> On Jan. 12, 2014, 4:11 p.m., Vishesh Handa wrote: > > This is awesome. Ship it! > > If you want you can also get rid of the IndexCleaner completely. I think ti > > might be better to move it into the KCM, and only run it when the > > configuration changes. There isn't much point running it each time. > > Simeon Bird wrote: > Thanks! > > Do you mean remove IndexCleaner from the IndexScheduler, and have the KCM > do something like this, for a cleaner which is implemented, and can clean > specific paths: > > if some folders were removed from the path, or the exclude filters > changed, or a total re-index was requested: > suspend indexing > call cleaner (maybe through dbus to baloo_file?) > connect cleaner done signal to resume indexing > > ? > > If so, that makes a lot of sense to me...we can get rid of restartCleaner > as well.
Yup. We can either do it in baloo_file or even create a separate process which the KCM just spawns when the configuration changes. > On Jan. 12, 2014, 4:11 p.m., Vishesh Handa wrote: > > src/file/indexscheduler.cpp, line 185 > > <https://git.reviewboard.kde.org/r/114983/diff/1/?file=233869#file233869line185> > > > > I'm a little bit concerned about this because one might decide that the > > BasicIQ should not check every file on startup. > > > > Look at the fileindexer.cpp, there is a check over there to see if > > initial check is enabled / disabled. > > Simeon Bird wrote: > hmm, good point. Do you think the logic should be like this? > > m_fileIQ->fillQueue(); > > if (m_basicIQ->isEmpty() && m_fileIQ->isEmpty()) { > m_eventMonitor->suspendDiskSpaceMonitor(); > setIndexingStarted(false); > } > else { > slotScheduleIndexing(); > } > > It shouldn't do any harm to try to fill the queue, and it does seem to > make it a bit more robust. > I'm not sure. Because the slotStartedIndexing will never be called unless the basic/fileIQ has some data in it. So this code would never be run. - Vishesh ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/114983/#review47256 ----------------------------------------------------------- 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/114983/ > ----------------------------------------------------------- > > (Updated Jan. 12, 2014, 3:43 p.m.) > > > Review request for Baloo and Vishesh Handa. > > > Repository: baloo > > > Description > ------- > > Misc. File queue changes - the top edit should perhaps be applied to > IndexScheduler::slotCommitted > as well. > > commit ea35771246890c46f63d30f765a0d19844bd3d63 > Author: Simeon Bird <bla...@gmail.com> > Date: Sat Jan 11 14:57:39 2014 -0500 > > Instead of unconditionally resuming the file queue when the basic queue > is done, call slotScheduleIndexing. > > Otherwise the file queue would start once the basic queue was done, even > on battery. > > commit b6aeea48286e53206726af48dc65da026c3dd7a8 > Author: Simeon Bird <bla...@gmail.com> > Date: Sat Jan 11 14:34:17 2014 -0500 > > Refactor a function in indexscheduler.cpp to reduce duplication > > commit 4aa06719a98f740fb55c51232ac306409a860077 > Author: Simeon Bird <bla...@gmail.com> > Date: Sat Jan 11 14:21:18 2014 -0500 > > Do not call fillQueue when the file queue is created, but when the basic > queue is done. > > We only know what work needs to be done in the file queue once the basic > queue is complete, so call fillQueue at that time, not when the queue is > created. > > Before we could be filling the file queue pointlessly with > tasks to be cleaned, or, if the file queue was empty when > the basic queue started, it might not start when it should have done. > > commit 406a6a4ae4f5facbb33b7de433347d84f77182c5 > Author: Simeon Bird <bla...@gmail.com> > Date: Sat Jan 11 12:44:52 2014 -0500 > > FileIndexingQueue doesn't need to listen to ConfigChanged to refill the > indexing queue, because that is handled in the IndexScheduler already > > > Diffs > ----- > > src/file/fileindexingqueue.h c112c97 > src/file/fileindexingqueue.cpp 0f22caf > src/file/indexscheduler.h 015d23a > src/file/indexscheduler.cpp 79cd5bf > > Diff: https://git.reviewboard.kde.org/r/114983/diff/ > > > Testing > ------- > > Compiled, ran with it. > > > Thanks, > > Simeon Bird > >
>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<
