Previously, a QTimer was used to start processing of our event queue in
the main Qt event loop. Unfortunately, QTimers are not thread-safe, and
disallow starting of a timer from a different thread than it was
created in.
Instead, use a dummy QObject to post a custom QEvent to the main loop
from whatever thread we like, and process our event queue when it is
received by our dummy object.
The logic for `EventLoopImplementationQt::pump()` caused calls to
`EventLoop::pump(PumpMode::DontWaitForEvents)` to not consume events
posted to the Qt event queue. This caused the window to be unresponsive
even when polling the event loop, if waiting was not desired.
Fixes infinite spinning in the cases when CSSPixels does not have
enough precision to represent increase per track which happens when
very small extra_space got divided by affected tracks number.
Rework the write_if_changed logic to properly truncate the output file.
The original logic would not truncate the file, leading to extra junk
at the end of files if the IDL files shrunk between builds.
3dd3120a8a changed open mode from Write to
ReadWrite, which stopped truncating files.
This could be noticed by building a slightly older branch, as compiling
it gave compile errors.
Only the top 4 entries from this file actually show up on the GitHub
interface, so instead of cutting some people off, let's just not have
the list.
There are already sponsorship links in README.md anyway :^)
Change associativity in computing of replaced element size to improve
precision of division.
Fixes vertically squashed image from Mozilla splash page MDN example.
Specifically, `CodeBlock::render_lines_for_terminal()` eats up blank
lines because it uses `DeprecatedString::split()` without the
`SplitBehavior::KeepEmpty` enum. Easy fix: use the enum.
The Track itself caches the Samples after each processing step which
allows it to be queried without the need to process it every time.
This result is queried by the WaveWidget which then caches the result to
prevent unnecessary heap allocations every paint event.
Removes the Sample struct inside Piano and replaces it with the struct
from LibDSP.
It automatically scales the height of the wave depending on the maximum
amplitude, as the Samples now contain floats and not integers.
This patch checks for visible items to determine the menu height. Now
the last visible item is used to determine the height of the menu.
Before this patch that menu height could be wrong e.g. if the last item
was not visible.
This patch adds the ability to refine a editing mask by a color-range
based on the pixels of the content image. This is useful for image
editing where mask restirction via luminosity might not fit or just
certain color ranges should be edited.
With this change the image modifications of a filter is only applied to
regions of the image where a editing mask is defined. If no editing
mask is defined the filter modifications are applied to the whole
image.
This adds a function where editing masks can be refined by selecting
a luminosity range that is applied to the content image and mapped to
the editing mask. This function allows the editing of image regions
that match only certain luminosity values.
This patch adds a function to make the editing-eask visible while
beeing in mask-mode so that the user can see which parts are covered
by the masks and can therefore be modified by other tools that support
editing masks.
Previously any expansion closing sequence would've caused the entire
expansion chain to be terminated, fix this by keeping track of active
expansions and running the parser in 'skip' mode.
Fixes#19110.
- Requesting an unsupported image type will now fallback to PNG
(which is now always the case),
- Errors should return 'data:,' instead of empty string,
- Added spec comments
For some reason, the tests are flaky for the aarch64 architecture when
running on a CI runner. This causes a lot of unnecessary red crosses for
PRs and since the test outcome doesn't seem to be too trustworthy, we
are better off not running these tests for now.
A previous commit made it so that SeekTable doesn't provide a seek
point from `seek_point_before()` if there is not a seek point before
the requested sample index. However, MP3Loader was only setting a seek
point after the first 10 frames, meaning that it would do nothing when
seeking back to 0.
To fix this, add a seek point at byte 0 for the first sample, so that
`seek_point_before()` will never fail.
Instead, perform the filtering for each rule as we go. This avoids
creating a separate list of rules, which was ~5% of runtime when
mousing around on the Discord web interface.
