Sql Server 2008 Full-Text Index (characters issue)

问题

The Full-Text Index searching is working perfect but suddenly I noticed that it fails when it comes to some characters variance in Arabic

In Arabic we have a trailing letter say {I} that can be written like { i } or { I } , It's the same letter but different ASCII code.. exactly like the English variance between { i } & { I }

the "Contains" function can get "ALi" but not "ALI"

• both {ALi & ALI} exist..
• {ALi} returns result but without the result of {ALI}
• {ALI} return 0 records when using full-text searching..

Any ideas??

Thanks

回答1:

did you implement a different index to arabic? You need one index to each different collation

来源：https://stackoverflow.com/questions/9763650/sql-server-2008-full-text-index-characters-issue