Friday, October 29, 2004 07:30 AM
My two bits on this..
This essentially depends on the rules defined for the QueryParser. You can use your own analyzer and define the rules separately. However if the application is searching for textual information, having these starting characters will not have much significance.
A different situtation is where you want to search for something like C++ or c# and the parser will strip off the + or # characters and end up searching for 'c'. In these situations the way to go will definitely change the grammer, the simplest of which would be to create these as additional special tokens. (This was suggested to me by Doug Cutting in reply to a question).