The cache was introduced to improve performance by not querying the
OS for the window process path every time we need to check if the window
is interesting to FancyZones. Since then other changes were made to the
the way we check the windows. Right now, the IsInterestingWindow function is
called when:
1) WinKey + arrows are used
2) window is started to be dragged
3) window is created
1) and 2) are initiated by the user, happen only once per interaction so
their performance impact can be dismissed. The 3) happens all the time
but for the most part the check for WS_CHILD or
GetAncestor(window, GA_ROOT) == window will filter those out. In the
end, only top-level windows will be queried for their path.
Removing the cache improves code readability and will make code
maintenance easier.
Do not swallow WinKey+Arrow events for non-interesting windows. This makes apps that are in the "excluded apps" list behave as if "Override windows snap keys" is disabled - they will react to the Windows default snap.
* localized dllmain.cpp of fancyzones project
* localized FancyZones.cpp
* format fancyzones.rc file
* Moved SuperFancyZones back to being a string instead of having it in the resource file as it is the window class name
* reverted changes for window name
* Formatted fancyzones rc file
* Localized shortcut_guide.cpp
* localized overlay_window.cpp
* formatting changes
* Localize overlay window
* removed the README link from the set of localized resources
* Typo: changed upper to lower
* removed hardcoded strings from CanvasEditorWindow.xaml
* removed hardcoded strings from GridEditorWindow.xaml
* loc
* Localized MainWindow
* reverting MainWindow.xaml as it is not rendering the window as expected
* Changed the resource settings from internal to public
* the culture is set based on the culture of the system UI set in the system settings
* Removed the french resource files used for testing
* Localized canvasWindow and mainwindow
* Removed setting the UI culture explicitly as it would be implicitly set to the culture of system UI
* Removed redundant header file
* fix typo
* make function obsolete
it is not used in the code
* rewrite the function that converts chinese chars to pinyin
1. Only difference in this rewrite is instead of returning 2D array, return as a combined single string of all the possible pinyin combination. Since fuzzy search does character matching, this shouldn't be a problem.
2. Added a function that returns a custom language converter. In this case Pinyin converter. New converters can be added.
* Use new language converter param + strip out ScoreForPinyin method
* update
* Change parameter name
* fix failing tests
* WIP
* Remove todo
There should be some distinction between score after precision filter and actual raw score derived from FuzzySearch. Although so far RawScore is used in testing, but it seems to describe the structure. Originally it was to avoid assigning score directly as it would be hard to reason about that output of FuzzySearch score is.
* Add constructors, remove default to enforce required properties
* remove setting rawscore in SearchPrecision
* Change method name to reflect intention
* Change parameter name + update comment
* update
* Remove params comment
Co-authored-by: theClueless <14300910+theClueless@users.noreply.github.com>
