Closed (won't fix)
Project:
Apache Solr Search
Version:
6.x-1.x-dev
Component:
Code
Priority:
Normal
Category:
Bug report
Assigned:
Unassigned
Reporter:
Created:
17 Dec 2008 at 12:07 UTC
Updated:
23 Jun 2009 at 04:09 UTC
Jump to comment: Most recent file
Comments
Comment #1
vladimir.dolgopolov commentedHere is the patch and tests fixing it.
Comment #2
gábor hojtsyUnfortunately *.test files are not viewable on drupal.org right now. Logged a bug report on that here: http://drupal.org/node/348164
Until that is fixed, it would be a good idea if you could repost these in a patch form or as .txt files.
Comment #3
pwolanin commentedYes, sorry about that - obviosuly a total bug here:
$fq = $this->fq[] = implode(" {$operator} ", $subfq);I think the patch is not quite right, however - we need to implode with the operator.
Comment #4
pwolanin commentedcommitted the patch in #3
Comment #5
pwolanin commentedSeems like we need some further fixes here as well - per testing and discussion w/ Jacob.
Comment #6
vladimir.dolgopolov commentedHere is the tests from #1 renamed as txt.
Comment #7
vladimir.dolgopolov commented@Peter #3
I think the code:
should produces (result of the patch #1):
instead of (result of the patch #3):
I think so because the code (notice I enter similar $keys instead of $filters param in apachesolr_drupal_query()):
produces:
P.S.
I'm thinking we should parenthesize the parent query as well:
Comment #8
JacobSingh commentedHey Guys,
The nodeaccess patch above looks good, and it looks like Peter already committed it.
I made one change to use parenthesis as Vlad suggests. Attached is a patch in progress of the subquery tests being handled a little better now. Since filter queries are always AND some of the previous tests no longer make sense AFAICT.
Comment #9
pwolanin commented@Vladamir - I was not actually thinking about subqueries being in the URL, rather as being added later - I'm not sure that the parsing code correctly handles them from the URL.
Comment #10
pwolanin commentedJacob - did you commit this patch? It looks like it from the CVS log.
Comment #11
pwolanin commentedpatch to fix nodeaccess handling for mlt block queries.
Comment #12
JacobSingh commentedThe handler can be determined form $params if needed, otherwise, committed the above. (patch attached).
Comment #13
vladimir.dolgopolov commentedThere are fixed tests.
Notes:
1. solr_base_query.test
It passes if you apply the patch Pure negative queries don't work
2. solr_base_subquery.test
It includes (commented) testStaticFactory() method intended to test apachesolr_current_query() static cache. But there is a problem.
The original test assumes option (b). Here I see option (a).
We shoud decide which approach to follow.
3. solr_index_and_search.test
The original one is based on search.module test so it cannot be used for testing dismax functionality.
It uses a query type 'standard'.
The dismax test will be created soon.
Comment #14
pwolanin commentedComment #15
JacobSingh commentedThis is pretty ancient.