Review Request 126789: Don't propagate dataChanged upwards when we are not filtering.

[Volker Krause]

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/126789/
-----------------------------------------------------------

Review request for KDE Frameworks and Stephen Kelly.


Repository: kitemmodels


Description
-------

This avoids expensive layoutChange signals from QSFPM. We can only be sure
we are not filtering if filterAcceptRows isn't overridden, which we can
check if we assume sub-classes have the Q_OBJECT macro set.

This will however break sub-classes overriding filterAcceptRows without
using the Q_OBJECT macro, and using their own filter criteria. However,
this is a significant performance improvement in the common case of
KRFPM being used for string-based searching in a tree view, and the search
being unused.


Diffs
-----

  src/krecursivefilterproxymodel.cpp dbb6eba421c0e680fffe43582f210ea3e42e6e7f 

Diff: https://git.reviewboard.kde.org/r/126789/diff/


Testing
-------

Tested in GammaRay, where the layoutChanges are particularly painful as the source model is in a different process.


Thanks,

Volker Krause

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kde-frameworks-devel/attachments/20160118/ac761b7a/attachment-0001.html>