What's New in Oracle Text?

This section describes new features of the Oracle Database 11g Release 2 (11.2) edition of Oracle Text and provides pointers to additional information. Information about new features from previous releases is also retained to help you migrate to the current release.

The following sections describe the new features in Oracle Text:

• Oracle Database 11g Release 1 (11.1) New Features in Oracle Text

• Oracle Database 10g Release 2 (10.2) New Features in Oracle Text

Oracle Database 11g Release 1 (11.1) New Features in Oracle Text

• On Windows systems, the executable that you specify for the USER_FILTER command attribute must now exist in the %ORACLE_HOME%/ctx/bin directory instead of %ORACLE_HOME%/bin.

  See Also:

  USER_FILTER

• Zero downtime for applications with new incremental indexing and online index creation.

  See Also:

  "Creating a CONTEXT Index Incrementally with POPULATE_PENDING" in Oracle Text Application Developer's Guide

• New features for recreating an index online and finer control for maintenance processes.

  See Also:

  "Recreating an Index" in Oracle Text Application Developer's Guide

• New Oracle Text Manager in Oracle Enterprise Manager with which you can:

  Monitor health of Text indexes.

  Modify index settings.

  Generate index-level statistics about disk space, fragmentation, garbage, frequency of words, and more.

  Synchronize, optimize, and rebuild indexes.

  Diagnose problems, and resume failed operations.

  Manage logs.

  See Also:

  "Text Manager in Oracle Enterprise Manager" in Oracle Text Application Developer's Guide

• New support for Composite Domain Index for CONTEXT indextype for improved mixed query performance.

  See Also:

  "Composite Domain Index (CDI) in Oracle Text" in Oracle Text Application Developer's Guide

• Improved query performance and scalability.

  See Also:

  "Parallelizing Queries Across Oracle RAC Nodes" in Oracle Text Application Developer's Guide

• New SDATA section type and SDATA operator that enables range searches on metadata.

  See Also:

  "SDATA Section" in Oracle Text Application Developer's Guide

• New user-defined scoring feature, DEFINESCORE and DEFINEMERGE.

  See Also:

  "DEFINESCORE" and "DEFINEMERGE"

• New AUTO_LEXER lexer type that performs language identification, word segmentation, document analysis, part-of-speech tagging, and stemming. The AUTO_LEXER also enables customization of these components.

  See Also:

  "AUTO_LEXER"

• New values for INDEX_STEMS attribute of BASIC_LEXER to enable better query performance for stem ($) queries.

  See Also:

  "BASIC_LEXER"

• NOPOPULATE option for ALTER INDEX to support incremental indexing.

  See Also:

  "POPULATE | NOPOPULATE"

• New limit for the number of partitions in Oracle Text is now the same as the maximum for Oracle Database.

  See Also:

  "Partitioned Tables and Indexes" in Oracle Text Application Developer's Guide

• New feature usage tracking.

  See Also:

  "Database Feature Usage Tracking in Oracle Enterprise Manager" in Oracle Text Application Developer's Guide

Oracle Database 10g Release 2 (10.2) New Features in Oracle Text

• New AUTO_FILTER Filter

  With Oracle Text 10g Release 2, the INSO_FILTER filter has been deprecated in favor of a new filter, AUTO_FILTER. AUTO_FILTER is backward-compatible with INSO_FILTER.

  Additionally, the INSO_TIMEOUT and INSO_FORMATTING attributes of the MAIL_FILTER have been deprecated in favor of AUTO_FILTER_TIMEOUT and AUTO_FILTER_OUTPUT_FORMATTING, respectively. Moreover, the INSOFILTER directive used in the mail configuration file of the MAIL_FILTER has been deprecated in favor of the new AUTO_FILTER directive.

  The system-defined preference CTXSYS.INSO_FILTER has also been deprecated in favor of a new preference, CTXSYS.AUTO_FILTER.

  With these changes, the list of document formats supported by Oracle Text has changed.

  See Also:

  Filter Types, Appendix B, "Oracle Text Supported Document Formats", and the Migration chapter of the Oracle Text Application Developer's Guide

• Changes in Asian Language Support

  Chinese, Japanese, and Korean now support the CTXRULE index type. All three languages also support mixed-case query searches, as does the WORLD_LEXER.

  Additionally, the KOREAN_LEXER has been deprecated. Use the KOREAN_MORPH_LEXER instead.

  See Also:

  Lexer Types

• New Stopwords

  New default stopwords have been provided for English, Finnish, Italian, Spanish, and Swedish.

  See Also:

  Appendix E, "Oracle Text Supplied Stoplists"

• Key Word in Context (KWIC)

  Two new procedures, CTX_DOC.SNIPPET and CTX_DOC.POLICY_SNIPPET, return text fragments containing keywords found in documents. This format enables users to see the keywords in their surrounding text, providing context for them.

  See Also:

  SNIPPET and POLICY_SNIPPET

• New ALTER INDEX Syntax

  ALTER INDEX now has two new parameters. ALTER INDEX PARAMETERS enables you to modify the parameters of a non-partitioned index or a local partitioned index (including all partitions) without rebuilding the index.

  This command works at the index level.

  ALTER INDEX MODIFY PARTITION PARAMETERS enables you to modify the metadata of an index partition.

  See Also:

  ALTER INDEX

• New Procedure for Handling Failed Index Creation

  The new CTX_ADM.MARK_FAILED procedure enables you to change an index's status from LOADING to FAILED; such a change is useful when CREATE or ALTER INDEX fails and it is necessary to recover the index.

  See Also:

  MARK_FAILED