Closed (fixed)
Project:
Apache Solr Search
Version:
6.x-3.x-dev
Component:
Code
Priority:
Major
Category:
Bug report
Assigned:
Unassigned
Reporter:
Created:
20 Sep 2012 at 23:50 UTC
Updated:
5 Oct 2012 at 15:01 UTC
Jump to comment: Most recent file
Comments
Comment #1
pwolanin commentedIt seems it also doesn't update them on save, nor when a node type is changed, so a bit incomplete in general.
Also - we we're clearing the environment cache when a node type changed.
Comment #3
pwolanin commentedoops - left the cache clearing out of the end of the save function.
Comment #4
pwolanin commentedWith some added doxygen and D6 backport.
Comment #5
nick_vhI'm not sure what you mean with not updating them on save? I just tried it and looks like it works on save?
Could you be a bit more explanatory?
Comment #6
nick_vhRevised version with some clutter removed and tests added to check the bundle array in the environment and the cloned environment.
Comment #7
nick_vhI'm not sure yet why this is failing if I don't clear the cache here
Comment #8
pwolanin commentedYou have to remember the 2 different threads, remote vs local. This change is wrong:
Comment #9
pwolanin commentedComment #10
nick_vhIf apachesolr_environment_save() calls apachesolr_environment_clear_cache(), I don't see why I should do it again in the test (and the tests confirm that).
Please elaborate?
Comment #11
pwolanin commentedIf you are doing an edit via http POST, then the local thread's static variable cache is not cleared. I wasn't looking at the full context of the test, but if you do a load inside the test code you'd still have the old values.
I don't know why the test failed to run - here's the patch again w/ just a minor comment change.
Comment #13
nick_vhI must have uploaded an old version before. Anyway, corrected one here.
I did not inherit the code comment because you added a lot of spaces (for the sake of the test I think?)
Comment #14
nick_vhComment #16
nick_vhdoh, juggling too much at once - changed the wrong line.
Comment #17
pwolanin commentedI think this approach is better.
Comment #18
nick_vhYou still have those spaces :)
Comment #19
pwolanin commentedlots lof tabs. fixed.
Comment #20
nick_vhgood to go
Comment #21
pwolanin commentedWith some more test cleanup and making the naming of the new function consistent with 6.x
Comment #22
pwolanin commentedcommitting #21, here's a 6.x version.
Comment #23
nick_vhLooks good, go ahead and commit
Comment #24
pwolanin commented