> On Dec. 6, 2015, 11:47 a.m., Vishesh Handa wrote: > > src/engine/mtimedb.cpp, line 98 > > <https://git.reviewboard.kde.org/r/126227/diff/2/?file=421133#file421133line98> > > > > Could you please sort these values as well? This is possible why we are > > getting duplicates. The WriteTransaction assumes the data is sorted when it > > gets it.
MtimeDB uses mtime as key, it's fairly easy to make mtimedb contain multiple mtime for the same docId. I don't see WriteTransaction uses MtimeDB::get. If you want to reproduce the problem, just find a file on the day A other than today and confirm timeline:// can show it. Then touch the file, and then you can find that file in both timeline:/ today and day A. - Xuetian ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/126227/#review89170 ----------------------------------------------------------- On Dec. 6, 2015, 5:29 a.m., Xuetian Weng wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://git.reviewboard.kde.org/r/126227/ > ----------------------------------------------------------- > > (Updated Dec. 6, 2015, 5:29 a.m.) > > > Review request for Baloo and Vishesh Handa. > > > Repository: baloo > > > Description > ------- > > Test with > - baloosearch 'modified=>...' -d $HOME > - baloosearch 'modified=>...' > > This two queries should return same result since I only index home folder. > But they are not, because result from mtimedb is not sorted. > > In some other rare case (On my machine, regularly happens to torrent > downloaded files), it could also return duplicate result. not sure if this > should be considered as a indexer bug. > > Sort and remove duplication from the result in VectorPostingIterator when > requesting first result with next(). > > > Diffs > ----- > > src/engine/mtimedb.cpp f7283b5 > > Diff: https://git.reviewboard.kde.org/r/126227/diff/ > > > Testing > ------- > > No more duplicate result or missing result. > > > Thanks, > > Xuetian Weng > >
>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<
