views. For more
information about creating custom views, see “Creating Custom XML Views” on page 65.
When you view an XML source definition in the XML Editor, you can see the relationships
between each element in the XML hierarchy. For each relationship between views, the XML
Editor generates links based on the type of relationship between the views.

To generate entity relationships:

1. In the Source Analyzer, select Sources-Import XML Definition. The XML Wizard opens.
2. Navigate to the source you want to import and click Open.
3. Enter a name for the file and click Next.
4. Select Entity Relationships and click Finish.
The XML Wizard generates an XML definition that uses entity relationships.

Generating Entity Relationships 63

Generating Hierarchy Relationships
When you create hierarchy relationships, each reference to a component is expanded under its
parent element. The XML Wizard selects the default root and uses default settings to create
XML groups. For information about default settings for XML groups, see “Generating
Hierarchical View Relationships” on page 38.

To generate hierarchy relationships:

1. In the Source Analyzer, select Sources-Import XML Definition. The XML Wizard opens.
2. Navigate to the source you want to import and click Open.
3. Enter a name for the file and click Next.
4. Select Hierarchy Relationships.

5. Select Normalized XML Views or Denormalized XML Views and click Finish.
The XML Wizard generates XML groups using hierarchy relationships.

64 Chapter 3: Working with XML Sources

Creating Custom XML Views
You can create custom XML views using the XML Wizard. When you create custom views,
you can choose roots and specify how to generate metadata. You can choose to include or
exclude global elements based on whether the root information applies to the data you intend
to process. For example, if your schema contains information about stores and customers, you
might want to create an XML definition that processes just the customers.
You can specify how you want to generate metadata associated with the view. You can reduce
metadata explosion for elements, complex types, and inherited complex types by generating
entity relationships. If you do not reduce the metadata references, the Designer generates
hierarchy relationships and expands all child elements under their parent elements.
For more information about how the Designer generates hierarchy and entity relationships,
see “Working with XML Views” on page 36.

To create a custom view using the XML Wizard:

1. In the Source Analyzer, select Sources-Import XML Definition.

The XML Wizard opens.
2. Navigate to the source you want to import and click Open.
3. Enter a name for the file and click Next.
4. Select Create Custom XML Views and click Next
Note: To manually create all the XML views in the XML Editor, select Skip Create XML
Views. The XML Wizard creates the schema in the Repository, but does not create the
XML views. For more information about creating views in the XML Editor, see “Creating
an XML View” on page 81.

Creating Custom XML Views 65

 5. Select root elements from the list of global root elements and click Next.
For information about selecting root elements, see “Selecting Root Elements” on page 66.
6. Choose to reduce metadata for elements, complex types, or inherited complex types and
click Finish.
For information about reducing metadata explosion, see “Reducing Metadata Explosion”
on page 66.

Selecting Root Elements

When you create a custom view, you can choose from the global elements in the imported
schema to set the root for the XML instance document. Global elements are elements in an
XML schema hierarchy that fall directly beneath the top root element.
Figure 3-2 shows the Root selection page:

Figure 3-2. Root Selection Page

Bookstore
element
selected as root.

Book123
element cleared
as root element.

Reducing Metadata Explosion

When the Designer creates an XML definition based on an XML schema that uses
inheritance, it can expand the metadata for each referenced element or group within the view
that references it. Or, it can create a separate view for the referenced object and create
relationships between it and other views.

66 Chapter 3: Working with XML Sources

If you use references within an XML schema, you might want to reduce the number of times
the Designer includes the metadata associated with a reference. The XML Wizard provides
the following options to reduce metadata references:
♦ Reduce element explosion. The Designer creates a view for any multiple-occurring
element or any element that is referenced by more than one other element. Each view can
have multiple hierarchical relationships with other views in the definition.
♦ Reduce complex type explosion. The Designer creates an XML view for any referenced
complex type or multiple-occurring element. The XML view can have multiple type
relationships with other views. If the schema uses inherited complex types, you can also
reduce explosion of inherited complex types.
♦ Reduce complex type inheritance explosion. For any inherited type, the XML Wizard
creates a type relationship.
Figure 3-3 shows the Reduce Metadata Explosions page:

Figure 3-3. Reduce Metadata Explosions Page

When you reduce metadata explosion, the Designer creates entity relationships between the
XML views it generates.

Creating Custom XML Views 67

Synchronizing XML Definitions
When you work with XML definitions, the files or sources you used to create the XML
definition might change. For example, you might add a new element or complex type to an
XSD file. You can synchronize an XML definition with any of the following repository
definitions or files you used to create the XML definition:
♦ Relational source definitions
♦ Relational target definitions
♦ Flat files
♦ URLs
♦ XML files
♦ DTDs
♦ Schema files
Note: Informatica recommends you use schema files to synchronize XML definitions.

To synchronize XML source definitions:

1. In the Source Analyzer, select Sources-Import XML Definition. The XML Wizard opens.
2. Navigate to the repository definition or file that you used to create a source or target
definition, and click Open.
3. In Step 1 of the Wizard, click Next. The Wizard ignores any change you make to the
name.
4. In Step 2 of the XML Wizard, choose to synchronize the XML definition and click Next.

68 Chapter 3: Working with XML Sources

5. The XML Wizard skips to Step 5. In Step 5 of the XML Wizard, choose the XML
definition you want to synchronize.

The XML Wizard synchronizes the source with the selected definition.
You can use this method to synchronize XML target definitions. If you modify an XML
source definition, you may also need to synchronize the target definition.
Note: Verify that you synchronize the XML definition with the source you used to create the
definition. If you synchronize an XML definition with a source that you did not use to create
the definition, the Designer cannot synchronize the definitions and loses it metadata. Choose
Edit-Revert to Saved to restore the XML definition.

Synchronizing XML Definitions 69

Editing XML Source Definition Properties
After you import an XML source definition, you can edit the source definition properties
such as the definition name. You also add metadata extensions.

To edit XML source definition properties:

1. Right-click the source in the Source Analyzer and select Edit.

2. On the Table tab, edit the following settings as necessary:

Table Settings Description

Rename button Edit the name of the source definition and enter a business name.

Business Name Descriptive name for the source definition. You can edit by clicking the Rename
button.

Owner Name Not applicable for XML files.

Description Optional description of the source. Character limit is 2,000 bytes/K, where K is
the maximum number of bytes for each character in the repository code page.
Enter links to business documentation.

Database Type Indicates the source or database type.

Code Page Indicates the repository code page. The XML source definition uses the code
page of the repository, not the code page defined in the XML source file.

3. On the Columns tab, you can view information about the columns in the definition. To
change column names or values, use the XML Editor.

70 Chapter 3: Working with XML Sources

 You can view the following information:

Columns Settings Description

Select Table Displays the source definition you are editing.

Column Name The name of the column.

Datatype Displays datatype.

Precision Displays length of data.

Scale Displays number of decimal positions for numerical data.

Not Null Indicates whether the column can accept nulls.

Key Type When the Designer creates default views for XML sources, it generates the
primary and foreign keys.

XPath Indicates the element referenced by the current column in the XML hierarchy.
XPath does not display for generated primary or foreign keys.

Business Name User-defined descriptive name for the column. If it is not visible in the window,
scroll to the right to view or modify the column.

4. Click the Metadata Extensions tab to create, edit, and delete user-defined metadata
extensions. For more information about metadata extensions, see the Repository Guide.
5. Click OK.
6. Choose Repository-Save to save changes to the repository.
Note: You cannot edit the Properties tab for an XML source definition.

Editing XML Source Definition Properties 71

Creating XML Definitions from Repository Definitions
You can import an XML source or target definition from relational or flat file definitions in
the repository. When you import an XML definition from a repository definition, the XML
Wizard creates an XML hierarchy from the relationships between the selected objects. The
XML Wizard creates a root element for the hierarchy. You can choose a root from the groups
you create, or you can create a separate root and relate the groups to it.
When you create an XML target, the XML Wizard generates keys to relate each group to the
root.

To create an XML definition from repository sources or targets:

1. In the Source Analyzer, click Sources-Import XML Definition.

or
In the Warehouse Designer, click Targets-Import XML Definition.
2. In the XML Import dialog box, click either Sources or Targets.

Enter a root name.

Selected Source or Target

Definitions
Add a definition.

List of Repository Sources

and Targets

Choose input type.

3. Select a definition from the list of sources or targets. Click the Arrow button to add the
definition to the selected source list.
You can select more than one input, and more than one input type. If the definitions are
related through primary and foreign keys, the XML Wizard uses the keys to relate groups
when it generates the hierarchy.
4. Click Open. The XML Wizard displays.
5. Enter the Optional XML Rootname to have the XML Wizard create a separate group for
the root element and relate all other groups to it. The root name defaults to XRoot.
Remove the root name if you want to use one of the other groups as the root. The XML
Wizard creates a group for each input source or target definition you select and generates

72 Chapter 3: Working with XML Sources

 a primary key in each group. It creates a foreign key in each group that points to the root
group link key.
6. Use the XML Wizard to generate the source or target groups.

Creating XML Definitions from Repository Definitions 73

Troubleshooting XML Sources
How can I put two multiple-occurring elements that both have the same parent element
into one view? For example, I would like to put all the elements of EMPLOYEE in one
view:
<!ELEMENT EMPLOYEE (EID, EMAIL+, PHONE+)>

Although in this example EMAIL and PHONE belong to the same parent element, they do
not belong in the same parent chain. Ordinarily, you cannot put them in the same
denormalized view.
To put all the elements of employee in one view, you need to pivot one of the multiple
occurring elements. For example, you can create an EMPLOYEE view and follow these steps
to add all the elements:
1. Add the EID and EMAIL elements to the EMPLOYEE view.
2. Pivot the number of occurrences of EMAIL that you want to include. If you get warnings
while pivoting any of the occurrences, confirm that you want to proceed.
3. Add the PHONE element.
Pivoting in this case turns an element into a single occurring element in a view. After you
pivot EMAIL, the Groups At element for the EMPLOYEE view is PHONE because it
becomes the only multiple-occurring element in the view.

I have the following element definition in my DTD:

<!ELEMENT EMPLOYEE (EMPNO, SALARY)+>

How can I match the EMPNO and SALARY in the same view?
This example of a DTD element definition is ambiguous. It is equivalent to the following:
<!ELEMENT EMPLOYEE (EMPNO+, SALARY+)>

In this definition, there appears to be an EMPNO and a SALARY for each employee.
However, when the definition is written this way, the number of occurrences for EMPNO is
separate from and may not match the number of occurrences of SALARY.
You can try either of the following solutions:
♦ Rewrite the element definition to make it unambiguous.
In most cases, the EMPLOYEE element is more correctly defined in this way:
<!ELEMENT EMPLOYEES (EMPLOYEE+)>
<!ELEMENT EMPLOYEE (EMPNO, SALARY)>

Redefined this way, there is one EMPNO and one SALARY for each EMPLOYEE, and
both elements go into the same EMPLOYEE view.

74 Chapter 3: Working with XML Sources

♦ Leave the elements in separate views and use the source definition twice in a mapping.
Even when EMPNO and SALARY are in different views, you can still combine the data in
a mapping. Use two instances of the same source definition and join them in a Joiner
transformation. For an example of using two instances of an XML source in a mapping, see
“Using the XML Source Qualifier in a Mapping” on page 129.

I imported an XML file with the following structure:

<Bookstore>
<Book>Book Name</Book>

<Book>Book Name</Book>

<ISBN>051022906630</ISBN>
</Bookstore>

When I import this file, the Designer drops the ISBN element. Why does this happen?
How can I get the Designer to include the ISBN element?
♦ Use the schema to import the XML definition. When you use an XML file to import an
XML definition, the Designer reads the first element as simple content because it has no
child elements. Based on this, the Designer determines that the second instance of Book is
also simple content and discards the child element, ISBN. If you use a schema to import
the definition, the Designer uses the structure defined in the schema to determine how to
read XML data.
♦ Ensure that the XML instance document accurately represents the associated schema. If
you use an XML file to import a source definition, ensure that the document is an accurate
representation of the structure in the corresponding XML schema.

Troubleshooting XML Sources 75

76 Chapter 3: Working with XML Sources
 Chapter 4

Using the XML Editor

This chapter includes the following topics:

♦ Overview, 78
♦ Creating and Editing Views, 81
♦ Maintaining View Relationships, 89
♦ Viewing Schema Components, 93
♦ Validating XML Definitions, 99
♦ Setting XML View Options, 100
♦ Troubleshooting XML Editing, 102

77
Overview
When you import XML definitions with the Designer, you can create XML definitions using
entity or hierarchy relationships, custom XML views, or no views. After you create a
definition, you must use the XML Editor to make changes to it.
Use the XML Editor to create views, modify components, add columns, and maintain view
relationships in the workspace. When you update an XML definition, the Designer
propagates the changes to any mapping that includes the source. Some changes to XML
definitions can invalidate mappings.
Note: If you made significant changes to the source you used to create an XML definition, you
can synchronize it to the XML definition rather than editing the definition manually. For
information about synchronizing XML source definitions, see “Synchronizing XML
Definitions” on page 68.
Figure 4-1 shows the XML Editor:

Figure 4-1. XML Editor

Navigator

XML
Workspace

Columns Window

78 Chapter 4: Using the XML Editor

 The XML Editor uses icons to represent XML component types. To view a legend that
describes the icons in the XML Editor, click View-Legend.
The Legend dialog box displays.

The XML Editor contains the following windows:

♦ Navigator
♦ Workspace
♦ Columns

XML Navigator Window

The Navigator displays the schema in a hierarchical form and provides information about
selected components. You can expand a component to see components below it in the
hierarchy.
The Navigator toolbar provides shortcuts to most of the functions that you can complete in
the Navigator. It also provides navigation buttons that enable you to traverse components in
the schema.

Properties Tab
The Properties tab displays information about a component you select in the Navigator. If the
component is a complex element, you can view element properties in the schema, such as
namespace, type, and content. When you view a simple element or attribute, the Properties
tab shows the type and length. The Properties tab also displays any annotations for schema
components.
If you import the definition from an XML file, you can edit the datatypes and cardinality
from the Properties tab. If you create the definition from a DTD file, you can edit the
component type.
You can change the namespace prefix and location for an element if the schema uses
namespace. The prefix identifies the element or attribute declarations that belong to a
namespace.

Actions Tab
The Actions tab lists options that you can use to see more information about a selected
component. It also provides a utility that enables you to reverse changes you make to
components in the Navigator.

Overview 79
 The following options might display on the Actions tab, depending on the properties of the
component you select:
♦ ComplexType references. Displays the references to a selected complex type.
♦ ComplexType hierarchy. Displays the complex types derived from the selected
component.
♦ Element references. Displays the components that reference the selected element.
♦ Child components. Displays the global schema components that the selected component
uses.
♦ Revert simpleType. Changes the type, length, and precision values back to the original
value if you have changed them.
♦ XML view references. Displays all the XML views and columns that reference the selected
component.

Workspace Window
The XML Workspace window displays a graphical representation of the XML views and the
relationships between the views. You can create XML views in the workspace and define
relationships between views.
The XML workspace toolbar provides shortcuts to most of the functions that you can do in
the workspace.
You can modify the size of the XML workspace in the following ways:
♦ Hide the Columns window. Click View-Columns Properties.
♦ Hide the Navigator. Click View-Navigator.
♦ Reduce the workspace. Click the Zoom button on the Workspace toolbar.

Columns Window
The Columns window displays the columns for a view that you select in the workspace. You
can use the Columns window to name columns that you add. If you use pivoted columns, you
use the Columns window to select and rename occurrences of multiple-occurring elements.
You can also specify options such as the Not Null option to prevent null data in instance
documents, and the Force Row, Hierarchy or Type Relationship Row, and Non-Recursive
Row options. These options affect how the PowerCenter Server writes data to XML targets.
For more information about the Columns window, see “Setting XML View Options” on
page 100.

80 Chapter 4: Using the XML Editor

Creating and Editing Views
Use the XML Editor to create custom XML views, or to edit XML views that you created with
the XML Wizard. To create views, you define the view and specify the columns in the view. If
the schema has multiple-occurring elements, you can specify which element occurrences to
include in the view. You can also create special ports for XML target file names, and pass-
through ports for XML Parser and Generator transformations.
You can perform the following tasks to edit XML views:
♦ Create XML views.
♦ Add columns to views.
♦ Delete columns from views.
♦ Add reference ports.
♦ Add FileName columns.
♦ Add pass-through fields.

Creating an XML View

You can create views in the XML workspace. When you create an XML definition with no
views, the XML workspace is empty when you open the XML Editor. You can create an empty
view and add normal columns, pivoted columns, or a view row column to the view.
Use the following procedure to add views to the XML definition.

To create a new XML view in the workspace:

1. Choose XML Views-Create XML View.

The XML Editor creates a blank view in the workspace and displays empty columns in
the Columns window.
2. Enter a name for the view in the Columns window.
The name displays on the XML view in the workspace.
3. Choose Components-XPath Navigator.

Creating and Editing Views 81

 The XPath Navigator displays in the Navigator window.
Mode Button

4. Set the Column Mode in the XPath Navigator to View Row Mode in order to add the
view row.
5. Select the element in the Navigator and drag it to the view in the workspace. The XML
Editor highlights the view row in blue.
The first time you add a column to a view, the Designer verifies the column can be a view
row. This occurs even if you do not specify to add a view row.
6. To change the view row to another column, right-click the appropriate row in the view
and select Set As view row.

Adding Columns to Views

You can add columns to XML views inside the XML workspace. To add columns, select them
from the XPath Navigator.
You can add components to an XML view when the following conditions are true:
♦ The component path starts from the element in the schema that is the view root for that
view.
♦ The component is not an enclosure element.
♦ The component does not violate normalization or pivoting rules. For example, you cannot
add more than one multiple-occurring element to a view.
♦ You can add an element with mixed content as two different columns: one as a simple
type, and one as a complex type.
♦ Two views can share the same column, and a view might contain multiple identical
columns.

To add columns to a view:

1. Select the XML view in the workspace.

82 Chapter 4: Using the XML Editor

 2. Highlight a parent element in the Navigator that contains the components you want to
add.
3. Choose Components-XPath Navigator.
The XPath Navigator displays in the Navigator window.

Mode Button

4. Click the Mode button and select to add a column, or a view row. Select Advanced to add
a pivoted column if you have a multiple-occurring element that you want to pivot into
separate columns in the view. Advanced Mode also allows you to add normal columns.
Note: You cannot create pivoted columns in XML target definitions.

5. Drag a column from the XPath Navigator into the appropriate view in the XML
workspace. You can select multiple columns at a time.
The XML Editor validates the column you add. If the column is invalid for the view, a
message displays in the status bar while you are dragging the column. As you add new
columns to views, they display in the Columns window.

Adding Pivoted Columns

You can pivot any multiple-occurring element in an XML view. You can also pivot view rows.
If you add a pivoted column, a default occurrence number displays in the Columns window.
This number indicates which occurrence from the schema you want to use in the column. You
can change the occurrence number to correspond to any occurrence you need. You can
rename the column and add another occurrence.

To add pivoted columns to a view:

1. Select the XML view in the workspace.

2. Highlight the element you want to pivot in the Navigator. Or, you can highlight any
element in the parent chain.
3. Choose Components-XPath Navigator.

Creating and Editing Views 83

 4. Select Advanced Mode
5. Drag the column to pivot from the XPath Navigator to the view in the XML workspace.
The Designer adds a default occurrence in the Columns window. This number indicates
which occurrence from the schema you want to use in the column.
Figure 4-2 shows a pivoted Product Column and YTD Sales column. The view contains
two occurrences of YTD Sales for the third occurrence of product for Store:

Figure 4-2. Pivoted Columns

6. To change the occurrence number, click the XPath link.

7. Change the pivot to correspond to the occurrence you need and click OK.

Figure 4-3. Specify XPath Pivot Column

8. You can rename the column and add another occurrence.

Note: You cannot change a pivot value in a column XPath if the pivot value is part of a
view row.
For more information about pivoting, see “Pivoting Columns” on page 51.

Expanding a Complex Type

A schema might have a complex type that is a base type for more than one type. For example,
a Publication can be a Magazine or a Newspaper. When you create XML views, you can
choose how to use Publication, either as a Magazine or a Newspaper type.
When you view a complex type in the XPath Navigator, you can view its derived types.

84 Chapter 4: Using the XML Editor

 To expand a complex type:

1. Highlight the complex type in the XPath Navigator.

The Expand Complex Types list shows any derived types.
2. Select the type you want to use. If you add the component to an XML definition, the
definition contains the type you select.
For more information about complex types, see “Complex Types” on page 19.

Applying Content to the Any Content Element

PowerCenter imports XML schemas that use the ANY content element, however, in order to
use the element in a view, you must define content for it. When you import a schema. the
ANY content element displays in the Schema Navigator with no properties.
For example, the following schema element includes an ANY content element:
<xs:element name="person">
<xs:complexType>
<xs:sequence>
<xs:element name="firstname" type="xs:string"/>
<xs:element name="lastname" type="xs:string"/>
<xs:any minOccurs="0"/>
</xs:sequence>
</xs:complexType>
</xs:element>

Figure 4-4 shows the ANY content element in the Schema Navigator:

Figure 4-4. Any Content Element in the Schema Navigator

Any Content Element

To apply content to the ANY element:

1. Click the any link in the Schema Navigator.

The Edit Any or AnyAttribute type content dialog box displays.

Creating and Editing Views 85

 Figure 4-5 shows the Edit Any or Any Attribute dialog box:

Figure 4-5. Edit Any or Any Attribute Dialog Box

Add Type

2. Click Add Type.

A new row displays.
3. Choose a type from the list of valid elements. Select a cardinality and click OK.
The substituted type displays in the Schema Navigator.

Figure 4-6. New Type Added to the Any Content Element

ANY Content
substituted type

Deleting Columns from Views

You can delete columns from views.

To delete a column from a view:

1. Right-click the column in the view in the workspace.

2. Select Delete This Column.
The XML editor prompts you to confirm that you want to delete the column.
3. Click Yes to confirm.
The XML Editor removes the column from the view. The column still displays in the
Navigator.

86 Chapter 4: Using the XML Editor

 Deleting a Pivoted Column
To delete an occurrence of a pivoted column, select and delete the column from the Columns
window.
To delete a pivoted column:
1. Right-click the column you want to delete in the Columns window.
2. Select Delete-Pivot.
3. Click Yes to confirm the delete.

Adding a Pass-Through Port

If you use an XML Generator transformation in a pipeline, you can add pass-through ports to
the transformation. Pass-through ports let you pass non-XML data through the XML
Generator transformation. When you define the port in the transformation, you add it to
either the DataInput group in the XML Parser transformation or the DataOutput group in
the XML Generator transformation.
Once you generate this port, you can add a reference port to pass the data through a view in
the XML definition. For more information about adding pass-through and reference ports,
see “Generating Pass-Through Ports” on page 147.

Finding Corresponding Reference Ports

If you have several pass-through ports in an XML definition, you can determine the
corresponding reference ports.

To determine the referenced port for a pass-through port:

1. Right-click the pass-through port.

2. Select Navigate to-Referenced Column.
The XML Editor highlights the referenced column in the workspace.

Adding a FileName Column

In a session, the PowerCenter Server outputs a new target XML file each time a new root
value occurs in the data. You can add a FileName column to an XML view to generate unique
file names for the XML files. The file name overrides the default output file name in the
session properties.

To create a FileName column in an XML view:

1. Right-click the view in the XML Editor.

2. Select Create FileName Column.
The XML Editor creates a new string column in the view.

Creating and Editing Views 87

 3. Change the column name in Columns window.
4. Exit the XML Editor.
The column displays in your XML definition. The XPath is $Filename.
5. You must set up an Expression transformation or other transformation in your mapping
to generate a file name to pass to the XML target.
Note: When you use XML filename ports, you need to specify the directory name for the file.

For more information about the FileName column, see “Naming XML Files Dynamically” on
page 116.

88 Chapter 4: Using the XML Editor

Maintaining View Relationships
You can use the XML Editor to maintain relationships between views in an XML definition.
You can maintain relationships by the following methods:
♦ Create a relationship between views. Define relationships between views in the
workspace.
♦ Create a type relationship. Define a type relationship between a column in a view and a
type view in the workspace.
♦ Recreate entity relationships. Generate new views and relationships using the same
options as in the XML Wizard.

Creating Relationships Between Views

You can use the XML Editor to create hierarchical or inheritance relationships between views.
You cannot create a relationship between an XML Source and a non-XML Source.

To create a relationship between XML views:

1. Right-click the top of the child XML view in the workspace.

2. Select Create Relationship.
3. Move the pointer to the parent view to establish a relationship.
As you move the pointer, a link displays between the views, and the XML Editor verifies
that the relationship is valid. If a relationship is not valid, an error message displays in the
status bar.
4. If no error message displays, click the parent view to establish the relationship.
The XML Editor creates the relationship and adds the appropriate foreign keys:

The XML Editor adds the foreign

key to Employee.

Maintaining View Relationships 89

 5. To view details about the relationship, place the cursor over the link between the views.
The Editor displays the relationship type and the primary and foreign keys.

Creating Type Relationships

You can create a type relationship between a column in a view and a type view. If the column
is pivoted, you can choose which occurrences to include in the relationship.

To create a type relationship:

1. Right-click the column in the view you want to use.

2. Select Create Relationship.
3. If the column is pivoted, select which occurrence to use:

4. Move the pointer to the type view to establish a relationship.

As you move the pointer, a link displays between the views and the XML Editor verifies
that the relationship is valid. If a relationship is not valid, an error message displays in the
status bar.
5. If no error message displays, click the parent view to establish the relationship.
The XML Editor creates the relationship:

Recreating Entity Relationships

You can recreate entity relationships for an XML definition using the Recreate Entity
Relationships dialog box. This function enables you to generate new XML views using the
same options as in the XML Wizard. When you regenerate views, you can choose to keep the
views you have already created. The XML definition contains all the views.

90 Chapter 4: Using the XML Editor

To recreate entity relationships for an XML definition:

1. Open the XML definition in the XML Editor.

2. Highlight the XML root in the Navigator.
If you highlight another component, the XML Editor uses that component as the root.
3. Choose XML Views-Create Entity Relationship.

4. Choose from the following options to reduce metadata explosion:

♦ Reduce element explosion. For any multiple-occurring or reference element, the XML
Wizard creates one XML view with multiple hierarchical relationships.
♦ Reduce complex type explosion. For any multiple-occurring or referenced complex
type, the XML Wizard creates one XML view with multiple type relationships. If the
schema uses inherited complex types, you can also reduce explosion of inherited
complex types.
♦ Reduce complex type inheritance explosion. For any inherited type, the XML Wizard
creates one XML view using multiple type relationships.
♦ Share existing XML views. Do not remove existing XML views.
♦ Refresh shared XML views. Save existing views but update them.
5. Click Next.

Maintaining View Relationships 91

 The Recreate Entity Relationships dialog box displays.

Generate a
view.

Display child
components.

Exclude a
child
component.

6. To display a child component, select a shared element or complex type and click the
name.
7. To exclude a child component, clear the element in the Exclude Child Components pane.
To generate a new view, select the element or complex type. When you create the new
entity relationships, you generate a view with that element as a view root.

92 Chapter 4: Using the XML Editor

Viewing Schema Components
You can perform the following tasks to view components in the Navigator and workspace:
♦ Update the namespace. Change the location of a schema.
♦ Sort components. Change the order of the schema components in the Navigator.
♦ Navigate to components. Find components by navigating from a component to another
component or area of the XML Editor window.
♦ Arrange views in the workspace. Arrange the views in the workspace hierarchically.
♦ Search for components in the schema. Find components in the Navigator.
♦ Search for components in the workspace. Find components in the workspace.
♦ View XML metadata. View an XML file, schema, or DTD that the XML Editor creates
from the XML definition.
♦ Preview XML data. Display an XML view using sample data from an external XML file.
♦ Validate XML definition. Validate views and relationships between the views.

Updating the Namespace

If the schema uses a namespace, you can change the prefix and schema location for element.

To update the namespace schema location for an element:

1. Select the element in the Navigator.

2. Click the Properties tab.
3. Click the Namespace value.

4. Update the prefix and schema location.

The schema location for a global namespace is a schema file. You can have multiple
schema files for a single prefix only with global namespaces.

Viewing Schema Components 93

 Sorting Components in the Components Pane
You can sort components in the Navigator by type, hierarchy, or namespace.

To sort components:

1. Click Components-Sort By.

2. Select Component Type, Hierarchy, or Namespace.
The components in the Navigator display in the order you specify.

Navigating to Components
If you have large XML definitions you can quickly find components by using the Navigate
option. To use this option, select a component to navigate from and select a navigation
option. For example, if you click a foreign key, you can navigate to the associated primary key
or to the column in the Columns window. You can navigate between components in the
workspace, the Columns window, and the Navigator.

To navigate to components:

1. Right-click a component in the workspace or in the Columns window.

2. Select Navigate to.
3. Select an available option.
You can select from the following options, depending on the component you select to
navigate from:
♦ Schema component. Highlights the component in the Navigator.
♦ PowerCenter column. Highlights the column in the Columns window.
♦ Primary key. Highlights the primary key associated with a selected foreign key.
♦ Referenced column. Highlights the referenced column associated with a pass-through
port in an XML Parser or Generator transformation.
♦ XPath Navigator. Displays the path to the selected the component.
♦ XML view. Highlights a view in the workspace that contains the selected column from
the Columns window.

Arranging Views in the Workspace

When you work with views in the workspace you can move them around in order to add links
or view relationships. You can reorganize the views into a hierarchical arrangement.
To arrange views in the workspace, Choose Layout-Arrange, or you can right-click the
workspace and select Arrange.

94 Chapter 4: Using the XML Editor

Searching for Components in the Schema
You can search for components in the schema and namespaces. You can search for
components by name or by type. You can also specify properties to search for, such as an
annotation, a fixed or default value, or a length. You can search for legal values using the
enumeration property.

To search for components in the schema:

1. Choose Edit-Search In Schema.

The Search Components dialog box displays.
2. To search by component properties, click the Advanced Options link to view the
Properties you can search for.
3. Enter a name, a type, or a property to search. If you want to search in a specific
namespace, click All, and select a namespace from the list.
4. Click Search.
The search results display in the bottom of the dialog box.

Click Advanced
Options to
search by
component
properties.

Click a search
result to view it
in the Properties
window.

5. Click a search result to view the component in the Properties window.

Searching for Components in XML Views

You can search for components and columns in an XML definition. If you search by
component name, you can find all the associated columns in the definition. For example, if
you search for a component “Number,” the results contain the views and columns that use the
component “Number.”
To search using a partial key, enter the first few characters of the column or component name.

Viewing Schema Components 95

 To search for components in XML views:

1. Choose Edit-Search XML Views.

The Search XML Views and Columns dialog box displays.
2. Enter search criteria. You can search for all column types, regular column types,
generated keys, or other types. Other types include FileName columns, reference Ports,
and pass-through fields.
3. Click Search.
The search results display in the bottom of the dialog box.
Figure 4-7 shows the Search XML Views dialog box:

Figure 4-7. Search XML Views and Columns

4. To clear search fields, click New Search.

Viewing a Simple Type Hierarchy

You can view a hierarchy of the XML simple types in the schema definition. To view a
hierarchy of simple types, choose View-SimpleType Hierarchy. A window displays a hierarchy
of the simple types.

Viewing Complex Types in the Schema

You can display a hierarchy of the complex types in the schema definition. To view a hierarchy
of complex types, choose View-ComplexType Hierarchy. A window displays a hierarchy of the
complex types in the schema. Select a component from the ComplexType Hierarchy window
to navigate to it in the schema.

96 Chapter 4: Using the XML Editor

Viewing XML Metadata
You can view an XML definition as an XML, DTD, or XML schema file.

To view XML metadata:

1. If you want to view the metadata as a sample XML document, choose a global
component in the Navigator.
2. Click View-XML Metadata.
The View XML Metadata dialog box displays:

3. Choose how you want to view the XML definition.

If you use multiple namespaces, choose the namespace to use.
A default application or text editor displays the metadata.
4. To save a copy of the XML, DTD, or schema, choose Save As.
5. Enter a new file name.
If the default display application is a text editor, you need to include the file suffix with
the file name and change the file type from.txt to All Files.

Previewing XML Data

You can preview XML data using an XML definition and an external XML file. To preview
data, you must have a valid XML definition in the repository and some sample data in an
external XML file that is valid for the definition. You can view data for one XML view at a
time.
For example, you might import a schema called EmployeeData.xsd and you create an XML
definition with several custom views. You have an XML file called Employees.xml that
contains sample data from 20 employees. You can preview each custom view using the sample
data from the XML file.

Viewing Schema Components 97

 Use the following procedure to preview data for a view in the XML Editor workspace.

To preview the XML data:

1. Choose XML Views-Preview Data.

2. Browse for the XML file containing data to preview.
If you have not selected a view from the Workspace, or if you have more than one view
selected, the Preview XML Data dialog box prompts you to select one view.
The Preview XML Data dialog box displays data from the XML file using the view that
you specify. If you want to use a different XML file, click the Select File icon on the
dialog box.
Figure 4-8 shows the Preview XML Data window:

Figure 4-8. Preview XML Data

98 Chapter 4: Using the XML Editor

Validating XML Definitions
You can validate an XML definition in the XML Editor.

To validate an XML definition:

1. Open the definition in the XML Editor.

2. Click inside the workspace.
3. Choose XML Views-Validate XML Definition.
A Validate window displays the results:

Validating XML Definitions 99

Setting XML View Options
The Columns window displays the columns for a selected XML view. You can select the Not
Null option to prevent null data from occurring in the instance document. For source
definitions, you can pivot a multiple-occurring leaf element. For more information about
pivoting, see “Pivoting Columns” on page 51.
Figure 4-9 show the Columns window:

Figure 4-9. Columns Window

If you are working with an XML source definition, the Columns window contains XML View
Options. These options give you flexibility about when you generate rows or foreign keys in a
session. You can select the following options:
♦ All hierarchy foreign keys
♦ Non-recursive row
♦ Hierarchy relationship row
♦ Force row
♦ Type relationship row

All Hierarchy Foreign Keys

Select the All Hierarchy Foreign Keys option to generate key values for all foreign keys in a
view. By default, the PowerCenter Server generates foreign key values for just the closest
parent to the view. Select the All Hierarchy Foreign Keys option to generate values for all the
foreign keys in the view.

Non-Recursive Row Option

Select Non-Recursive Row for an XML view in a circular relationship. Use it to stop recursion
after one occurrence of the element during a session.

100 Chapter 4: Using the XML Editor

Hierarchy Relationship Row Option
Select Hierarchy Relationship Row to generate a row only when parent view data occurs with
the child view data in a session. For example, a definition might have a hierarchy that includes
an employee view and an address view. Employee is the parent view. During a session, the
PowerCenter Server generates rows for the address view only when it has data for the
employee view.

Force Row Option

Select Force Row to generate a row for the XML view even if the view row element does not
appear. For example, a view row contains product\sales, you can choose to output product,
even if the view row has no sales data in it. You can use this option only for denormalized
source XML views.

Type Relationship Row Option

Select Type Relationship Row to generate a row only when parent type data occurs with the
derived view data in a session. For example, a definition might have a hierarchy that includes
an Address type and a Canadian address type. During a session, the PowerCenter Server
generates rows for the Canadian address view only when it has data for the Address view.

Setting XML View Options 101

Troubleshooting XML Editing
When I validate my XML definition I get an error that says my XML definition is too large.
Why does this occur?
If you import an XML file that has components defined with infinite lengths, you can easily
exceed the 500 MB limit for total column length. You can change the column lengths in the
XML Editor, or you can set an option to override all infinite lengths and reimport the file. For
more information about overriding infinite lengths, see “Importing an XML Source
Definition” on page 57.

I cannot find the DTD or XML schema file that I created when I viewed XML metadata.
The DTD or XML schema file that you can view is a temporary file that the Designer creates
only for viewing. If you want to use the file for other purposes, save it with another name and
directory when you view it.

When I add columns to my XML source views, the hierarchy in my source XML file
remains the same.
When you add columns to XML source views, you do not add elements to the underlying
hierarchy. The XML hierarchy that you import remains the same no matter how you create
the views or how you map the columns in a view to the elements in the hierarchy. You can
modify the datatypes and the cardinality of the elements, but you cannot modify the structure
of the hierarchy.

102 Chapter 4: Using the XML Editor

 Chapter 5

Working with XML Targets

This chapter includes the following topics:

♦ Overview, 104
♦ Importing an XML Target Definition from XML Files, 105
♦ Creating a Target from an XML Source Definition, 106
♦ Editing XML Target Definition Properties, 107
♦ Validating XML Targets, 112
♦ Using an XML Target in a Mapping, 114
♦ Troubleshooting, 118

103
Overview
You can create XML target definitions in the following ways:
♦ Import the definition from an XML file. Create a target definition from an XML, DTD,
or XML schema file. You can import XML file definitions from a URL or a local machine.
If you import an XML file with an associated DTD, the XML Wizard uses the DTD to
generate the XML document. For more information about importing XML files, see
“Importing an XML Target Definition from XML Files” on page 105.
♦ Create an XML target definition based on an XML source definition. Drag an existing
XML source definition into the Warehouse Designer. If you create an XML target
definition, the Designer creates a target definition based on the hierarchy of the XML
definition. For more information about creating an XML target from other definitions, see
“Creating a Target from an XML Source Definition” on page 106.
♦ Create an XML target based on a relational file definition. You can import an XML
target definition from a relational or flat file repository definition. For more information
about importing XML definitions from relational or flat file definitions, see “Creating
XML Definitions from Repository Definitions” on page 72.
In addition to creating XML target definitions, you can perform the following tasks with
XML targets in the Warehouse Designer:
♦ Edit target properties. Edit an XML target definition to add comments or update them to
reflect changed target XML, DTD, or XML schema files. For information on editing XML
target definitions, see “Editing XML Target Definition Properties” on page 107.
♦ Synchronize target definitions. You can synchronize your target XML definition to an
updated schema if you need to make changes. Synchronizing enables you to update the
XML definition instead of reimporting the definition when the schema changes. For more
information about synchronizing XML definitions, see “Synchronizing XML Definitions”
on page 68.

104 Chapter 5: Working with XML Targets

Importing an XML Target Definition from XML Files
You can import XML definitions from DTD, XML, or XML schema files. You can import
local files or reference files using a URL.
You can choose from the following options to create XML views:
♦ Create entity relationships. This option enables you to create entities for some
components so they do not expand each time the data references them. You create
relationships between entities instead of creating one large hierarchy.
♦ Create hierarchical relationships. This option creates a root and XML components
expand under this root. You can choose to create normalized or denormalized views. If you
choose normalized, every element or attribute appears once. One-to-many relationships
become separate XML views with keys to relate the views. If you create denormalized XML
views, all elements and attributes display in one hierarchical group. For more information
hierarchical relationships, see “Generating Hierarchical View Relationships” on page 38.
♦ Create custom XML views. This option enables you to select multiple global elements as
roots for XML views and select options for reducing metadata explosion.
♦ Skip creating XML views. This option creates no views. You can use the XML Editor to
create XML views in the workspace at a later time.
♦ Synchronize XML definitions. You can use this option to avoid reimporting the XML
definition when the associated XML schema file or DTD file changes.
Note: Informatica recommends that you import DTD or XML schema files instead of XML
files. PowerCenter has a size limit of 100K for XML files. If you import an XML file with an
associated DTD, the XML Wizard uses the DTD.

To import XML target definitions:

1. In the Warehouse Designer, select Targets-Import XML Definition. The Import XML
Definitions window opens. The default displays schema files in a local folder.
2. Click Local File or URL to browse for XML files.
3. To browse for DTD or XML files, select the appropriate file extension from the Files of
Type list.
For more information about using the XML Wizard see “Importing an XML Source
Definition” on page 57.

Importing an XML Target Definition from XML Files 105

Creating a Target from an XML Source Definition
When you want to create a target definition that closely resembles an existing source
definition, you use the source definition or a shortcut to the source definition to create the
target definition. You drag the XML source definition into the Warehouse Designer to create
an XML target definition or a relational target definition.
Use the following guidelines to create XML targets:
♦ When you create an XML target definition from an XML source definition, you create a
duplicate of the XML source definition.
♦ A valid XML source definition does not necessarily create a valid XML target definition.
To ensure that you create a valid target definition, validate the target.
Note: XML target definitions cannot contain pivoted columns.

Use the following guideline to create relational targets:

♦ If you create a relational target, the Designer creates the relational target definitions based
on the groups in the XML source definition. Each group in the XML source definition
becomes a target definition.
♦ The Designer creates the same relationship between the target definitions as the
relationship among the groups in the source definition.

To create a target definition from an XML source definition:

1. Drag an XML source definition from the Navigator into the Warehouse Designer
workspace.
The XML Export dialog box appears:

2. Select the type of target you want and click OK.

The target definition appears in the Warehouse Designer workspace. If you select
relational targets, more than one target definition might appear in the workspace,
depending on the source.
3. Choose Repository-Save to save the target definition in the repository.

106 Chapter 5: Working with XML Targets

Editing XML Target Definition Properties
After you create an XML target definition, you can edit the properties to reflect changes in
the target data, to add business names and comments, or to change the code page.

To edit XML target properties:

1. Open the XML target in the Warehouse Designer.

2. Right-click and select Edit.

3. On the Table tab, edit the settings as necessary:

Table Settings Description

Rename button Edit the name of the target definition and enter a business name.

Business Name Descriptive name for the target table. Edit the Business Name using the Rename
button.

Constraints Not applicable to XML targets. Any entry is ignored.

Creation Options Not applicable to XML targets. Any entry is ignored.

Description Optional description of target table. Character limit is 2,000 bytes/K, where K is
the maximum number of bytes for each character in the repository code page.
Enter links to business documentation.

Code Page Select the code page to use in the target definition. For more information about
XML code pages, see “Code Pages” on page 27.

Editing XML Target Definition Properties 107

 Table Settings Description

Database Type Indicates that the target definition is an XML target.

Keywords Allows you to keep track of your targets. As development and maintenance work
continues, the number of targets increases. While all of these targets may
appear in the same folder, they may all serve different purposes. Keywords can
help you find related targets. Keywords can include developer names, mappings,
or the associated schema.
You can use keywords to perform searches in the Repository Manager. For
details on keyword searches in the Repository Manager, see “Using the
Repository Manager” in the Repository Guide.

4. On the Columns tab, you can view XML column definitions.

The Columns tab displays the following properties:

Columns Settings Description

Select Table Displays the target definition you are editing. To choose a different definition to
edit, select one from the list of definitions you have available in the workspace.

Column Name Name of the column.

Datatype XML datatype for the column.

Precision Size of column. You can change precision only for some datatypes, such as
string.

Scale Number of decimals in a decimal datatype.

108 Chapter 5: Working with XML Targets

 Columns Settings Description

Not Null Indicates if the column can be blank.

Key Type The type of key the XML Wizard generates to link the views.

XPath The path through the XML document hierarchy that enables you to locate an
item.

5. On the Properties tab, you can modify the transformation attributes of the target
definition. If you are using a source-based commit session or Transaction Control
transformation with the XML target, you can define how you want to flush data to the
target. For more information, see “Working with XML Targets in a Session” on page 155.

Editing XML Target Definition Properties 109

 You can edit the following attributes:

Columns Settings Description

Select Table Displays the source definition you are editing. To choose a different source
definition to edit, select it from the list.

Duplicate Group Choose one of these options to handle processing duplicate rows in the target:
Row Handling - First Row. The PowerCenter Server passes the first duplicate row to the
target. Rows following with the same primary key are rejected.
- Last Row. The PowerCenter Server passes the last duplicate row to the
target.
- Error. The PowerCenter Server passes the first row to the target. Rows with
duplicate primary keys increment the error count. The session fails when the
error count reaches the error threshold.
For more information about duplicate group row handling, see “Handling
Duplicate Group Rows” on page 159.

DTD Reference DTD or XML schema file name for the target XML file. The PowerCenter
Server adds the document type declaration to the XML file when you create it.
For more information about using the DTD or schema file name, see “DTD and
Schema Reference” on page 160

On Commit The PowerCenter Server can generate multiple XML documents or append to
one XML document after a commit. You can use one of the following options:
- Ignore Commit. The PowerCenter server creates an XML document and
writes to it at end of file only.
- Create New Document. Creates a new XML document at each commit.
- Append to Document. Writes to the same XML document after each commit.
For more information about flushing XML on commits, see “Flushing XML on
Commits” on page 160.

Cache Directory The directory for the XML target cache files. The default is the $PMCacheDir
server variable. For more information about working with caches, see “XML
Caching Properties” on page 162.

Cache Size The total size in bytes for the XML target cache. The default is 10,000,000
bytes.

6. On the Metadata Extensions tab, you can create, modify, delete, and promote non-
reusable metadata extensions, as well as update their values. You can also update the

110 Chapter 5: Working with XML Targets

 values of reusable metadata extensions. For more information about metadata extensions,
see the Repository Guide.

7. Click OK.
8. Choose Repository-Save.

Editing XML Target Definition Properties 111

Validating XML Targets
You can create customized XML views that describe how to extract data to an XML
document. However, not all view structures or relationships between views are valid in an
XML definition. Some view structures might be valid for an XML source, but not for an
XML target. The Designer prevents you from creating ambiguous definitions.
PowerCenter validates target XML views when you perform different tasks:
♦ The Designer does limited validation when you save or fetch an XML target from the
repository.
♦ The XML Editor validates each step when you edit XML in the XML workspace.
♦ You can validate a target definition in the XML Editor.
♦ PowerCenter validates XML target connections when it validates mappings.
The Designer uses rules to validate the following types of relationships:
♦ Hierarchy relationships
♦ Type relationships
♦ Inheritance relationships

Hierarchy Relationship Validation

The Designer uses the following rules to validate hierarchy relationships:
♦ A view that has a root at a type cannot be a standalone view. It must be a child in an
inheritance relationship or it must have a type relationship with another view. An XML
target is invalid if it has no views that are rooted at an element.
♦ You must connect a view with a multiple-occurring view row to another view.
♦ Two views cannot have the same effective view row.
♦ An XML target is invalid if it has no view root at an element.
♦ You can separate parent and child views by other elements, but if you have a choice of two
parents for a view, you must use the closest one. Determine the closest parent by the path
of the effective view row. One parent comes before the other in the path. Choose the view
that comes second in the path.
♦ You must connect all views with the same view root in the same hierarchy. The definition
cannot contain multiple trees for the same view root.
♦ An XML view can have a hierarchical relationship to itself only if the view row and the
view root are identical for the view.

112 Chapter 5: Working with XML Targets

Type Relationship Validation
A type relationship is a relationship between a column and a view. It is not a relationship
between two views. The following rules apply to type relationships:
♦ A column in view, V1, can have a type relationship to a view, V2, if the view roots are the
same type, or the V2 view root type is derived from the V1 view root. Both view roots
must be global complex types.
♦ If a column in a view has a type relationship to another view, you cannot expand the
column.

Inheritance Validation
You can create two types of inheritance relationships with XML views:
♦ View-to-view inheritance. A view is a derived type of another view. Both views must have
global complex view roots.
A view can have an inheritance relationship to another view only if its view root is a
complex type derived from the view root type of the other view.
A view can be a parent in multiple inheritance relationships, but a view can be a child in
only one inheritance relationship.
♦ Column-to-view inheritance. The column is an element of a local complex type, Type1,
and the view is rooted at a global complex type, Type2. Type1 is derived from Type2.
A column in a view can have an inheritance relationship to another view if the column is a
local complex type and the type is derived from the view root type of the other view.
If a column in a view, V1, has an inheritance relationship to a view, V2, you cannot put
the content of V2 into view V1.

Validating XML Targets 113

Using an XML Target in a Mapping
When you add an XML target to a mapping, you need to following mapping guidelines for
multigroup transformations. For more information about multigroup transformations, see the
Designer Guide.
The following components affect how you map an XML target in a mapping:
♦ Active sources
♦ Root elements
♦ Target port connections
♦ Abstract elements
♦ Transaction control points
♦ FileName columns

Active Sources
An active source is an active transformation the PowerCenter Server uses to generate rows.
The PowerCenter Server can load data from different active sources to an XML target.
However, all target ports within a single group must receive data from the same active source.
The following transformations are active sources:
♦ Aggregator
♦ Application Source Qualifier
♦ Custom, configured as an active transformation
♦ Joiner
♦ MQ Source Qualifier
♦ Normalizer (VSAM or pipeline)
♦ Rank
♦ Sorter
♦ Source Qualifier
♦ XML Source Qualifier
♦ Mapplet, if it contains one of the above transformations

Selecting a Root Element

If an XML definition has more than one possible root, you can specify a root element for a
target instance.

To specify the root element:

1. Right-click the target definition in the Mapping Designer and select Edit.

114 Chapter 5: Working with XML Targets

 2. Click the Properties tab.
3. Click the arrow in the Root Element value column.
The Select Root dialog box displays:

4. Select an element from the list.

Connecting Target Ports

When a session uses a mapping that contains an XML target, the PowerCenter Server must be
able to create the XML hierarchy in the target and determine where a row of data fits within
the hierarchical structure.
Use the following guidelines when you connect ports in a mapping:
♦ If you connect one port in a group, you must connect both the foreign key and primary
key ports for the group.
♦ If you connect a foreign key port in a group, you must connect the associated primary key
port in the other group. If you do not connect the primary key port of the root group, you
do not need to connect the associated foreign key ports in the other groups.
♦ If you use an XML schema with a default attribute value, you must connect the attribute
port to create the default attribute in the target. If you pass a NULL value through the
connected port, the PowerCenter Server writes the default value to the target.
You need to ensure that the PowerCenter Server can create a valid XML hierarchy during a
session by correctly connecting the XML target ports in a mapping. When you save or
validate a mapping with an XML target, the Designer validates the target port connections.

Connecting Abstract Elements

An abstract element cannot occur directly in an XML instance document. Instead, you must
use an element derived from the abstract element. By default, the Designer creates a view for
any abstract complex element. To reduce metadata, elements from the abstract type do not

Using an XML Target in a Mapping 115

 repeat in any derived type. When you map data to the abstract type, you need to also map
data to at least one derived type.
During a session, if the PowerCenter Server loads data to an abstract type, then it should also
have data for a non-abstract derived type associated with it. If the derived type has no data,
then the PowerCenter Server does not write the abstract element in the target XML
document.
For more information about abstract elements, see “Abstract Elements” on page 21.

Flushing XML Data to Targets

You can flush data to an XML target at each commit point in a session, however each input
group must receive data from the same transaction control point in the mapping. When you
create a session based on this mapping, you can append data to the XML file target at each
commit or create a new file at each commit. You can specify either option with the On
Commit target property.
When you connect the XML target input groups to multiple transaction control points, the
PowerCenter Server writes the data to the XML file target after it processes all source rows.
For more information about transaction control points, see “Understanding Commit Points”
in the Workflow Administration Guide.

Naming XML Files Dynamically

You can add a FileName column to an XML target definition to dynamically create file names
for XML documents. When the PowerCenter Server passes data to the FileName column, it
overrides the output file name in the target properties. For example, if you pass the string
“Harry” to the FileName column, the PowerCenter Server names the XML document Harry.
Note: If you are creating a new XML document on each commit, you need to dynamically
name each XML document you create. If you do not do this, the PowerCenter Server
overwrites the XML document from the previous commit.
The PowerCenter Server generates a new XML document for each distinct primary key value
in the root group of the target. You add a FileName column to set different names for each
document. Each name overrides the output file name in the session properties.
For information about creating a FileName column in the XML Editor, see “Adding a
FileName Column” on page 87.
For information about multiple XML document output see, “Multiple XML Document
Output” on page 163.

Example
The following example shows a mapping containing an XML target with a FileName column.
The Expression transformation generates a file name from the Country XML element and
passes the value to the FileName column. The mapping passes a country to the target root,

116 Chapter 5: Working with XML Targets

which is called Client. Whenever the Client value changes, the PowerCenter Server creates a
new XML file.
Figure 5-1 shows the mapping containing an XML target with a FileName column:

Figure 5-1. Filename Column in a Mapping

Define the Filename

filename. Column

The PowerCenter Server passes the following rows to the target:

Country,Region,Revenue
USA,region1,1000
France,region1,10
Canada,region1,100
USA,region2,200
USA,region3,300
USA,region4,400
France,region2,20
France,region3,30
France,region4,40

The session produces the following files by country name:

Canada.xml
France.xml
USA.xml
revenue_file.xml.lst

Using an XML Target in a Mapping 117

Troubleshooting
I imported a source definition from an XML file. Then I imported a target definition from
the same XML file. The default groups for the source and target definitions are not the
same.
The XML Wizard does not always create the same group structure for a source definition and
a target definition if you change some of the options when you import the target.
For example, the ContactInfo element in the following DTD is an enclosure element. It has
no text content, but has maxOccurs > 1. The child elements also have maxOccurs > 1.
<!ELEMENT HR (EMPLOYEE+)>
<!ELEMENT EMPLOYEE (LASTNAME,FIRSTNAME,ADDRESS+,CONTACTINFO+)>
<!ATTLIST EMPLOYEE EMPID CDATA #REQUIRED>
<!ELEMENT LASTNAME (#PCDATA)>
<!ELEMENT FIRSTNAME (#PCDATA)>
<!ELEMENT ADDRESS (STREETADDRESS,CITY,STATE,ZIP)>
<!ELEMENT STREETADDRESS (#PCDATA)>
<!ELEMENT CITY (#PCDATA)>
<!ELEMENT STATE (#PCDATA)>
<!ELEMENT ZIP (#PCDATA)>
<!ELEMENT CONTACTINFO (PHONE+,EMERGCONTACT+)>
<!ELEMENT PHONE (#PCDATA)>
<!ELEMENT EMERGCONTACT (#PCDATA)>

If you do not create an XML view for an enclosure element in the source definition, you do
not create the Contactinfo element in the source.
The XML Wizard creates the following source and target definitions:

Source Definition for HR.dtd Target Definition for HR.dtd

118 Chapter 5: Working with XML Targets

The wizard does not include the ContactInfo element in the source definition because of the
option setting. However, the wizard includes the ContactInfo element in the target definition.

The XML target created from my relational sources contains all elements, but no attributes.
How can I modify the target hierarchy so that I can mark certain data as attributes?
You cannot modify the component types that the wizard creates from relational tables.
However, you can view a DTD or AN XML schema file of the target XML hierarchy. Save the
DTD or XML schema file to your own directory and filename. Open this new file and modify
the hierarchy, setting the attributes and elements as needed. Then, you can use this file to
import a target definition with a new hierarchy. For more information about viewing XML
definitions, see “Viewing XML Metadata” on page 97.

Troubleshooting 119
120 Chapter 5: Working with XML Targets
 Chapter 6

XML Source Qualifier

Transformation
This chapter includes the following topics:
♦ Overview, 122
♦ Adding an XML Source Qualifier to a Mapping, 123
♦ Editing an XML Source Qualifier Transformation, 125
♦ Using the XML Source Qualifier in a Mapping, 129
♦ Troubleshooting, 135

121
Overview
Transformation type:
Active
Connected

When you add an XML source definition to a mapping, you need to connect it to an XML
Source Qualifier transformation. The XML Source Qualifier transformation defines the data
elements that the PowerCenter Server reads when it runs a session. It determines how the
PowerCenter reads the source data.
You can manually add a source qualifier transformation or you can create a source qualifier
transformation by default when you add a source definition to a mapping. For more
information about adding XML Source Qualifier transformations, see “Adding an XML
Source Qualifier to a Mapping” on page 123.
You can edit some of the properties and add metadata extensions to an XML Source Qualifier
transformation. For more information about editing an XML Source Qualifier, see “Editing
an XML Source Qualifier Transformation” on page 125.
When you connect an XML Source Qualifier transformation in a mapping, you must follow
some rules to create a valid mapping. For more information about using an XML Source
Qualifier transformation in a mapping, see “Using the XML Source Qualifier in a Mapping”
on page 129.

122 Chapter 6: XML Source Qualifier Transformation

Adding an XML Source Qualifier to a Mapping
An XML Source Qualifier transformation always has one input/output port for every column
in the XML source. When you create an XML Source Qualifier transformation for a source
definition, the Designer links each port in the XML source definition to a port in the XML
Source Qualifier transformation. You cannot remove or edit any of the links. If you remove an
XML source definition from a mapping, the Designer also removes the corresponding XML
Source Qualifier transformation. You can link one XML source definition to one XML Source
Qualifier transformation.
You can link ports of one XML Source Qualifier group to ports of different transformations
to form separate data flows. However, you cannot link ports from more than one group in an
XML Source Qualifier transformation to ports in the same target transformation.
If you drag columns of more than one group to a transformation, the Designer copies the
columns of all the groups to the transformation. However, it links only the ports of the first
group to the corresponding ports of the columns created in the transformation.
You can add an XML Source Qualifier transformation to a mapping by dragging an XML
source definition into the Mapping Designer workspace or by manually creating one.

Automatically Creating an XML Source Qualifier Transformation

When you drag an XML source definition into the Mapping Designer workspace, the
Designer creates an XML Source Qualifier transformation by default.

To create an XML Source Qualifier transformation by default:

1. In the Mapping Designer, create a new mapping or open an existing one.

2. Click and drag an XML source definition into the mapping.
The Designer creates an XML Source Qualifier transformation and links each port in the
XML source definition to a port in the XML Source Qualifier transformation.

Manually Creating an XML Source Qualifier Transformation

You can create an XML Source Qualifier transformation in a mapping if you have a mapping
that contains XML source definitions without Source Qualifiers or if you delete the XML
Source Qualifier transformation from a mapping.

To manually create an XML Source Qualifier transformation:

1. In the Mapping Designer, create a new mapping or open an existing one.

Make sure that there is at least one XML source definition without a source qualifier in
the mapping.
2. From the menu, choose Transformation-Create.

Adding an XML Source Qualifier to a Mapping 123

 The Create Transformation dialog box appears:

3. Select XML Source Qualifier transformation, and type a name for the new
transformation.
The naming convention for XML Source Qualifier transformations is
XSQ_TransformationName.
4. Click Create.
The Designer lists all the XML source definitions in the mapping with no corresponding
XML Source Qualifier transformations.

5. Select a source definition and click OK.

The Designer creates an XML Source Qualifier transformation in the mapping and links
each port of the XML source definition to a port in the XML Source Qualifier
transformation.

124 Chapter 6: XML Source Qualifier Transformation

Editing an XML Source Qualifier Transformation
You can edit XML Source Qualifier transformation properties, such as transformation name
and description.

To edit an XML Source Qualifier transformation:

1. In the Mapping Designer, open the XML Source Qualifier transformation.

2. On the Transformation tab, edit the following properties:

Transformation Setting Description

Select Transformation Displays the transformation you are editing. To choose a

different transformation to edit, select it from the list.

Rename Button Edit the name of the transformation.

Description Description of the transformation.

Editing an XML Source Qualifier Transformation 125

 3. Click the Ports tab to view the details of the XML Source Qualifier transformation ports.
Figure 6-1 shows the XML Source Qualifier transformation ports:

Figure 6-1. XML Source Qualifier Transformation Ports

Generated Key
Sequence Numbers

Use the Sequence column to set start values for generated keys in XML groups. You can
enter different values for each generated key. Whenever you change these values, the
sequence numbers restart the next time you run a session using the transformation.
4. Click the Properties tab to configure properties that affect how the PowerCenter Server
runs the mapping during a session.

126 Chapter 6: XML Source Qualifier Transformation

 Table 6-1 lists the XML Source Qualifier properties you can edit:

Table 6-1. XML Source Qualifier Properties

Properties Setting Description

Select Transformation Displays the transformation you are editing. To choose a different transformation to
edit, select it from the list.

Tracing Level Determines the amount of information about this transformation the PowerCenter
Server writes to the session log when it runs the workflow. You can override this
tracing level when you configure a session.

Reset At the end of a session, the PowerCenter Server resets the start values to the start
values for the current session. For more information, see “Setting Sequence Numbers
for Generated Keys” on page 128.

Restart At the beginning of a session, the PowerCenter Server starts the generated key
sequence for all groups at one. For more information, see “Setting Sequence Numbers
for Generated Keys” on page 128.

5. Click the Metadata Extensions tab to create, edit, and delete user-defined metadata
extensions.

Add a metadata
extension.

Delete a metadata
extension.

You can create, modify, delete, and promote non-reusable metadata extensions, as well as
update their values. You can also update the values of reusable metadata extensions. For
more information, see “Metadata Extensions” in the Repository Guide.
6. Click OK.
7. Choose Repository-Save to save changes.

Editing an XML Source Qualifier Transformation 127

 Setting Sequence Numbers for Generated Keys
Each view in the XML Source Qualifier definition has a primary key and sequence value for
the key. During a session, the PowerCenter Server generates keys from sequence values and
increments the values.
At the end of the session, the PowerCenter Server updates each sequence value in the
Repository to the current value plus 1. These values become the start values the next time the
PowerCenter Server runs a session using the transformation.
The Repository maintains the following sequence values:
♦ Default value. The sequence value for a key that displays in the XML Source Qualifier
when you first create it. The default value is 1 for each key.
♦ Start value. A sequence number value for a key at the start of a session. You can view the
start values in the XML Source Qualifier transformation before you run a workflow.
♦ Current value. A sequence value for a key during a session.
The start values for the generated keys display in the Sequence column in the XML Source
Qualifier.
Note: If you edit the sequence start values on the Ports tab, you must save the changes and exit
the Designer before you run a workflow.

Changing Sequence Start Values

You can change sequence start values before or after a session by using the following options
on the XML Source Qualifier transformation Properties tab:
♦ Reset. At the end of a session, the PowerCenter Server resets the start values back to the
start values for the current session. For example, at the beginning of a session, the start
value of a key is 2000. At the end of a session, the current value is 2500. When the session
completes, the start value in the Repository remains at 2000. You might use this option
when you are testing and you want to generate the same key numbers the next time you
run a session.
♦ Restart. At the beginning of a session, the PowerCenter Server restarts the start values
using the default value. For example, if the start value for a key is 1005, and you select
Restart, the PowerCenter Server changes the start value to 1. You might use this option if
your keys are getting large and you will have no duplicate key conflicts if you restart
numbering.

128 Chapter 6: XML Source Qualifier Transformation

Using the XML Source Qualifier in a Mapping
Each group in an XML source definition is analogous to a relational table, and the Designer
treats each group within the XML Source Qualifier transformation as a separate source of
data.
The Designer enforces concatenation rules when you connect objects in a mapping.
Therefore, you need to organize the groups in the XML source definition so that each group
contains all the information you require in one pipeline branch.
Consider the following rules when you connect an XML Source Qualifier transformation in a
mapping:
♦ You can link ports from only one group in an XML Source Qualifier transformation to
ports in one input group of another transformation. You can copy the columns of several
groups to one transformation, but you can link the ports of only one group to the
corresponding ports in the transformation.
Figure 6-2 shows that ports of two groups in one XML Source Qualifier transformation
cannot link to ports in one single input group transformation:

Figure 6-2. Linking XML Source Qualifier Transformations to One Input Group
XML SQ Single Input Group
Transformation Transformation
Group 1
Column11 Column1
Column12 Column2
Group 2 Column3
Column21 X Column4
X
Column22

♦ You can link ports from one group in an XML Source Qualifier transformation to ports
in more than one transformation. Each group in an XML Source Qualifier transformation
can be a source of data for more than one pipeline branch. Data can pass from one group
to several different transformations.
♦ You can link multiple groups from one XML Source Qualifier transformation to
different input groups in a transformation. You can link multiple groups from one XML
Source Qualifier transformation to different input groups in most multiple input group
transformation, such as a Joiner or Custom transformation. However, you can only link
multiple groups from one XML Source Qualifier transformation to one Joiner
transformation if the Joiner has sorted input. To connect two XML Source Qualifier
transformation groups to a Joiner transformation with unsorted input, you must create
two instances of the same XML source. For an example on connecting two XML Source
Qualifier transformations to Joiner transformations, see “Joining Two XML Source
Qualifier Transformation Groups” on page 133.

Using the XML Source Qualifier in a Mapping 129

 Figure 6-3 shows how to link multiple groups from one XML Source Qualifier
transformation to different input groups in a multiple input group transformation:

Figure 6-3. Linking XML Source Qualifier to Multiple Input Group Transformations
XML SQ Joiner Transformation configured
Transformation1 for sorted input
Group 1
Column11 Column1 (master)
Column12 Column2 (master)
Group 2 Column3 (detail)
Column21 Column4 (detail)
Column22

XML SQ Custom Transformation

Transformation2

Group 1 Input Group 1

Column11 Column11
Column12 Column12
Group 2 Input Group 2
Column21 Column21
Column22 Column22

130 Chapter 6: XML Source Qualifier Transformation

XML Source Qualifier Transformation Example
This section shows an XML Source Qualifier transformation in a mapping. Figure 6-4 shows
the element hierarchy for the StoreInfo.xml file:

Figure 6-4. Sample XML File StoreInfo.xml

You might want to calculate the total YTD sales for each product in the StoreInfo.xml
regardless of region. Besides sales, you also want the names and prices of each product. To do
this, you need both product and sales information in the same transformation. However,
when you import the StoreInfo.xml file, the default groups that the Designer creates include a
Product group for the product information and a Sales group for the sales information.

Using the XML Source Qualifier in a Mapping 131

 Figure 6-5 shows the default groups for the StoreInfo file with the product and sales
information in separate groups:

Figure 6-5. Invalid use of XML Source Qualifier Transformation in Aggregator Mapping

You cannot combine groups in a single input

group transformation. The Designer cannot pass
data from both Product and Sales groups to one
input group.

Since you cannot link both the Product and the Sales groups to the same single input group
transformation, you can create the mapping in one of the following ways:
♦ Use a denormalized group containing all required information.
♦ Join the data from the two groups using a Joiner transformation.

Using One Denormalized Group

You can reorganize the groups in the source definition so that all the information you need are
in the same group. For example, you can combine the Product and Sales groups into one
denormalized group in the source definition. One denormalized group enables you to process
all the information for the sales aggregation through one data flow.

132 Chapter 6: XML Source Qualifier Transformation

Figure 6-6 shows a denormalized group Product_Sales containing a combination of columns
from both the Product and Sales groups:

Figure 6-6. Using a Denormalized Group in a Mapping

To create the denormalized group, edit the source definition in the Source Analyzer. You can
either create a new group or modify an existing group. Add to the group all the product and
sales columns you need for the sales calculation in the Aggregator transformation. You can use
the XML Editor to create the group and validate it.
For more information about denormalized groups, see “Generating a Denormalized View” on
page 40.

Joining Two XML Source Qualifier Transformation Groups

You can also get data from two XML source groups into one data flow by joining data from
two groups using a Joiner transformation. When you configure the Joiner transformation for
sorted input, you can link two groups from one XML Source Qualifier transformation
instance to the Joiner transformation. When you use a Joiner transformation configured for
unsorted input, you must use two instances of the same XML source and link a group from
each XML Source Qualifier transformation instance to the Joiner transformation.
You can then send the data from the Joiner transformation to an Aggregator transformation to
calculate the YTDSales for each product.

Using the XML Source Qualifier in a Mapping 133

 Figure 6-7 illustrates how you can create two instances of the same XML source and join data
from two XML Source Qualifier transformations:

Figure 6-7. Using an XML Source Definition Twice in a Mapping

134 Chapter 6: XML Source Qualifier Transformation

Troubleshooting
When I drag two groups from an XML Source Qualifier transformation to a
transformation, the Designer copies the columns but does not link all the ports.
You can link only one group of an XML Source Qualifier transformation to one
transformation. When you drag more than one group to a transformation, the Designer
copies all the column names to the transformation. However, it links the columns of only the
first group.

I cannot break the link between the XML source definition and its source qualifier.
The XML Source Qualifier transformation columns match the corresponding XML source
definition columns exactly. You cannot remove or modify the links between an XML source
definition and its XML Source Qualifier transformation. When you remove an XML source
definition, the Designer automatically removes its XML Source Qualifier transformation.

Troubleshooting 135
136 Chapter 6: XML Source Qualifier Transformation
 Chapter 7

Midstream XML
Transformations
This chapter includes the following topics:
♦ Overview, 138
♦ XML Parser Transformation, 139
♦ XML Generator Transformation, 141
♦ Creating a Midstream XML Transformation, 143
♦ Editing Midstream XML Transformation Properties, 144
♦ Generating Pass-Through Ports, 147

137
Overview
You can use XML definitions to read or create XML data. However, sometimes you need to
extract or generate XML inside a pipeline. For example, you might want to send a message to
a TIBCO target containing an XML document as the data field. In this case, you need to
generate an XML document before sending the message to TIBCO. You can use a midstream
XML transformation to generate the XML.
You can create the following types of midstream XML transformations:
♦ XML Parser transformation. The XML Parser transformation reads XML from one input
port and outputs data to one or more groups.
♦ XML Generator transformation. The XML Generator transformation reads data from one
or more sources and generates XML. It has a single output port.
Use a midstream XML transformation to extract XML data from messaging systems such as
TIBCO, MQSeries, or from other sources, such as files or databases. The XML
transformation functionality is similar to the XML source and target functionality, except it
parses the XML or generates the document in the pipeline.
Midstream XML transformations support the same XML schema components that the XML
Wizard and Editor support. In addition, XML transformations support the following
functionality:
♦ Pass-through ports. You can use pass-through ports to pass non-XML data through the
midstream transformation. These fields are not part of the XML schema definition, but
you can use them to generate denormalized XML groups. You use these fields in the same
manner as top-level XML elements. You can also use a pass-through field as a primary key
for the top-level group in your XML definition. For more information, see “Generating
Pass-Through Ports” on page 147.
♦ Real-time processing. You can use a midstream XML transformation to process data as
BLOBs from messaging systems.
♦ Support for multiple partitions. You can generate different XML documents for each
partition.

138 Chapter 7: Midstream XML Transformations

XML Parser Transformation
Transformation type:
Active
Connected

The XML Parser transformation is similar to an XML source definition. When the
PowerCenter Server processes an XML Parser transformation, it reads a row of XML data,
parses the XML, and passes data through output groups. It also can pass non-XML data.
The XML Parser transformation has one input group, and one or more output groups. The
input group has one input port, “DataInput” which accepts a binary or string data BLOB as
an XML document.
When you create a midstream XML Parser transformation, you use the XML Wizard to
import an XML, DTD, or XML schema file. For example, you can import the following
Employee DTD file:
<!ELEMENT EMPLOYEES (EMPLOYEE+)>
<!ELEMENT EMPLOYEE (LASTNAME, FIRSTNAME, ADDRESS, PHONE+, EMAIL*,
EMPLOYMENT)>
<!ATTLIST EMPLOYEE EMPID CDATA #REQUIRED
DEPTID CDATA #REQUIRED>

<!ELEMENT LASTNAME (#PCDATA)>

<!ELEMENT FIRSTNAME (#PCDATA)>

<!ELEMENT ADDRESS (STREETADDRESS, CITY, STATE, ZIP)>

<!ELEMENT STREETADDRESS (#PCDATA)>
<!ELEMENT CITY (#PCDATA)>
<!ELEMENT STATE (#PCDATA)>
<!ELEMENT ZIP (#PCDATA)>

<!ELEMENT PHONE (#PCDATA)>

<!ELEMENT EMAIL (#PCDATA)>

<!ELEMENT EMPLOYMENT (DATEOFHIRE, SALARY+)>

<!ATTLIST EMPLOYMENT EMPLSTAT (PF|PP|TF|TP|O) "PF">
<!ELEMENT DATEOFHIRE (#PCDATA)>
<!ELEMENT SALARY (#PCDATA)>

XML Parser Transformation 139

 Figure 7-1 shows the XML Parser transformation that the Designer creates from the Employee
DTD file:

Figure 7-1. XML Parser Transformation

DataInput
Group

140 Chapter 7: Midstream XML Transformations

XML Generator Transformation
Transformation type:
Active
Connected

You can use an XML Generator transformation to combine input that comes from several
sources to create an XML document. For example, you can use the transformation to combine
the XML data from two TIBCO sources into one TIBCO target. One source might contain
employee and salary information, and the other might have employee phone and email
information.
The XML Generator transformation is similar to an XML target definition. When the
PowerCenter Server processes an XML Generator transformation, it writes rows of XML data.
The PowerCenter Server can also process pass-through fields containing non-XML data in the
transformation.
Figure 7-2 shows the XML Generator transformation:

Figure 7-2. XML Generator Transformation

Group 1
TIBCO Source 1

Group 2

Group 3
TIBCO Source 2

Group 4

Output TIBCO Target

The XML Generator transformation has one or more input groups and one output group.
The output group has one port, “DataOutput,” which allows a binary or string data BLOB as
an XML document. This group also contains the output port when you create pass-through
fields.

XML Generator Transformation 141

 Figure 7-3 shows a sample XML Generator transformation:

Figure 7-3. Sample XML Generator Transformation

DataOutput Group

142 Chapter 7: Midstream XML Transformations

Creating a Midstream XML Transformation
When you create a midstream XML transformation, you use the XML Wizard and XML
Editor to define the XML groups. You can create the transformation in the Transformation
Developer and the Mapping Designer.

To create a midstream XML transformation:

1. Open the Transformation Developer or the Mapping Designer.

2. Click Transformation-Create.
The Create Transformation dialog box displays.

3. Select the Midstream XML Parser or Midstream XML Generator transformation type.
4. Enter a transformation name, and click Create.
The Import XML Definition dialog box displays.
5. Choose a file to import, and click Open.
The XML Wizard displays.
6. Create the XML definitions using the XML Wizard. For more information about the
XML Wizard, see “Importing an XML Source Definition” on page 57.
7. Click Finish in the XML Wizard.
The midstream XML transformation displays in the workspace.
8. To edit the midstream XML transformation properties, double-click the transformation
in the workspace.

Creating a Midstream XML Transformation 143

Editing Midstream XML Transformation Properties
You can edit some of the midstream XML transformation properties. However, because you
use the XML Wizard and Editor to define the transformation, you must use these tools to
change the XML definition.
If you create a midstream XML transformation in the Mapping Designer, the following rules
apply:
♦ If you make the transformation reusable, you can only change some of the transformation
properties from the Mapping Designer. You cannot add pass-through ports or metadata
extensions.
♦ If you create a non-reusable transformation, you can edit the transformation from the
Mapping Designer.
When you configure a midstream XML transformation, you can configure components on
the following tabs:
♦ Transformation tab. You can rename the transformation and add a description on the
Transformation tab.
♦ Properties tab. You can update the tracing level.
♦ Metadata Extensions tab. You can extend the metadata stored in the repository by
associating information with repository objects, such as an XML transformation. For more
information, see “Metadata Extensions” in the Repository Guide.
♦ XML Parser or XML Generator tab. You can create pass-through ports using this tab.
Pass-through ports enable you to pass non-XML data through the transformation. For the
XML Parser, you can choose to reset sequence numbers if you use sequence numbering to
generate XML column names. For the XML Generator, you can choose to create a new
XML document on commits.

Midstream XML Parser Tab

Use the Midstream XML Parser tab to modify the size of the DataInput port. You can also
add pass-through ports on this tab. For more information about pass-through ports, see
“Generating Pass-Through Ports” on page 147.
You can access the XML Editor from the Midstream XML Parser Tab. Click the XML Editor
button.
Note: When you access the XML Editor, you cannot update Edit Transformations until you
exit the XML Editor.

144 Chapter 7: Midstream XML Transformations

 Figure 7-4 shows the Midstream XML Parser tab:

Figure 7-4. Midstream XML Parser Tab

Add a pass-
through
port.
Edit the
XML
definition.

Reset key
sequence
numbers to
beginning
values after
session
completes.

Always
start key
sequence
numbers at
one.

Table 7-1 shows the options you can change on the Midstream XML Parser tab:

Table 7-1. Midstream XML Parser Settings

Transformation Setting Description

Restart Always start the generated key sequence for all groups at one.

Reset At the end of a session, reset the value sequence for all generated
keys in all groups. This resets the sequence number back to where
it was previously.

Note: The options in Table 7-1 affect the generated key numbers. If you do not choose either
option, the sequence numbers in the generated keys increase from session to session. If you
select the Restart or Reset option, it updates the Restart or Reset property that displays on the
Initialization Properties tab. You cannot change these options from the Initialization
Properties tab, however.

Midstream XML Generator Tab

Use the XML Generator tab to modify the size of the DataOutput port. You can also add
pass-through ports on these tab. For more information about pass-through ports, see
“Generating Pass-Through Ports” on page 147.

Editing Midstream XML Transformation Properties 145

 You can access the XML Editor from the Midstream XML Generator Tab. Click the XML
Editor button. When you access the XML Editor, you cannot edit transformation properties
until you exit the XML Editor.
Figure 7-5 shows the Midstream XML Generator tab:

Figure 7-5. Midstream XML Generator Tab

Table 7-2 shows the options you can change on the XML Generator transformation tab:

Table 7-2. Midstream XML Generator Settings

Transformation Setting Description

On Commit The PowerCenter Server can generate multiple XML documents

after a commit. You can use one of the following options:
- Ignore Commit. The PowerCenter Server creates the XML
document and writes to it at end of file only. Use this option if two
different sources are connected to the XML Generator
transformation.
- Create New Document. Creates a new XML document at each
commit. Use this option if you are running a real-time session.
When a session uses multiple partitions, the PowerCenter Server
generates a separate XML document for each partition, regardless
of On Commit settings. If you select Create New Document, the
PowerCenter Server creates new documents for each partition.

Note: The Designer sets the transformation scope to all input when you choose to ignore
commits. Its sets the transformation scope to the transaction level if you set On Commit to
Create New Doc.

146 Chapter 7: Midstream XML Transformations

Generating Pass-Through Ports
Pass-through ports are columns that pass non-XML data through a midstream XML
transformation. For example, you can pass message IDs with XML for MQSeries sources and
targets. You can use message ID to correlate input and output messages for requests and
replies.
When you define a pass-through port in the midstream transformation, you add it to either
the DataInput group in the XML Parser transformation or the DataOutput group in the
XML Generator transformation.
Once you generate the port, you use the XML Editor to add a corresponding reference port to
another view in the XML definition. In the XML Parser transformation, the pass-through
port is an input port, and the corresponding reference port is an output port. In the XML
Generator transformation, the pass-through port is an output port and the associated
reference port is an input port. See Figure 7-6 on page 149 for an example of an XML
definition with the pass-through port and the reference port.

To create a pass-through port in a midstream XML transformation:

1. Open the transformation in the Transformation Developer or Mapping Designer.

2. Double-click the transformation to open Edit Transformations.
3. Click the Midstream XML Generator or Midstream XML Parser tab.
The window displays either the DataInput or DataOutput port depending on the
transformation type.
4. Click the Add button to add an output port for the pass-through.
A default field displays in the Field Name column.
5. Modify the field name. You can also modify type, precision, and scale depending on the
file you used to create the definition.
6. Click XML Editor to open the XML definition for the transformation.
The XML workspace displays the XML views in your XML definition.
7. Right-click a the top of a view to add the reference port.
8. Select Add a Reference Port.

Generating Pass-Through Ports 147

 The Reference Port dialog box opens.

The dialog box lists the pass-through ports you added in the transformation.
9. Select the pass-through port that will correspond to the new reference port in the view
and click OK.
The corresponding output reference port displays in the view. You can rename the port to
a more meaningful name in the Columns window.

Output
Reference
Port
Input Pass-Through Port

10. Click Apply Changes and exit the XML Editor.

11. Click OK in the transformation.
12. Choose Repository-Save.
Non-XML data comes through the input port called Pass_thru_field and passes through the
corresponding COL_0 reference output port.

148 Chapter 7: Midstream XML Transformations

Figure 7-6 shows an XML Parser transformation with pass-through ports:

Figure 7-6. Pass-Through Ports

Pass-Through Output Port

Pass-Through Input Port

Generating Pass-Through Ports 149

150 Chapter 7: Midstream XML Transformations
 Chapter 8

Working with XML

Sessions
This chapter includes the following topics:
♦ Working with XML Sources in a Session, 152
♦ Working with XML Targets in a Session, 155
♦ Server Handling for XML Targets, 157
♦ Working with Midstream XML Transformations, 165

151
Working with XML Sources in a Session
When you create a session to read data from an XML source, you can configure source
properties for that session. For example, you might want to override the source file name and
location in the session properties.
Figure 8-1 shows the Mapping tab in session properties:

Figure 8-1. Properties Settings for an XML Source

Table 8-1 describes the properties you can override for XML readers in a session:

Table 8-1. XML Reader Options

XML Source Required/

Description
Option Optional

Treat Empty Optional Treat empty XML components as null.

Content as Null

Source File Optional Location of the XML file. By default, the PowerCenter Server looks in the
Directory server variable directory, $PMSourceFileDir.
You can enter the full path and file name. If you specify both the directory and
file name in the Source Filename field, clear this field. The PowerCenter
Server concatenates this field with the Source Filename field when it runs the
session.
You can also use the $InputFileName session parameter to specify the file
directory.
For details on session parameters, see “Session Parameters” in the Workflow
Administration Guide.

152 Chapter 8: Working with XML Sessions

 Table 8-1. XML Reader Options

XML Source Required/

Description
Option Optional

Source Filename Required Enter the file name, or file name and path. Optionally use the $InputFileName
session parameter for the file name.
If you specify both the directory and file name in the Source File Directory
field, clear this field. The PowerCenter Server concatenates this field with the
Source File Directory field when it runs the session. For example, if you have
“C:\XMLdata\” in the Source File Directory field, then enter “filename.xml” in
the Source Filename field. When the PowerCenter Server begins the session,
it looks for “C:\data\filename.xml”.
For details on session parameters, see “Session Parameters” in the Workflow
Administration Guide.

Source Filetype Required The source filetype option enables you to configure multiple file sources by
using a file list. Choose Direct or Indirect. The option indicates whether the
source file contains the source data, or whether it contains a list of files with
the same file properties. Choose Direct if the source file contains the source
data. Choose Indirect if the source file contains a list of files.
When you select Indirect, the PowerCenter Server finds the file list and reads
each listed file when it runs the session. For details on file lists, see the
Workflow Administration Guide.

Table 8-2 describes the properties you can override for an XML Source Qualifier in a session:

Table 8-2. XML Source Qualifier Options for a Session

XML Source Required/

Description
Option Optional

Validate XML Required The Validate XML Source option provides flexibility for validating an XML
Source source against a schema or DTD file. Select Do Not Validate to skip
validation, even if the instance document has an associated DTD or schema
reference. Select Validate Only if DTD is Present to validate only when the
XML source has a corresponding DTD or schema file. The session fails if the
instance document specifies a DTD or schema and one is not present. Select
Always Validate to always validate the XML file. The session fails if the DTD
or schema does not exist or your data is invalid.

Partitionable Optional Allows you to create multiple partitions for the source pipeline.

Server Handling for XML Sources

The PowerCenter Server can distinguish empty values from NULL values in an XML source.
You can choose to pass empty strings as NULL values by selecting the Treat Empty Content
As NULL option in the Mapping tab of the session properties. By default, empty content is
not NULL.
You can choose to omit fixed elements from the XML source definition. If the DTD or XML
schema specifies a fixed or default value for an element, the value appears in the XML source
definition.

Working with XML Sources in a Session 153

 You can define attributes as required, optional, or prohibited in an element tag. You can also
specify fixed or default values for attributes. When a DTD or XML schema contains an
attribute with a fixed or default value, the PowerCenter Server passes the value into the
pipeline even if the element tag in the instance document does not contain the attribute. If
the attribute does not have a fixed or default value, the PowerCenter Server passes a NULL
value for the attribute. A parser error occurs when a required attribute is not present in an
element or a prohibited attribute appears in the element tag. The PowerCenter Server writes
this error to the session log.

154 Chapter 8: Working with XML Sessions

Working with XML Targets in a Session
When you configure a session to load data to an XML target, you define writer properties on
the Mapping tab of the session properties.
Figure 8-2 shows the XML writer properties in the session properties:

Figure 8-2. Properties Settings for an XML Writer

Working with XML Targets in a Session 155

 Table 8-3 describes the properties you define in the XML Writer:

Table 8-3. XML Writer Options

XML Targets Required/

Description
Options Optional

Output File Directory Optional Enter the directory name in this field. By default, the PowerCenter Server writes
output files in the server variable directory, $PMTargetFileDir.
You can enter the full path and file name. If you specify both the directory and
file name in the Output Filename field, clear this field. The PowerCenter Server
concatenates this field with the Output Filename field when it runs the session.
You can also use the $OutputFileName session parameter to specify the file
directory.
For details on session parameters, see “Session Parameters” in the Workflow
Administration Guide.

Output Filename Required Enter the file name, or file name and path. By default, the Workflow Manager
names the target file based on the target definition used in the mapping:
target_name.xml.
If the target definition contains a slash character, the Workflow Manager
replaces the slash character with an underscore.
Enter the file name, or file name and path. Optionally use the $OutputFileName
session parameter for the file name.
If you specify both the directory and file name in the Output File Directory field,
clear this field. The PowerCenter Server concatenates this field with the Output
File Directory field when it runs the session.
For details on session parameters, see “Session Parameters” in the Workflow
Administration Guide.
Note: If you specify an absolute path file name when using FTP, the
PowerCenter Server ignores the Default Remote Directory specified in the FTP
connection. When you specify an absolute path file name, do not use single or
double quotes.

Validate Target Optional Validate XML target data against the schema.

Format Output Optional Format the XML target file so the XML elements and attributes indent.
Otherwise, each line of the XML file starts in the same position.

XML Datetime Required Choose local time, the difference in hours between the server time zone and
Format Greenwich Mean Time or Greenwich Mean Time (UTC).

Null Content Required Choose how to represent null content in the target. For more information, see
Representation “Null and Empty String” on page 158.

Empty String Content Required Choose how to represent empty string content in the target. For more
Representation information, see “Null and Empty String” on page 158.

156 Chapter 8: Working with XML Sessions

Server Handling for XML Targets
You can configure some of the settings the PowerCenter Server uses when it loads data to an
XML target:
♦ Character set. You can configure the PowerCenter Server to run sessions with XML targets
in either ASCII or Unicode data movement mode. For more information about character
sets, see “Character Set” on page 157.
♦ Special characters. The PowerCenter Server adds escape characters to some of the special
characters in XML targets. For more information about special characters, see “Special
Characters” on page 158.
♦ Null and empty string. You can choose how the PowerCenter Server handles null data or
empty strings when it writes data to an XML target. For more information about null and
empty strings, see “Null and Empty String” on page 158.
♦ Handling duplicate group rows. You can choose how the PowerCenter Server handles
rows with duplicate primary keys. For more information about handling duplicate group
rows, see “Handling Duplicate Group Rows” on page 159.
♦ DTD and schema reference. You can specify a DTD or schema file name for the target
XML file. For more information about specifying a DTD or schema reference, see “DTD
and Schema Reference” on page 160.
♦ Flushing XML on commits. You can configure the PowerCenter Server to periodically
flush data to the target. For more information about flushing XML on commits, see
“Flushing XML on Commits” on page 160.
♦ XML Caching properties. You can specify a cache directory for an XML target. For more
information about XML caches, see “XML Caching Properties” on page 162.
♦ Session logs for XML targets. You can view session logs for an XML session. For more
information about locating XML session logs, see “Session Logs for XML Targets” on
page 162.
♦ Multiple XML output. The PowerCenter Server outputs a new XML document if the data
in the root changes. For more information about multiple XML output, see “Multiple
XML Document Output” on page 163.

Character Set
You can configure the PowerCenter Server to run sessions with XML targets in either ASCII
or Unicode data movement mode. XML files contain an encoding declaration that indicates
the code page used in the file. The most commonly used code pages are UTF-8 and UTF-16.
PowerCenter supports UTF-8 code pages for XML targets only.
PowerCenter supports the same set of code pages for XML files that it supports for relational
databases and other files. For details on code page compatibility, see “Globalization
Overview” in the Installation and Configuration Guide. For a list of supported code pages, see
“Code Pages” in the Installation and Configuration Guide.

Server Handling for XML Targets 157

 For XML targets, PowerCenter uses the code page declared in the XML file. When you run
the PowerCenter Server in Unicode data movement mode, the XML target code page must be
a superset of the PowerCenter Server code page and the source code page.

Special Characters
The PowerCenter Server adds escape characters to the following special characters in XML
targets:
< & > ”

Null and Empty String

You can choose how the PowerCenter Server handles null data or empty strings when it writes
elements and attributes to an XML target file. By default, the PowerCenter Server does not
output element tags for NULL values, but it outputs tags with no content for empty strings.
To change these defaults, you can change the Null Content Representation and Empty String
Content Representation XML target properties. Choose one of the following options for each
property:
♦ No Tag. Do not output a tag.
♦ Tag with Empty Content. Output just the tag.
You can choose to output NULL and empty string values as empty content in element tags or
choose not to output element tags in an XML target. You can use both of the following XML
target options for leaf elements or any element that has no multi-occurring elements:
♦ NULL Content Representation. Allows you to choose how the PowerCenter Server
outputs NULL values. By default, the PowerCenter Server does not output element tags
for NULL values. You can choose to output NULL values as empty content in element
tags.
♦ Empty String Content Representation. Allows you to choose how the PowerCenter Server
outputs empty string values. By default, the PowerCenter Server outputs empty content in
element tags for empty string values. You can choose to not output element tags for an
empty string values.
You can specify fixed or default values for elements and attributes. When an element in an
XML schema or a DTD has a fixed or default value, the PowerCenter Server inserts the value
instead of writing empty content. If the XML schema or DTD does not specify a value for an
attribute and the attribute has a NULL value, the PowerCenter Server omits the attribute.
If a required attribute does not have a fixed value, the attribute must be a projected field.
When an attribute has a fixed value, the PowerCenter Server does not project the column and
the column does not appear in the target.
The PowerCenter Server does not output prohibited attributes to a target. An error occurs
when a prohibited attribute appears in an element tag. An error also occurs if a required
attribute is not present in an element tag. The PowerCenter Server writes these errors to the
session log or the error log when you enable row error logging.

158 Chapter 8: Working with XML Sessions

 Table 8-4 shows the formats of XML file elements and attributes that contain null values or
empty strings:

Table 8-4. Null and Empty String Output for XML Targets

Type of Output Type of Data Target File

Element Null <elem></elem>

Empty string <elem></elem>

Attribute Null <elem>...</elem>

Empty string <elem attrib=“”>...</elem>

Handling Duplicate Group Rows

Sometimes duplicate rows occur in source data. The PowerCenter Server can pass only one of
these rows to an XML target. You can configure duplicate row handling in the XML target
session properties. You can also configure the PowerCenter Server to write warning messages
in the session log when duplicate rows occur.
The PowerCenter Server writes duplicate rows to the session log. It does not write duplicate
XML target rows to the reject file.
The PowerCenter Server handles duplicate rows passed to the XML target root group
differently than it handles rows passed to other XML target groups:
♦ For the XML target root group, the PowerCenter Server always passes the first row to the
target. When the PowerCenter Server encounters duplicate rows, it increases the number
of rejected rows in the session load summary.
♦ For any XML target group other than the root group, you can configure duplicate group
row handling in the XML target definition in the Mapping Designer.
♦ If you choose to warn about duplicate rows, the PowerCenter Server writes all duplicate
rows for the root group to the session log. Otherwise, it drops the rows without logging
any error messages.
You can select which row the PowerCenter Server passes to the XML target:
♦ First Row. The PowerCenter Server passes the first row to the target. When the
PowerCenter Server encounters other rows with the same primary key, it increases the
number of rejected rows in the session load summary.
♦ Last Row. The PowerCenter Server passes the last duplicate row to the target. You can
configure the PowerCenter Server to write the duplicate XML rows to the session log by
setting the Warn About Duplicate XML Rows option.
For example, the PowerCenter Server encounters five duplicate rows. If you configure the
PowerCenter Server to write the duplicate XML rows to the session log, the PowerCenter
Server passes the fifth row to the XML target and writes the first four duplicate rows to the
session log. Otherwise, the PowerCenter Server passes the fifth row to the XML target but
does not write anything to the session log.

Server Handling for XML Targets 159

 ♦ Error. The PowerCenter Server passes the first row to the target. When the PowerCenter
Server encounters a duplicate row, it increases the number of rejected rows in the session
load summary and increments the error count.
When the PowerCenter Server reaches the error threshold, the session fails. When a session
containing an XML target fails, the PowerCenter Server does not write any row to the
XML target.
The PowerCenter Server sets an error threshold for each XML group. For details on the
error threshold, see “Working with Sessions” in the Workflow Administration Guide.

DTD and Schema Reference

When you edit the XML target in the Warehouse Designer, you can also specify a DTD or
schema file name for the target XML file. The PowerCenter Server adds a document type
declaration or schema reference to the target XML file and inserts the name of the file you
specify. For example, if you have a target XML file with the root element TargetRoot and you
set the DTD Reference option to TargetDoc.dtd, the PowerCenter Server adds the following
document type declaration after the XML declaration:
<!DOCTYPE TargetRoot SYSTEM "TargetDoc.dtd">

The PowerCenter Server does not check that the file you specify exists or that it is valid. The
PowerCenter Server does not validate the target XML file against the DTD or schema file you
specify.
Note: An XML instance document must refer to the full relative path of a schema if a
midstream XML transformation is processing the file. Otherwise, the full path is not required.

Flushing XML on Commits

When you process an XML file or stream, the XML parser takes the entire XML file, parses it,
and writes target XML data at end of file. The On Commit attribute enables you to
periodically flush the data to the target before reaching end of file. You can flush data
periodically into one target XML document, or you can generate multiple XML documents.
You might want to flush XML data in the following situations:
♦ Large XML files. If you are processing a large XML file of several gigabytes, the
PowerCenter Server may have reduced performance or it may run out of memory. You can
set the On Commit attribute to Append to Doc. This flushes XML data periodically to the
target document.
♦ Real-time processing. If you process real-time data that requires commits at specific times,
use Append to Doc.
You can set the On Commit attribute to one of the following values:
♦ Ignore commit. Generate and write to the XML document at end of file only.
♦ Append to document. Write to the same XML document at the end of each commit. The
XML document closes at end of file. This option is not available for XML Generator
transformations.

160 Chapter 8: Working with XML Sessions

♦ Create new document. Create and write to a new document at each commit. You create
multiple XML documents.
Warning: When you create new a document on commit, you need to provide a unique filename
for each document. Otherwise, the PowerCenter Server overwrites the document it generated
on the previous commit.
You can flush data only if all groups in the XML target are connected to the same single
commit or transaction point. The transformation at the commit point generates denormalized
output. The denormalized output contains repeating primary key values for all but the lowest
level node in the XML schema. The PowerCenter Server extracts rows from this output for
each group in the XML target.
You must have only one child group for the root group in the XML target.

Ignoring Commit
You can choose to generate the XML document after the session has read all the source
records. This option causes a session to store all of the XML data in cache as it processes, so
you should use this option when you are not processing a lot of data.

Appending to Document on Commit

When you append data to an XML document, you can use a source-based or user-defined
commit in the session. Use a single point in the mapping to generate transactions. All the
projected groups of an XML target must belong to the same transaction control unit.
For sessions using source-based commits, the single transaction point might be a source or
nearest active source to the XML target, such as the last active transformation before the
target. For sessions with user-defined commits, the transaction point is a transaction
generating transformation.

Creating XML Documents on Commit

You can choose to generate a separate XML document for each commit. To generate multiple
XML output documents, set On Commit to Create New Document. To define the commit,
you can turn on source-based commit in the session, or you can generate the commit from a
transaction generating transformation in the mapping.
When the PowerCenter Server generates multiple XML documents, it creates files with the
following name format:
<base file name>.number.xml

The base file name is a local file name for regular files and a remote file name for FTP files.
Each XML document that the PowerCenter Server creates contains the base file name and a
number, starting with 1. The PowerCenter Server generates a file list after writing all the
output files for the XML target. The file list contains all the output file names. For local files,
the file list contains a list of all fully qualified file names, one per line in server code page. For
FTP files, the file list contains file names with no path component.

Server Handling for XML Targets 161

 The file list name syntax is:
<base file name>.lst

Each time you run a session, the PowerCenter Server overwrites the target XML files and
generates a new file list. The PowerCenter Server places the file list into the XML target
directory.
Note: The first file that the PowerCenter Server generates does not contain a number. It is the
default base file. This file is included in the list file if the server writes data to it. If, however,
you use a FileName column to create file names, the server deletes the default file when
processing completes.
If the base file name does not have an extension, the PowerCenter Server appends a period
and the number at the end of the name. For example, instance “abc” becomes “abc.1”. If the
extension of a base file name is not “.xml” the PowerCenter Server retains the extension, but
applies a period and a number before the extension. For example “abc.txt” becomes
“abc.1.txt.”

XML Caching Properties

The PowerCenter Server uses a data cache to store XML row data while it generates an XML
document. The cache size is the sum of all the groups in the XML target instance. It includes
a primary key and a foreign key index cache for each XML group and one data cache for all
groups.
The default cache size is 10,000,000 bytes. When the memory requirements exceed the cache
size, the PowerCenter Server pages data to index and data files in the cache directory. When
the session completes, the PowerCenter Server releases cache memory and deletes the cache
files.
You can specify the cache directory and cache size for the XML target. The default cache
directory is $PMCacheDir, which is a server variable that represents the directory where the
PowerCenter Server stores cache files by default.

Session Logs for XML Targets

When you run a session with an XML target, the PowerCenter Server writes the target name
and group name into the session log. The session log lists the target and group names in the
following format:
Target Name : : Group Name.

For example, the following session log excerpt records the PowerCenter Server loading a target
table to the group DEPARTMENT in the target EMP_SALARY:
WRITER_1_1_1> WRT_8167 Start loading table [EMP_SALARY::DEPARTMENT] at:
Wed Nov 05 08:01:35 2003

162 Chapter 8: Working with XML Sessions

Multiple XML Document Output
The PowerCenter Server generates a new XML document for each distinct primary key value
in the root group of the target. To create separate XML files, you must pass data to the root
node primary key. When the value of the key changes, the PowerCenter Server creates a new
target file.
The PowerCenter Server creates multiple XML files when the root group has more than one
distinct primary key value. If the PowerCenter Server receives multiple rows with the same
primary key value, it chooses the first or last row based on the way you configure duplicate
row handling.
If you pass data to a column in the root group, but you do not pass data to the primary key,
the PowerCenter Server does not generate a new XML document. It writes a warning message
to the session log indicating that the primary key for the root group is not projected, and the
PowerCenter Server is generating only one document.

Example
The following example includes a mapping that contains a flat file source of country names,
regions, and revenue dollars per region. The target is an XML file. The root view contains the
primary key, XPK_COL_0, which is a string.
Figure 8-3 shows data mapped to the root of the XML definition:

Figure 8-3. Mapping Data to an XML Root

Each time the PowerCenter Server passes a new country name to the root view it generates a
new target file. Each target XML file contains country name, region, and revenue data for one
country.
The PowerCenter Server passes the following rows to the XML target:
Country,Region,Revenue
USA,region1,1000
Canada,region1,100
USA,region2,200
USA,region3,300
USA,region4,400
France,region1,10
France,region2,20
France,region3,30
France,region4,40

Server Handling for XML Targets 163

 The PowerCenter Server builds the XML files in cache. It creates one XML file for USA, one
file for Canada, and one file for France. If you specify “revenue_file.xml” as the output file
name in the session properties, the session produces the following files:
revenue_file.xml - contains the Canada rows
revenue_file.1.xml - contains the France rows
revenue_file.2.xml - contains the USA rows
revenue_file.xml.lst - contains a list of each XML file the session created

If the data has multiple root rows with circular references, but none of the root rows has a null
foreign key, the PowerCenter Server cannot find a root row. It outputs the following message
in the session log file:
XMLW_31108 Error: An appropriate start row was not found for XML root
group [B] with circular reference. No output was generated.

You can add a FileName column to XML targets in order to name XML output documents
based on data values.

164 Chapter 8: Working with XML Sessions

Working with Midstream XML Transformations
When you configure a session using a midstream XML transformation, you define properties
on the Mapping tab of the session properties.
Figure 8-4 shows the XML Generator transformation session properties:

Figure 8-4. Properties Settings for an XML Generator Transformation

Table 8-5 describes the properties you define in the XML Generator transformation:

Table 8-5. XML Generator Transformation Session Options

XML Targets Required/

Description
Options Optional

Validate Output Optional Validate XML target data against the schema.

Format Output Optional Format XML output so the XML elements and attributes indent. Otherwise,
each line of the XML file starts in the same position.

XML Datetime Required Select local time, local time with time zone, or UTC. Local time with time zone
Format is the difference in hours between the server time zone and Greenwich Mean
Time. UTC is Greenwich Mean Time.

Working with Midstream XML Transformations 165

 Table 8-5. XML Generator Transformation Session Options

XML Targets Required/

Description
Options Optional

Null Content Required Select No Tag or Tag with Empty Content. For more information see “Null and
Representation Empty String” on page 158.

Empty String Content Required Select No Tag or Tag with Empty Content. For more information see “Null and
Representation Empty String” on page 158.

Duplicate Group Row Required Select First Row, Last Row, or Error. For more information, see “Handling
Handling Duplicate Group Rows” on page 159.

Orphan Row Required Orphan rows are child rows that are missing parent data. Select Ignore to
Handling continue the session and ignore the orphan rows. Select Error to abort the
session when orphan rows occur.

DTD Reference Optional Associated DTD or XML schema file name to add to the XML file the
transformation creates. You must fully qualify this file name when you use it
with XML Generator transformations. For more information about using the
DTD or schema file name, see “DTD and Schema Reference” on page 160.

Cache Size Required The total size in bytes for the cache memory used by the transformation. The
default is 10,000,000 bytes.

Cache Directory Required The directory for the XML cache files. The default is the $PMCacheDir server
variable. For more information about working with caches, see “XML Caching
Properties” on page 162.

166 Chapter 8: Working with XML Sessions

Figure 8-5 shows the XML Parser transformation session properties:

Figure 8-5. Properties Settings for an XML Parser Transformation

Table 8-6 describes the properties you define in the XML Parser transformation:

Table 8-6. XML Parser Transformation Session Options

XML Targets Required/

Description
Options Optional

Validate XML Source Optional Validate XML source data against the schema.

Treat Empty Content Required Select No Tag or Tag with Empty Content. For more information see “Null and
as Null Empty String” on page 158.

Working with Midstream XML Transformations 167

168 Chapter 8: Working with XML Sessions
 Appendix A

XML Datatype Reference

This appendix includes the following topics:

♦ XML and Transformation Datatypes, 170

169
XML and Transformation Datatypes
PowerCenter supports all XML datatypes specified in the W3C May 2, 2001
Recommendation. Table A-1 lists the XML datatypes and compares them to the
transformation datatypes that display in the XML Source Qualifier transformation. For
details on XML datatypes, see the W3C specifications for XML datatypes at http://
www.w3.org/TR/xmlschema-2.
For more information about using transformation expressions and functions to convert
datatypes, see “Functions” in the Transformation Language Reference.
When you pass data to the target, make sure that it is in the correct format so that the
PowerCenter Server writes the data correctly in the target XML file.
You can change XML datatypes in XML definitions and in midstream XML transformations
if you import an XML file to create the definition. You cannot change XML datatypes when
you import them from an XML schema, and you cannot change the transformation datatypes
for XML sources within a mapping.
Table A-1 shows the XML and corresponding Transformation datatypes:

Table A-1. XML and Transformation Datatypes

Datatype Transformation Range

anyURI String 1 to 104,857,600 characters

base64Binary Binary 1 to 104,857,600 bytes

boolean Small Integer Precision 5; scale 0

byte Small Integer Precision 5; scale 0

date Date/Time Jan 1, 1753 AD to Dec 31, 9999 AD (precision to the second)

dateTime Date/Time Jan 1, 1753 AD to Dec 31, 9999 AD (precision to the second)

decimal Decimal Precision 1 to 28; scale 0 to 28

double Double Precision 15, scale 0

duration String 1 to 104,857,600 characters

ENTITIES String 1 to 104,857,600 characters

ENTITY String 1 to 104,857,600 characters

float Double Precision 15, scale 0

gDay Integer Precision 10; scale 0

gMonth Integer Precision 10; scale 0

gMonthDay Date/Time Jan 1, 1753 AD to Dec 31, 9999 AD (precision to the second)

gYear Integer Precision 10; scale 0

gYearMonth Date/Time Jan 1, 1753 AD to Dec 31, 9999 AD (precision to the second)

170 Appendix A: XML Datatype Reference

 Table A-1. XML and Transformation Datatypes

Datatype Transformation Range

hexBinary Binary 1 to 104,857,600 bytes

ID String 1 to 104,857,600 characters

IDREF String 1 to 104,857,600 characters

IDREFS String 1 to 104,857,600 characters

int Integer Precision 10; scale 0

integer Integer Precision 10; scale 0

language String 1 to 104,857,600 characters

long Integer Precision 10; scale 0

Name String 1 to 104,857,600 characters

Ncname String 1 to 104,857,600 characters

negativeInteger Integer Precision 10; scale 0

NMTOKEN String 1 to 104,857,600 characters

NMTOKENS String 1 to 104,857,600 characters

nonNegativeInteger Integer Precision 10; scale 0

nonPositiveInteger Integer Precision 10; scale 0

normalizedString String 1 to 104,857,600 characters

NOTATION String 1 to 104,857,600 characters

positiveInteger Integer Precision 10; scale 0

QName String 1 to 104,857,600 characters

short Small Integer Precision 5; scale 0

string String 1 to 104,857,600 characters

time Date/Time Jan 1, 1753 AD to Dec 31, 9999 AD (precision to the second)

token String 1 to 104,857,600 characters

unsignedByte Small Integer Precision 5; scale 0

unsignedInt Integer Precision 10; scale 0

unsignedLong Integer Precision 10; scale 0

unsignedShort Integer Precision 10; scale 0

Unsupported Datatypes
PowerCenter does not support the following XML datatypes:
♦ binary

XML and Transformation Datatypes 171

 ♦ century
♦ month
♦ nstring
♦ number
♦ recurringDate
♦ recurringDay
♦ recurringDuration
♦ timeDuration
♦ timeInstant
♦ timePeriod
♦ uriReference
♦ year

XML Date Format

PowerCenter supports the following date, time, and datetime datatype format:
CCYY-MM-DDThh:mm:ss:sss

You can use this format or any portion of this format if it conforms to the XML schema
specifications. For example, an element of type date, time, or datetime may use either of the
following formats within a session:
CCYY-MM

or
CCYY-MM-DD/Thh

PowerCenter does not support negative dates for datetime format.

The format of the first datetime element in an XML file determines the format of all
subsequent values of the element.
For example, if the first value of a datetime element uses the format
CCYY-MM-DDThh:mm:ss

then all the subsequent values for the same element must follow the same format.
If the PowerCenter Server reads in a value for the same date, time, or datetime element that
has a different format, even if it is a valid dateTime format, it rejects the row.
In this example, if the Informatica reads in a subsequent value with a different format
CCYY-MM-DD

then it rejects the row even if the new format is also a valid datetime format.

172 Appendix A: XML Datatype Reference

Tip: If you generate an XML target from a relational source, ensure that you use the supported
date, time, and datetime format specified in the associated schema or DTD file to ensure
validation.

XML and Transformation Datatypes 173

174 Appendix A: XML Datatype Reference
 Index

A C
absolute cardinality (XML) caching XML
description 14 properties 162
abstract elements cardinality
description 21 absolute 14
using in a mapping 115 relative 15
advanced mode types 14
using the XPath Navigator 83 child element (XML)
all group (XML) overview 4
description 24 choice group (XML)
all hierarchy foreign keys option description 23
overview 100 circular references
ANY element type (XML) description 47
in the XML Editor 85 non-recursive row option 100
overview 21 code pages
append to document importing XML sources 57
flushing XML 161 XML file 6, 27
arrange columns
organizing views in workspace 94 adding to XML views 82
atomic types deleting from an XML view 86
description 17 generating names 60
attributes pivoting 51
DTD syntax 7 size limitations 102
XML 60 Columns window
overview 81
setting XML View Options 100
commit
flushing XML 160

175
complex types XML targets 104, 107
creating type relationships 90 XML Views 82
description 19 element type ANY (XML)
expanding 84 overview 21
extended 19 elements
in XML schemas 19 description 3
restricted 19 DTD syntax 7
viewing the hierarchy 96 empty strings
composite keys (XML) XML target files 158
overview 54 enclosure element (XML)
concatenated columns (XML) creating views for 59
description 54 XML hierarchy 4
creating encoding declaration
a new XML view in workspace 81 describing code page in XML 6
relationships between XML views 89 entity relationships
custom XML groups generating 42
description 37 generating XML views 63
skip create view 65 enumeration
description 17
searching for values 95
D escape characters
in XML targets 158
datatypes
unsupported XML 171
XML metadata 13
default value
F
XML attributes 115 facets
default_value description 17
for a DTD attribute 7 filename column
deleting passing to XML target 116
columns from XML view 86 #FIXED option
denormalized XML groups description 7
description 40 flushing data
documentation appending to document 161
conventions xxvii create new documents 161
description xxvi ignore commit 161
online xxvii XML targets 116
DTD file Force Row option
description 7 overview 101
metadata from DTD files 32 foreign keys
schema reference 160 generating values for all in a view 100
DTM buffer size errors
fixing 59
duplicate group row handling G
overview 159
generated keys
description 36
E sequence numbering 126
generating column names
editing setting option 60
XML source definitions 70 global declarations
XML Source Qualifier transformation 125 option to create 59

176 Index
global element (XML) legend
overview 4 XML Editor icons 79
limitations
XML sources and targets 54
H lists (XML)
description 18
hierarchy (XML) local element (XML)
description 13 overview 4
hierarchy relationships
circular references 47
element relationships 13
generating 38
M
Relationship Row option 101 mappings
connecting abstract elements 115
using XML targets 114
I XML Source Qualifier transformation 129
XML target ports 115
ignore commit message IDs
flushing XML 161 XML Generator transformations 147
ignore fixed element metadata
setting option 59 viewing as an XML schema 97
ignore prohibited attributes metadata explosion
setting options 60 example 43
#IMPLIED option reducing 66
description 7 metadata extensions
importing in XML source qualifiers 127
XML sources 57 in XML sources 71
XML targets 104, 105 in XML targets 110
infinite length columns midstream XML transformation
overriding 102 creating 143
infinite precision general properties 144
overriding 59 Generator properties 145
Informatica overview 138
documentation xxvi Parser properties 144
Webzine xxviii pass-through ports 147
reset generated key sequence 145
mode button
K using the XPath Navigator 83
multiple XML output
keys example 162
generated key sequence numbers 126 generating 161
generating multiple foreign keys in a view 100 multiple-occurring element (XML)
XML view keys and relationships 36 overview 4

L N
layout name tag
arranging views 94 description 13
leaf element (XML) namespace
overview 4 description 11
updating in XML Editor 93

Index 177
naming columns pivoting
option 60 deleting pivoting columns 86
Navigator in Advanced Options 59
viewing complex types 96 rules 51
viewing simple types 96 with Xpath Navigator 82
new line character XML columns 51
XML attributes 60 ports
non-recursive row option pass-through 147
description 100 XML Source Qualifier transformation 129
normalized XML groups XML targets 115
description 38 PowerCenter Server
null constraint (XML) handling XML targets 157
description 16 precision
null data overriding infinite length 59
XML target files 158 prefix
updating namespace 93
previewing
O XML data 97
properties
on commit midstream XML transformation 144
append to document 161 XML caching 162
create new documents 161 XML Generator transformation 145
ignore commit 161 XML Parser transformation 144
options 160
options
for creating views 59
output files
R
session properties 156 reference ports
adding to views 147
finding in workspace 87
P relative cardinality (XML)
description 15
parent chain (XML) #REQUIRED option
description 5 description 7
parent element (XML) reset
description 5 midstream generated key sequence 145
partitionable restart
XML source option 153 midstream generated key sequence 145
passive transformations root element
XML Source Qualifier 122 specifying in a target 114
pass-through ports
adding to XML views 144, 145
finding the reference ports 87
generating 147
S
overview 87 saving
pattern facet metadata as XML/DTD/schema 97
description 17 schema
files 9
location in namespace 93
searching for components 95
searching in XML Editor 95
simple types 17

178 Index
searching transaction control point
in schema 95 XML targets 116
in XML Editor 95 transformation
select columns XML Source Qualifier 122
creating type relationships 90 transformation datatypes
sequence comparing to XML 170
numbering generated keys 126 troubleshooting
sequence group (XML) XML Source Qualifier transformation 135
description 23 XML sources 74
server handling XML targets 118
XML sources 153 type relationships
XML targets 157 creating in the workspace 90
session logs Type Relationship Row option 101
XML targets 162
session properties
output files, XML 156 U
XML Generator transformation 165
XML Parser transformation 167 unions (XML)
XML sources 152 description 18
XML targets 156
simple types
description 17 V
viewing a hierarchy 96
single-occurring element (XML) validating
overview 5 target rules 112
Skip Create XML Views validate target option 156
setting custom views 65 XML definitions 99
source filename XML source option 153
XML sources option 153 view
source filetype arranging in workspace 94
XML source option 153 creating relationships 89
sources generating entity relationships 42
creating XML targets from XML sources 106 hierarchical relationships 38
special characters setting options 59
parsing 158 view row
start value description 49
generated keys 126 using the Force Row option 101
substitution groups viewing
example 45 XML metadata as XML/DTD/Schema 97
in XML schema files 24
synchronizing
XML definitions 68 W
webzine xxviii

T
targets X
duplicate group row handling 159
XML
setting DTD/schema reference 160
attributes 60
specifying a root element 114
character encoding 27
code pages 27, 57

Index 179
 comparing datatypes to transformation 170 XML metadata
datatypes 170 cardinality 14
description 3 datatype 13
metadata from 31 description of types 11
previewing data 97 from DTD file 32
special characters 158 from substitution groups 24, 45
synchronizing definitions with schemas 68 from XML file 31
XML definitions from XML schema 33
creating from flat files 72 hierarchy 13
creating from relational files 72 name 13
editing source definition properties 70 null constraint 16
synchronizing with sources 68 viewing as a schema 97
XML Editor XML Parser transformation
adding columns to views 82 example 139
Columns window 81 overview 138
creating a new view 81 session properties 167
deleting columns 86 XML rules
editing views 82 pivoting groups 51
icons legend 79 XML source and target groups from relational tables 35
options 59 XML target port connections 115
pass-through fields 144, 145 XML schema
searching for components 95 complex types 19
validating definitions 99 metadata from 33
XML Generator transformation setting default attributes 115
example 141 XML Source Qualifier transformation
overview 138 adding to mapping 123
pass-through ports 147 automatically creating 123
session properties 165 editing 125
XML groups manually creating 123
all group 24 overview 122
choice group 23 port connections 129
custom 37 troubleshooting 135
denormalized groups 40 using in a mapping 129
element and attribute groups 23 XML sources
modifying source groups 57 creating targets from 106
normalized groups 38 editing XML definitions 70
options for creating 59 importing 57
source and target groups from relational tables 35 limitations 54
substitution groups 24, 45 partitionable option 153
XML hierarchy server handling 153
child element 4 session properties 152
creating hierarchy relationships 64 source filename 153
enclosure element 4 source filetype option 153
global element 4 source location 152
leaf element 4 troubleshooting 74
local element 4 validate option 153
multiple-occurring element 4 XML targets
parent chain 5 active sources 114
parent element 5 creating 104
single-occurring element 5 creating from XML sources 106
duplicate group row handling 159

180 Index
 editing 104, 107
editing target properties 110
flushing data 116
groups from relational tables example 35
importing 104, 105
in sessions 155
limitations 54
multi-line attributes 60
On Commit session property 116
outputting multiple files 161
overview 104
port connections 115
server handling 157
session log entry 162
session properties 156
setting default attributes 115
setting DTD/schema reference 160
troubleshooting 118
using in mapping 114
validate option 156
XML View Options
setting in the Columns window 100
XML views
adding columns 82
adding pass-through fields 144, 145
creating 61
creating a new 81
creating hierarchy relationships 64
creating relationships between 89
creating with XML Wizard 61
generating custom views 65
generating entity relationships 63
pivoting columns 51
Skip Create XML View option 65
XML Wizard
generating custom XML views 65
generating entity relationships 63
generating hierarchy relationships 64
importing sources 57
selecting root elements 66
synchronizing XML definitions 68
XPath
See also XML Path
expanding complex types 84
using the Navigator 82

Index 181
182 Index