[TurboGears] Re: Search options: Are there any, should we be considering them?

2006-06-05 Thread Robin Haswell
FWIW the only search solution I managed to get flying after trying all of the proposals in this thread was a PyLucene based search servce running on an XML-RPC server (using SimpleXMLRPCServer). This is because PyLucene doesn't play friendly with threads unless you use this: from PyLucene import

[TurboGears] Re: Search options: Are there any, should we be considering them?

2006-06-04 Thread Ben Lee
2 pieces of useful data: 1) http://wiki.apache.org/jakarta-lucene/LucyProposal Doug Cutting, Dave Balmain, Marvin Humphrey have got ASF approval for a C backend to lucene for all dynamic languages (Perl, Ruby, Java, et al.) I expect a Python port to soon follow. 2) generalized inverted

[TurboGears] Re: Search options: Are there any, should we be considering them?

2006-05-13 Thread Kevin Dangoor
On 5/12/06, Michele Cella [EMAIL PROTECTED] wrote: Some links... just in the case you haven't catched them: http://swapoff.org/wiki/pyndexter Cool! I hadn't seen that. At least someone has managed to get Hype to build. If we can make eggs of that, we may be all set. I do like the idea of a

[TurboGears] Re: Search options: Are there any, should we be considering them?

2006-05-12 Thread Kevin Dangoor
On 5/12/06, Robin Haswell [EMAIL PROTECTED] wrote: So I was wondering, should we be thinking about a portable way of providing search for TG apps? Or at least, I think we should provide an integration method for something which is quite easily obtainable, if not shipped with TG itself. Yes,

[TurboGears] Re: Search options: Are there any, should we be considering them?

2006-05-12 Thread anders pearson
On 2006-05-12 13:00:22 +0100, Robin Haswell wrote: Let's face it, every application on the web these days need some form of search, and as far as I can see, TG + SO/any ORM is going to make that tricky. Agreed. This won't work for everyone, but I've been using PostgreSQL's tsearch2 engine

[TurboGears] Re: Search options: Are there any, should we be considering them?

2006-05-12 Thread Robin Haswell
In the past I've also used Lupy (combining the slowness of a pure-python implementation with the awkwardness of a Java style API. With bonus bugginess!) and MySQL's full-text search (if it doesn't have stemming, I have a hard time considering it real full-text search). Our MySQL 4.1

[TurboGears] Re: Search options: Are there any, should we be considering them?

2006-05-12 Thread Robin Haswell
Well I've wasted a midday and I can tell you PyLucene is a big no-no. A google for cherrypy pylucene brings too many horror stories to mention. Basically something in PyLucene causes CP to bomb out - on my dev setup, insantiating (sp) PyLucene.IndexSearcher() causes an immediat autoreload.

[TurboGears] Re: Search options: Are there any, should we be considering them?

2006-05-12 Thread Robin Haswell
Here's a third choice that doesn't appear to suffer from the problems of the other two, but I haven't used it: http://hype.python-hosting.com/ This doesn't build on my Debian Sarge system, bombs out with a missing file error (_hype.c!). Plus it depends on hyperes..something, which also

[TurboGears] Re: Search options: Are there any, should we be considering them?

2006-05-12 Thread Kevin Dangoor
On 5/12/06, Robin Haswell [EMAIL PROTECTED] wrote: GPL GPL GPL GPL (fine for some things, but not for TG core) Yeah I read about this - what's the problem with using GPL software? I'm no license expert, AFAIK it doesn't affect our (ie. my company's) business as we don't ship code. Like I

[TurboGears] Re: Search options: Are there any, should we be considering them?

2006-05-12 Thread Michele Cella
Some links... just in the case you haven't catched them: http://swapoff.org/wiki/pyndexter http://blog.case.edu/bmb12/2006/03/merquery_text_indexing_and_search_with_a_focus http://blog.case.edu/bmb12/2006/03/more_on_merquery Ciao Michele Robin Haswell wrote: Hey guys Let's face it, every

[TurboGears] Re: Search options: Are there any, should we be considering them?

2006-05-12 Thread Ronald Jaramillo
Hi Robin, I have been trought pretty much the same evaluation trip (and share your findings). I ended writting a basic database- base search dingie that works fine for small documents sets ( 1k documents ). For bigger sets, I'll rather install Tomcat + Lucene and expose Lucene trough