* Drupal 6.2, search has been enabled, permissions given, and re-indexed. Site http://pssnet.com/~devone/drupaltest/
* The search picks up certain strings (இன்றைய காலகட்டத்தில்) while not others (மேற்குறித்த கட்டுரைக்கு).
* Database is utf8_unicode encoded, the web pages are displaying correctly.
* There seems to be several known issues, but no one has posted a clear solution for this issue (http://drupal.org/node/604002). Tried to apply some of the suggested patches, and they failed.
* Tamil Words are broken into three letter meaningless phases. It may have worked if the breakup was for full characters. But the collation screws up.

Comments

jhodgdon’s picture

Category: support » bug
Status: Active » Closed (duplicate)

If Tamil is being broken up into meaningless 3-character phrases, then maybe you need to turn off the CJK tokenizing?

Meanwhile, this appears to be a duplicate of #604002: Poor search support of some Unicode scripts, which hasn't been fixed yet, sorry.

Natkeeran’s picture

Thank you for the reply. Changing Minimum word length to index to 1 helps with the search. This is for 6.2x.

Unicode is not even rendering it properly in D7. The db is encoded in utf.