Merge lp:~jamesh/mediascanner2/no-notify-on-unchanged-close into lp:mediascanner2
Proposed by
James Henstridge
Status: | Merged |
---|---|
Approved by: | Michi Henning |
Approved revision: | 323 |
Merged at revision: | 324 |
Proposed branch: | lp:~jamesh/mediascanner2/no-notify-on-unchanged-close |
Merge into: | lp:mediascanner2 |
Diff against target: |
285 lines (+215/-4) 5 files modified
CMakeLists.txt (+1/-0) src/daemon/SubtreeWatcher.cc (+5/-3) src/daemon/SubtreeWatcher.hh (+1/-1) test/CMakeLists.txt (+7/-0) test/test_subtreewatcher.cc (+201/-0) |
To merge this branch: | bzr merge lp:~jamesh/mediascanner2/no-notify-on-unchanged-close |
Related bugs: |
Reviewer | Review Type | Date Requested | Status |
---|---|---|---|
PS Jenkins bot (community) | continuous-integration | Needs Fixing | |
Michi Henning (community) | Approve | ||
Review via email:
|
Commit message
Don't emit the InvalidateResults signal if a file was opened for writing and then closed, but not actually modified
Description of the change
Don't emit the InvalidateResults signal if a file was opened for writing and then closed, but not actually modified.
This prevents Michi's thumbnailer taglib branch from triggering lots of invalidations, since TagLib opens files for writing for no good reason.
The sleeps in the new test are needed because the invalidation code tries to batch changes to reduce the number of invalidations. If I add more tests to this fixture, I'll definitely want to get rid of the sleeps, but for now it is probably okay.
To post a comment you must log in.
I installed this on the phone and verified that the it fixes the problem with the taglib branch. James, thanks heaps for helping with this in record time!