Creating a Text Configuration Object

Create a text configuration for use with a text index.

Prerequisites

Authenticate with Sybase IQ using an account that has DBA or Resource authority.

If you have DBA authority, you can assign any user as the owner of text configuration objects you create.
If you have Resource authority, all text configuration objects you create are owned by you.

Task

For detailed information on stoplists, see the Sybase IQ documentation: Unstructured Data Analytics in Sybase IQ > TEXT Indexes and Text Configuration Objects > Text Configuration Objects > Text Configuration Object Settings > Stoplist Setting (STOPLIST).

From the application menu bar, select View > Open > Administration Console.
In the Administration Console, expand IQ Servers > Schema Objects.
Select Text Configuration Objects, click the drop-down arrow that appears to the right, and select New.
In the wizard, select the Sybase IQ server on which to create the text configuration.
Select the Sybase IQ user who will own the text configuration.
Enter a name for the text configuration.
Select the database collation for this text configuration. Each database has a CHAR collation and an NCHAR collation, configured when the database was created. Sample collations:
- ISO_BINENG – Binary ordering, English ISO/ASCII 7-bit letter case mappings
- UCA – Standard default Unicode Collation Algorithm collation
(Optional) Enter a comment describing the text configuration.
Choose a term breaker algorithm and specify the minimum and maximum term lengths.
(Optional) If you chose the Generic term breaker algorithm, you can specify an external library function to break the text into terms. Use one of these formats:
- function-name@library-file-name
- Windows-function-name@library-file-name.dll
- UNIX:UNIX-function-name@library-file-name.so
(Optional; available only for CHAR collations) Specify an external library to perform document filtering before term breaker processing.
(Optional) Create a stoplist by entering terms to omit from the text index. (Terms in the stoplist are also ignored in queries.) Separate terms with spaces.
Many nonalphanumeric characters are ignored in stoplists; others (including spaces, apostrophes, and dashes) are interpreted as term delimiters. Consequently, including contractions and hyphenated terms may lead to undesirable results, even when you enclose the terms in quotes. How the stoplist is parsed depends on the term breaker and term lengths you specified in step 9 and step 10.
Click Finish to create the text configuration.

Next

Create a text index that uses your new text configuration.