Configuring Oracle search in language_languageCode.xml

Guidewire provides the ability to configure language-appropriate linguistic search capabilities through the <LinguisticSearchCollation> element. This subelement of <GWLanguage> is defined in language_languageCode.xml. You use the strength attribute of this subelement to configure and control specialized search behavior.

The meaning of the strength attribute depends on the specific language. In general, the settings mean the following:

• A strength of primary considers only character weights. This setting instructs the search algorithms to consider just the base, or primary letter, and to ignore other attributes such as case or accents. Thus, the collation rules consider the characters é and E to have the same weight. For more information on this attribute, see Configuring database sort in language_languageCode.xml.
• A strength of secondary, the default, considers character weight and accent differences, but not case differences. Thus, the collation rules consider the characters e and é to be different and thus the rules treat them differently. The collation rules do not, however, treat e and E differently.

To summarize, the strength attribute can take the following values, with the default being secondary.

The following language_ja.xml file is an example for Japanese with suggested settings.

<?xml version="1.0" encoding="UTF-8"?>
<Language xmlns="http://guidewire.com/language">
  <GWLanguage code="ja" name="Japanese" typecode="ja">
    <LinguisticSearchCollation strength="primary"/>
    <SortCollation strength="primary"/>
  </GWLanguage>
</Localization>