> On Jan. 12, 2014, 4:17 p.m., Vishesh Handa wrote: > > Seems pretty cool. You rock!
Cheers! > On Jan. 12, 2014, 4:17 p.m., Vishesh Handa wrote: > > src/file/indexingqueue.cpp, line 55 > > <https://git.reviewboard.kde.org/r/114981/diff/1/?file=233863#file233863line55> > > > > I seem to have a vague memory of why this was not done, but now I > > cannot remember it, so go ahead. > > > > At some point we really need unit tests for the queues and the > > scheduler. > > > > PS: Could you please also fix the coding style? Space after if > > Maybe we discover why it was done like this the hard way then! Also, yes, you're right, we totally need unit tests. I'll fix the coding style in the uploaded version. - Simeon ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/114981/#review47257 ----------------------------------------------------------- On Jan. 12, 2014, 3:43 p.m., Simeon Bird wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://git.reviewboard.kde.org/r/114981/ > ----------------------------------------------------------- > > (Updated Jan. 12, 2014, 3:43 p.m.) > > > Review request for Baloo and Vishesh Handa. > > > Repository: baloo > > > Description > ------- > > Rework slotScheduleIndexing so that it is clearer when each queue runs, > and fix a few corner case bugs where the queues would run or not run in the > wrong places. > > Then fix a few strange things that turn up when I do this; the IndexCleaner > stuff is a bit pointless, > because the IndexCleaner does nothing right now. > > commit be744870904645be78f50bd51b186805e8c902ed > Author: Simeon Bird <[email protected]> > Date: Sat Jan 11 16:25:51 2014 -0500 > > Fix up the IndexCleaner so that it starts suspended, then set resume to > just emitResult and finish instantly. This stops the cleaner hanging > forever, as it did previously. > > Also we don't need m_suspended as that is handled by the parent KJob > class. > > commit e8aa2268601ee40296bf120385e0506bfc6241b9 > Author: Simeon Bird <[email protected]> > Date: Sat Jan 11 18:02:38 2014 -0500 > > Make it safe to call resume on an empty queue by not doing > callForNextIteration when the queue is empty > > commit e8d46a4517fe21a5382fdb21f61eed90f15b53af > Author: Simeon Bird <[email protected]> > Date: Sat Jan 11 17:59:23 2014 -0500 > > Don't call resume on an empty queue. > > This leads to a call loop: > resume->callForNextIteration->finishedIndexing > ->slotFinishedIndexing-> slotScheduleIndexing-> resume > > commit 96622c959f5dfd84df772a151a63df7d02ba9992 > Author: Simeon Bird <[email protected]> > Date: Sat Jan 11 14:16:45 2014 -0500 > > Refactor slotScheduleIndexing to make it clearer when each queue runs. > > Along the way, fix a couple of odd edge cases: > 1. The file queue would sometimes not be started correctly > 2. The cleaner would stop when going onto battery, but if we > were already on battery when starting, it would still be started. > Change this so the cleaner runs even when on battery. > > Also make some adjustments to the index cleaner so that it works when > resumed: these are only necessary because it is not yet implemented. > > > Diffs > ----- > > src/file/indexcleaner.h 7d68ebf > src/file/indexcleaner.cpp 685a675 > src/file/indexingqueue.cpp ca2abc3 > src/file/indexscheduler.h 015d23a > src/file/indexscheduler.cpp 79cd5bf > > Diff: https://git.reviewboard.kde.org/r/114981/diff/ > > > Testing > ------- > > Compiled, ran with it a bit. Some of the edge cases are fixed now. > > > Thanks, > > Simeon Bird > >
>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<
