Date and time formats in indexes

Date and time indexes can be adapted to several languages and regions Recognized formats are based on the ICU (International Component for Unicode) and ISO 8601 standards.

Each format is associated with a language to allow recognition of spelled month and days (January, February, ..., Monday, Tuesday, etc.) in any supported language.

To specify a date or a time format, you must use the <format> element and specify the appropriate language locale in its NAME attribute.

For example, to index American English dates written as Wednesday, May 15, 2002 and French dates written as Mercredi 15 mai 2002, you would define your index as follows:

<index NAME = "MyDateIndex"> 
   <dateindex> 
      <dateindexproperties> 
         <datevalidation VALUE="YearMonthDay"/> 
         <formats> 
            <format NAME="en_US"/> 
            <format NAME="fr_FR/> 
         </formats> 
      </dateindexproperties> 
   </dateindex> 
</index>

Each locale allows a set of date and time patterns. For example, Tuesday, June 11th, 2002 is one of the possible patterns for dates in American English.

When you specify a locale in date and time indexes, TEXTML Server parses the documents for all the patterns allowed. Once a pattern is recognized, the date or the time is indexed.

The complete list of patterns is available in the DATETIME_FORMATS.XML document, located in TEXTML Server’s Documentation folder. This system document is also available in the system document section of the document base’s repository.