Re: [sword-devel] FFe application for sword being considered, and patches split out

2009-08-31 Thread Jonathan Marsden
Matthew Talbert wrote: > However, I recommend that you switch to using the SimpleAnalyzer > rather than StandardAnalyzer. All that's needed it to replace > occurrences of standard::StandardAnalyzer to SimpleAnalyzer (no > namespace needed), and remove all references to stop_words. Based on > my ex

Re: [sword-devel] FFe application for sword being considered, and patches split out

2009-08-31 Thread Dmitrijs Ledkovs
heya all I've just read motu release meeting minutes and from that they point out that a clear bug-fix release from upstream needs a bug in Launchpad and sponsor (not the big and scary ffe). So if clear bugs with clear fixes will be crafted into 1.6.0.1 tarball might work instead of carrying in m

Re: [sword-devel] FFe application for sword being considered, and patches split out

2009-08-31 Thread Jonathan Marsden
DM Smith wrote: For a fix that doesn't break backward compatibility. How about getting the stopword list from the analyzer and checking the query for those words. If any are found then gracefully fail the query, or merely delete them and potentially create a bad query, or ... That prevents the

Re: [sword-devel] FFe application for sword being considered, and patches split out

2009-08-31 Thread Jonathan Marsden
Matthew Talbert wrote: I may package 0.9.23 for Windows, but the linux distros aren't going to move to 0.9.23 until it's regarded as stable. Indeed. As confirmation: rmadison -s karmic libclucene-dev outputs libclucene-dev | 0.9.20-3 |karmic | amd64, i386 So that is what will

Re: [sword-devel] FFe application for sword being considered, and patches split out

2009-08-31 Thread Greg Hellings
On Mon, Aug 31, 2009 at 12:55 PM, DM Smith wrote: > On 08/31/2009 12:55 PM, Matthew Talbert wrote: >> >> On Mon, Aug 31, 2009 at 12:52 PM, DM Smith  wrote: >> >>> >>> On 08/31/2009 12:42 PM, DM Smith wrote: >>> I don't think the solution regarding segfaults on a search of only stopwo

Re: [sword-devel] FFe application for sword being considered, and patches split out

2009-08-31 Thread DM Smith
On 08/31/2009 12:55 PM, Matthew Talbert wrote: On Mon, Aug 31, 2009 at 12:52 PM, DM Smith wrote: On 08/31/2009 12:42 PM, DM Smith wrote: I don't think the solution regarding segfaults on a search of only stopwords is the right way to solve the problem. The problem is in clucene or an

Re: [sword-devel] FFe application for sword being considered, and patches split out

2009-08-31 Thread Matthew Talbert
> 0.9.23 is "unstable" and is over 13 months old.  0.9.21b is the latest > stable, and even that dates back to October of last year.  Judging > from the last two commits on their project page, they're in a > refactoring stage - no idea how long that is expected to last.  Since > 0.9.21b post-dates

Re: [sword-devel] FFe application for sword being considered, and patches split out

2009-08-31 Thread Greg Hellings
On Mon, Aug 31, 2009 at 11:55 AM, Matthew Talbert wrote: > On Mon, Aug 31, 2009 at 12:52 PM, DM Smith wrote: >> On 08/31/2009 12:42 PM, DM Smith wrote: >>> >>> I don't think the solution regarding segfaults on a search of only >>> stopwords is the right way to solve the problem. The problem is in c

Re: [sword-devel] FFe application for sword being considered, and patches split out

2009-08-31 Thread Matthew Talbert
On Mon, Aug 31, 2009 at 12:52 PM, DM Smith wrote: > On 08/31/2009 12:42 PM, DM Smith wrote: >> >> I don't think the solution regarding segfaults on a search of only >> stopwords is the right way to solve the problem. The problem is in clucene >> or an inappropriate use of it. A bug should be entere

Re: [sword-devel] FFe application for sword being considered, and patches split out

2009-08-31 Thread DM Smith
On 08/31/2009 12:42 PM, DM Smith wrote: I don't think the solution regarding segfaults on a search of only stopwords is the right way to solve the problem. The problem is in clucene or an inappropriate use of it. A bug should be entered into the clucene project for it. The bug has already bee

Re: [sword-devel] FFe application for sword being considered, and patches split out

2009-08-31 Thread DM Smith
I don't think the solution regarding segfaults on a search of only stopwords is the right way to solve the problem. The problem is in clucene or an inappropriate use of it. A bug should be entered into the clucene project for it. I found a related, and now fixed bug, where "a AND b" where "a"

Re: [sword-devel] FFe application for sword being considered, and patches split out

2009-08-31 Thread Matthew Talbert
> Since clucene isn't aiming for either UTF-16 OR UTF-32, I don't > believe you'll be able to. A better approach would be to get the size > of "content" and set a value based on that. FWIW, I just did this for both searching and index creating. I over-allocated the length by 500, which is overkill

Re: [sword-devel] FFe application for sword being considered, and patches split out

2009-08-31 Thread Matthew Talbert
> I do appreciate and plan to incorporate Matthew's suggestions into svn > head.  I believe I still may research clucene and see if we can use our > UTF-16 or 32 conversion routines so we won't need to allocate 1 meg of > static buffer. Since clucene isn't aiming for either UTF-16 OR UTF-32, I don

Re: [sword-devel] FFe application for sword being considered, and patches split out

2009-08-31 Thread Matthew Talbert
> > Troy A. Griffitts wrote: > > > Quick note: I remember scanning the patch originally sent seeing a > > 'new' without an obvious delete.  Please be sure to check for this. > > I may be wrong; it was a quick observation. > > You could well be right, the stopwords patch > >  http://crosswire.org/~j

Re: [sword-devel] FFe application for sword being considered, and patches split out

2009-08-30 Thread Jonathan Marsden
Troy A. Griffitts wrote: > Quick note: I remember scanning the patch originally sent seeing a > 'new' without an obvious delete. Please be sure to check for this. > I may be wrong; it was a quick observation. You could well be right, the stopwords patch http://crosswire.org/~jmarsden/17_no_st

Re: [sword-devel] FFe application for sword being considered, and patches split out

2009-08-30 Thread Troy A. Griffitts
Quick note: I remember scanning the patch originally sent seeing a 'new' without an obvious delete. Please be sure to check for this. I may be wrong; it was a quick observation. I do appreciate and plan to incorporate Matthew's suggestions into svn head. I believe I still may research clucene a