This is so nice!
- Queries
- All Stories
- Search
- Advanced Search
Advanced Search
Oct 16 2019
Oct 15 2019
Oct 14 2019
Move comment to header
Oct 13 2019
- Use changedEvent() instead of overriding
Bad arcanist foo, sorry
@elvisangelaccio, about your question, reviewing the code, there are four places (apart from setter and getter) where m_searchPath is used:
- For baloo search (l.502), it is only used if m_everywhere is not checked, so it would cancel the condition in the getter, returning m_searchPath anyway
- For regular search (l.135), there is a similar condition as in searchPath().
- Menu action to call an external tool (KFind) (l.423)
- In the entry point fromUrlSearch(), introduced in this patch to reset it on certain paths
- Rebase to master
- Fix non-indexed corner case
- Make use of searchPath()
Almost there!
I still don't understand if we are fixing an issue or if this is just a removal of a redundant signal. AFAICS we are only removing a selectionChanged() instance called by clearSelection() in "search mode", but this signal is also emitted in many other places (e.g. is emitted 3 times if I change selection with the mouse).
In D24602#546343, @ndavis wrote:Why do the comboboxes look so thin in the video?
- Fix combos' height
In D24602#546318, @ngraham wrote:I think we could even get rid of the More Options/Fewer Options toggle button and just always show the search options now that they don't take up much space.
Baloo now only outputs multiple entries in the property map instead of stringlist as of D19088 and clients all handle this since Applications 19.08
Why do the comboboxes look so thin in the video?
This is a lot nicer. I think we could even get rid of the More Options/Fewer Options toggle button and just always show the search options now that they don't take up much space.
- Validate terms before applying
Oct 12 2019
Oct 11 2019
Any progress on this?
Oct 10 2019
- Fix typo and correct i18n markers
More info on the i18n markers here: https://api.kde.org/frameworks/ki18n/html/prg_guide.html#good_ctxt
Rebasing
Oct 9 2019
- Add setFacetsVisible() method
- Update toggle button on showEvent
In D23557#543977, @meven wrote:In D23557#543918, @elvisangelaccio wrote:@meven Looks like this patch is not enough: https://build.kde.org/job/Applications/job/dolphin/job/kf5-qt5%20SUSEQt5.12/78/testReport/projectroot.src/tests/placesitemmodeltest/
Could you have a look?
It seems to me the test fails because is uses an older kio version than 5.63 with D7446 applied creating "Recently Saved" group instead of group "Recent".
FAIL! : PlacesItemModelTest::testGroups() Compared values are not the same
Actual (groups.at(2).second.toString()): "Recently Saved" Expected (QStringLiteral("Recent")) : "Recent"The actual shows that the old "Recently Saved" group was created when it should not have been.
I don't know how this jenkins job works, but I would recommend cleaning the cached artifact KIO 5.63.0 for dolphin somehow.
We can alternatively limit this test run only on KF5 >= 5.64
@bruns are you good with this now?
In D23557#543918, @elvisangelaccio wrote:@meven Looks like this patch is not enough: https://build.kde.org/job/Applications/job/dolphin/job/kf5-qt5%20SUSEQt5.12/78/testReport/projectroot.src/tests/placesitemmodeltest/
Could you have a look?
Oct 8 2019
@meven Looks like this patch is not enough: https://build.kde.org/job/Applications/job/dolphin/job/kf5-qt5%20SUSEQt5.12/78/testReport/projectroot.src/tests/placesitemmodeltest/
In D20532#542673, @fbg13 wrote:In D20532#542522, @elvisangelaccio wrote:Adding a new setting makes the code and the UI more complex and harder to maintain in the long term. We should add new settings only when there is no other way (this usually happens if a new feature makes sense only for a subset of users).
But even if we go with just the "Go up" action, I think there should be an option to turn it off.
I often miss the file or folder when double clicking and going up would be annoying.
Thanks
Nice code improvement
Oct 7 2019
Split out refactor of isIndexingEnabled and depend on D24478
In D24446#543268, @elvisangelaccio wrote:Would you mind to move the isIndexingEnabled() refactoring to another commit?
Would you mind to move the isIndexingEnabled() refactoring to another commit?
- Name change
useBalooSearch() is not accurate; it's phrased as a command, not a question. I'd recommend changing it to usesBalooSearch or indexingEnabled or balooIndexingEnabled or something like that.
Oct 6 2019
- Add action to switch to the last tab
- Clean patch into several revisions
In D20532#542522, @elvisangelaccio wrote:Adding a new setting makes the code and the UI more complex and harder to maintain in the long term. We should add new settings only when there is no other way (this usually happens if a new feature makes sense only for a subset of users).
But even if we go with just the "Go up" action, I think there should be an option to turn it off.
I often miss the file or folder when double clicking and going up would be annoying.
I think Alt+0 for the tenth tab is fine, but maybe we should have it switch to the last tab instead. I doubt it's that common for people to have ten tabs open in Dolphin!
As far as I know "Switch to tab n" shortcuts are not identical between applications and vary quite a bit. Some applications use 0-9, some use 1-9. Some have a "Switch to last tab" shortcut assigned to 0 or 9. Some use Ctrl, some use Alt instead or even both.
In the case of web browsers I would assume they don't use 0 because Ctrl + 0 resets the zoom level, and Firefox on Windows uses Ctrl instead of Alt, Chrome uses both Ctrl and Alt I think but I don't use Chrome.
In D24422#542666, @bruns wrote:Please do one change at a time.
Please do one change at a time.
- Add parsing for filename/content
I don't see a compelling reason why this feature shouldn't work when you're using the systemwide double-click mode. I use that mode and this feature works perfectly fine with it.
Another thing that I forgot: the double click trigger should only be enabled in single-click mode, probably.
Yep, the i18n issue if fixed now.
In D7446#542407, @dfaure wrote:This commit breaks kfileplacesmodeltest, please take a look.
Patch seems to work fine, but the lack of commit message makes it hard to understand where the actual bugfix is. Please simplify the patch as mentioned inline.
This commit breaks kfileplacesmodeltest, please take a look.