MoinMoin: MoinMoinChat/Logs/moin-dev/2006-02-11

2006-02-11T00:05:01  <xorAxAx> [PERIODIC ANNOUNCEMENT] Logs can be found on http://moinmoin.wikiwikiweb.de/MoinMoinChat/Logs/moin-dev
2006-02-11T16:53:42  <ThomasWaldmann> moin Fabi
2006-02-11T16:58:08  <ThomasWaldmann> Fabi: is it possible to feed external data into the lupy index so it shows up in search results, but doesnt need to be in pagelist?
2006-02-11T17:31:10  <Fabi> not yet
2006-02-11T17:31:21  <Fabi> but should be easy
2006-02-11T17:31:27  <Fabi> if we can get lupy fixed
2006-02-11T17:34:31  <ThomasWaldmann> i want to add filesystem indexing data to the search index
2006-02-11T17:35:01  <ThomasWaldmann> and access the filesystem via virtual pages in moin
2006-02-11T17:36:17  <Fabi> this should work
2006-02-11T17:36:23  <Fabi> you just have to index the pages
2006-02-11T17:36:53  <ThomasWaldmann> due to the amount of files in that filesystem, i dont want to add them to pagelist
2006-02-11T17:37:29  <Fabi> yes, just modify the indexer
2006-02-11T17:37:38  <ThomasWaldmann> and due to security, i dont want to show context in search results, just the filename (virtual page name)
2006-02-11T17:37:40  <Fabi> to use the full list
2006-02-11T17:37:56  <Fabi> hmm
2006-02-11T17:38:56  <Fabi> then you'll have to modify the search result code, too
2006-02-11T17:39:48  <ThomasWaldmann> is the LupyIntegration page's bug section up-to-date and complete as far as we know?
2006-02-11T17:40:33  <xorAxAx> there might be separate bug reports in the tracker
2006-02-11T17:40:52  <ThomasWaldmann> SF?
2006-02-11T17:41:32  <ThomasWaldmann> in the wiki there is only that optimize breaks index bug
2006-02-11T17:42:37  * ThomasWaldmann switches on lupy search for moinmoin
2006-02-11T17:44:46  <xorAxAx> no, our tracker :)
2006-02-11T17:44:47  <ThomasWaldmann> is it ok that the build index script also runs optimize index afterwards or will that trigger that bug?
2006-02-11T17:45:26  <ThomasWaldmann> i have 4.7mb index for moinmoin
2006-02-11T17:45:47  <xorAxAx> "On this wiki, MoinMoinBugs did not show any results but no traceback either. At least a traceback/error should come up if locks do not work etc." was the worst bug
2006-02-11T17:45:53  <xorAxAx> and its not lupy related
2006-02-11T17:45:57  <xorAxAx> but a bug in the lupy interface
2006-02-11T17:46:10  <xorAxAx> some locking call without else branch i guess
2006-02-11T17:48:06  <ThomasWaldmann> hey, no bugs ;)
2006-02-11T17:48:28  <ThomasWaldmann> lupy fixed all our bugs
2006-02-11T17:50:00  <ThomasWaldmann> the / after MoinMoinBugs confuses it
2006-02-11T17:50:15  <starshine> rofl
2006-02-11T17:50:59  <ThomasWaldmann> but even without /, it only finds 26 pages
2006-02-11T17:54:40  * ThomasWaldmann rebuilds lupy index without optimize call
2006-02-11T17:56:20  <ThomasWaldmann> even while building the index, I already get 600+ bugs pages (title search)
2006-02-11T17:57:55  <ThomasWaldmann> looks like that optimize calls optimizes it to death
2006-02-11T18:02:46  <ThomasWaldmann>         TODO: needs special locking, so the index is readable until the
2006-02-11T18:02:46  <ThomasWaldmann>         optimization is finished.
2006-02-11T18:03:54  <ThomasWaldmann> do we have readable, correct index data while optimize is running?
2006-02-11T18:07:33  <ThomasWaldmann> Fabi: lupy.py:309 except:
2006-02-11T18:07:42  <ThomasWaldmann> shouldn't that be "finally:" ?
2006-02-11T18:09:00  <ThomasWaldmann> ehrm, no, forget that one
2006-02-11T18:13:17  <Fabi> IIRC the db should be usable during optimization
2006-02-11T18:13:32  <Fabi> never the less optimization is broken
2006-02-11T18:14:52  <ThomasWaldmann> Fabi: it does fall back to standard search while index building, right?
2006-02-11T18:20:00  <ThomasWaldmann>             try:
2006-02-11T18:20:00  <ThomasWaldmann>                 return self._decode(self._read())
2006-02-11T18:20:00  <ThomasWaldmann>             finally:
2006-02-11T18:20:00  <ThomasWaldmann>                 self.readLock.release()
2006-02-11T18:20:07  <ThomasWaldmann> does such a thing work?
2006-02-11T18:20:24  <xorAxAx> ThomasWaldmann: of course
2006-02-11T18:20:56  <ThomasWaldmann> so it releases the lock aftern returning?
2006-02-11T18:22:05  <xorAxAx> before returning
2006-02-11T18:23:36  <ThomasWaldmann> title search for "Bugs" : 100 results (all showing MoinMoinBugs/...)
2006-02-11T18:23:49  <ThomasWaldmann> title search for "MoinMoinBugs" : 26 results
2006-02-11T18:24:09  <xorAxAx> 100 is a dirty hard limit
2006-02-11T18:24:29  <xorAxAx> python basics:
2006-02-11T18:24:31  <xorAxAx> >>> def foo():
2006-02-11T18:24:32  <xorAxAx> ...  try:
2006-02-11T18:24:32  <xorAxAx> ...   return (lambda:sys.stdout.write("1"))() or 2
2006-02-11T18:24:32  <xorAxAx> ...  finally:
2006-02-11T18:24:32  <xorAxAx> ...   print "3"
2006-02-11T18:24:34  <xorAxAx> ...
2006-02-11T18:24:35  <ThomasWaldmann> that doesnt explain the 26
2006-02-11T18:24:36  <xorAxAx> >>> foo()
2006-02-11T18:24:39  <xorAxAx> 13
2006-02-11T18:24:42  <xorAxAx> 2
2006-02-11T18:24:44  <xorAxAx> indeed :)
2006-02-11T18:26:28  <ThomasWaldmann> ok
2006-02-11T18:33:22  <ThomasWaldmann> those lupy title search results are irrational
2006-02-11T19:01:06  <ThomasWaldmann> Fabi: did you try putting all the lupy stuff under a single dir?
2006-02-11T19:01:30  <ThomasWaldmann> like lupy_index -> lupy
2006-02-11T19:01:43  <ThomasWaldmann> lupy_index_lock -> lupy/lock
2006-02-11T19:02:19  <ThomasWaldmann> lupy_queue_lock -> lupy/queuelock
2006-02-11T19:02:37  * xorAxAx wonders if he has tried anything about lupy at all
2006-02-11T19:19:10  <ThomasWaldmann> argh
2006-02-11T19:19:22  <ThomasWaldmann> MoinMoin -> 4 results (all NOT CamelCase)
2006-02-11T19:19:31  <ThomasWaldmann> Moin Moin -> 100 results
2006-02-11T19:20:14  <ThomasWaldmann> because of the special treatment of CamelCase stuff
2006-02-11T19:21:31  * ThomasWaldmann removed that and builds index again
2006-02-11T22:07:24  * ThomasWaldmann cleans up search and lupy