[Koha-bugs] [Bug 1677] nozebra search doesnt handle multi-index searches

[bugzilla-daemon at pippin.metavore.com]

http://bugs.koha.org/cgi-bin/bugzilla/show_bug.cgi?id=1677





------- Comment #4 from mason.james at liblime.com  2007-12-19 17:47 -------
> so it seems that the 2 searches  (title and author), does correctly get a
> matching result for bib '641,circleoflo'
> 
> but NZAnalyze() doesnt successfully return it, as the index differs slighty, so
> the match fails because the indexes have different end value (-1, -2)
> 
> 641,circleoflo-1
> 641,circleoflo-2
> 
> what does this end value mean?


ok so the end value is 'weight', and the weighting seems to be the cause of the
search breaking


if i hard-code my rebuild_nozebra.pl script to use a weight value of 1 for all
indexes, the search works correctly ( yay)

here is a log dump of the previous search returning the correct results, using
weight=1 for indexes

### $leftresult:
'641,circleoflo-1;960,childcarea-1;993,loveisnote-1;1209,lovelaught-1;1229,howtoloveb-1;1510,siblingriv-1;1727,youregoing-1;'
### $rightresult: '641,circleoflo-1;'
### $operand: ' and '

### @leftresult: [
###                '641,circleoflo-1',
###                '960,childcarea-1',
###                '993,loveisnote-1',
###                '1209,lovelaught-1',
###                '1229,howtoloveb-1',
###                '1510,siblingriv-1',
###                '1727,youregoing-1'
###              ]

### $finalresult: '641,circleoflo-1;641,circleoflo-1;'
### $result: '641,circleoflo-1;641,circleoflo-1;'


so the question, is the problem that the 'rebuild_nozebra.pl' script is making
incorrect weight values for the indexes?

or, is the problem that the indexes are correct, but NZAnalyze isnt matching
then correctly, (ie: does it need to strip the weight value from the index
before it compares for a matching result)

Paul, any ideas?

Mason.




------- You are receiving this mail because: -------
You are the QA contact for the bug, or are watching the QA contact.