[Lucene-dev] Parsing query with mixed tokenized fields

[Tal Dayan]

This is for the FAQ,

Is it possible to use the QueryParser for queries that include both
tokenized
and non tokenized fields ?

Consider for example the query string:

  +body:dogs +site:http://www.yahoo.com

(Assume that first field is tokenzied and the second is not)

Since the query parser passes the
terms through the analyzer (for example, to normalize 'dogs' to 'dog' using
the Porter filter),
what will happen to the term of the second (non tokenized) field ?

If this is a problem indeed, how about extending the parser to support
literal terms that
are not tokenized such as

  +body:dogs +site:'http://www.yahoo.com'

Tal



_______________________________________________
Lucene-dev mailing list
[EMAIL PROTECTED]
http://lists.sourceforge.net/lists/listinfo/lucene-dev