In the Lucene build that we've got (2/21) the question mark does not do a single-character replace. Does anyone know why? We're using the StandardAnalyzer and the default QueryParser.
-----Original Message----- From: Peter Carlson [mailto:[EMAIL PROTECTED]] Sent: Saturday, February 23, 2002 5:23 PM To: Lucene Users List Subject: Re: Googlifying lucene querys Hi Jari, Lucene is designed as an API with different components broken out so a developer can create the uniqueness required. One part of Lucene is the QueryParser. The QueryParser takes a search string and create a set of classes based on the current QueryParser.jj implementation and turns it into a Lucene Query. This is meant to be a good solution for most people, but it is just a sample of what can be done. In the current implementation of QueryParser 'george bush "white house"' Will create an OR query of George OR bush OR "white house" Basically, the default is an OR between words unless otherwise specified. You can use other boolean operators like AND, and NOT So 'george AND bush OR "white house" NOT ford' Lucene and the current QueryParser supports wildcards with the * character Single character replace with the ? Character Fuzzy searches with the ~ character when next to a single word term Proximity searches (just added to QueryParser) with the ~3 next to a phrase term Again, you can create your own QueryParser to create your desired implementation. I hope this helps. --Peter On 2/23/02 8:19 AM, "Jari Aarniala" <[EMAIL PROTECTED]> wrote: >> +george +bush +white +house > > Well, that's pretty obvious even for me :) If you have separate words, > just tokenize the string and add a plus in front of each of the words. > But what I'm trying to do here is this: > > Let's say I have a more complicated query, say > > 'george bush "white house"' > > There you have two separate words, "george" and "bush" and then > "white house" enclosed in quotes. If I use a piece of simple > tokenization code, the above query becomes > > +georbe +bush +"white +house" > > See what I mean? That won't work the way expected. > Anyway, I'm still a bit confused the inner workings of Lucene, > so maybe I'll come up with something myself. > > Jari Aarniala > [EMAIL PROTECTED] > > > > -- > To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> > For additional commands, e-mail: <mailto:[EMAIL PROTECTED]> > > -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]> -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
