we have an application that uses context indexes. Normally we only had it in one language. Now there is a customer that uses 2 languages. So the application is now in both languages, also the documents inserted in the database could be both languages.
At installation time of the Oracle Text part, you also have to create the lexer preference, wordlists, stoplists ...
Can I, or should I load that in database for both languages? Is this possible and necessary?
It depends on the type of operators you are using or planning to use in your queries. For example, you can use STEM ($) for some languages included in the BASIC_LEXER, for other languages, you will need another LEXER...
We are currently working in both French and English indexed texts and only interested in NOT (~), AND (&) and OR (|) searches. As far as making both languages work well together is concerned, we just created a custom list of stopwords that include english and french stopwords. I would say that it is the minimum needed, since otherwise you would index lots of useless words, polluting searches as well as increasing index sizes. For now we haven't seen the need to change another parameter.
HTH & Regards,
ORA-000TK : No bind variable detected... Shared Pool Alert code 5 - Nuclear query ready .