The builder test was relying on default visiblity
for non-static functions. Make it explicit that we
want to export these functions, so the test keeps
working when we change the default visibility.
The relevant question here is about details, because we have to choose
if we declare alpha-only formats as having their (nonexistant) color
channels premultiplied or not, so that the code paths using them can do
the right thing.
Because we are premultiplied by default, it makes sense to treat alpha
like that, because then the alpha-only code doesn't need to do
workarounds for straight alpha.
Where this is relevant of course is when expanding the alpha channel
into color channels, where we want to end up with white.
So make sure we do color = alpha there instead of color = 1 like we did
before.
We need them for mask-only textures.
For tiffs, we convert the formats to RGBA (the idea that tiff can save
everything needs to be buried I guess) as tiffs can't do alpha-only.
Inverted alpha masks have an effect on the source, even if the mask
doesn't cover the source at all - or worse, is completely clipped out.
The GL renderer handles this fine, but Cairo and Vulkan had
optimizations that got this wrong.
In particular, fix the combination of luminance and alpha. We want to do
mask = luminance * alpha
and for inverted
mask = (1.0 - luminance) * alpha
so add a test that makes sure we do that and then fix the code and
existing tests to conform to it.
The result of calling update_property needs
to be that the property is marked as set
afterward, even if the value we pass happens
to match the default value.
After this change, scrollbars have value-now
show up as zero in the accessiblity page of
the inspector, even when that matches the lower
bound.
Test included.
Fixes: #5886
With the current approach, we get duplicate labels
in the accessible name: _Cancel Cancel. Change things
around to always set the labelled-by accessible relation
if we have a label, and not the label accessible property.
When running the tests, only run the random (and potentially large) size
download test once instead of 10 times.
There's no real benefit in doing that, both because it's unlikely to
fail only in the 2nd or 9th run and because the sizes are picked
randomly.
This also speeds up the test massively as the download test was
dominating the runtime.
Instead of picking a few numbers in advance and running them through the
test gauntlet every time, pick the random numbers at runtime.
This both increases the test coverage in that it ultimately tests more
combinations across many runs and it reduces the runtime of individual
runs because every tun only runs the download tests twice (with 1px and
the random size) instead of 5 times.
And that speedup benefits the CI, where the asan runs would cause this
test to timeout sometimes.
Make it use an alpha value that is well defined, ie 0.4 instead of 0.5.
0.4 * 255 = 102
0.5 * 255 = 127.5
This avoids rounding issues where some math may cause the resulting
alpha value to be 127, and some other math ends up with 128.
The idea is that for a rectangle intersection, each corner of the
result is either entirely part of one original rectangle or it is
an intersection point.
By detecting those 2 cases and treating them differently, we can
simplify the code to compare rounded rectangles.
When registering an observer, we send a notification and for that we need
to query the action's state and param type. When setting up a muxer parent,
same thing happens, except the action is queried on the parent instead.
This means that the muxer will notify observers about the parent's actions,
but not about its own.
Add a test to verify it works.
Fixes https://gitlab.gnome.org/GNOME/gtk/-/issues/5861
Add some odd-sized texture sizes to the
download tests, to trigger alignment issues
in the various upload code paths. And add
a size that is bigger than the max-texture-size
we force in one of our test setups.
To compensate, reduce the number of
runs per size from 20 to 10.
The GL renderers like to premultiply content that isn't, and due to the
data loss with alpha == 0 (transparent white, transparent black and
transparent anything are all represented by (0, 0, 0, 0) when
premultiplied) these values cannot be converted back.
There is no longer a need to use gdk_texture_download() and force
conversion to ARGB8 format. We can download the pixels in the original
format again.
That way we avoid testing the conversion code and avoid having to deal
with differences in representable colors.
However, some formats do do conversions, so we allow pixel comparisons
to be accurate (requires 16bit comparison accuracy) or inaccurate (we
only care about 8bit).
Note that for the default RGBA formats, this is identical and means they
need to be bit-exact the same, no matter what.
But the higher bit depth formats may be more different - floating point
can even have different values with high accuracy (the float mantissa is
23 bit, we only care about 16).
When we emit items-changed due to a section
sorter change, don't also emit sections-changed.
Instead make the items-changed signal cover the
whole range.
Tests included.
This one tests a crossfade between two non-overlapping nodes with a clip
region that covers neither of the two nodes.
This tests that renderers can deal with clip regions that doesn't
overlap nodes in a situation where they will most likely want to create
an offscreen.
As offscreens are typically clipped to the clip region, this would cause
an empty offscreen and that can cause failures.
This was an experiment where an offscreen was translated inside an
existing clip.
Because renderers try to limit offscreens to the clip rect, this is
interesting, because they might get the translation wrong.
Using gdk_texture_new_from_resource() is not valid here because we are
not sure if the given resource is valid.
Plus, the previous optimization is no longer relevant, because we are
not using gdk_pixbuf_new_from_resource() anymore - which was what this
optimization was about before it was ported to GdkTexture.
Test attached.
Add some tests for handling of failures.
The test data here is taking from gdk-pixbufs
tests/test-images/fail directory, excluding anything
but png, tiff and jpg images.
Instead of injecting `-fvisibility=hidden` depending on a compiler check
ourselves, let Meson do it for us.
This also avoids us having to filter `-fvisibility=hidden` when reusing
the common compiler flags.
Our notify tests would fall over if there was
a duplicate enum value (within the first 10 values).
Make it handle that, by skipping the duplicate value.
This is failing because I can't figure out
how to make wireplumber and pipewire work
in ci enough to let me add a new monitor :(
As usual, the test works fine locally.
The rounded-clip-in-clip-3d test fails in GL when
flipped. Given that it was already excluded from cairo,
and also fails cairo when flipped, give up on it for now.
The repeated tests were not careful enough to produce
the correct reference image to match what the repeat
node does.
With these changes, all cairo tests pass.
Add separate suites for running the gsk compare-render
tests with the --flip, --rotate or --repeat options.
A bunch of these fail currently, and need diagnosis.
Add options to the gsk compare-render test for
modifying the node (and do a matching change to
the reference image).
flip: negative scale flipping things horizontally
rotate: 90 degree rotation
repeat: 2x2 grid
There are a lot of cases where properties are implemented in classes but
the getters for these exist in an interface that class implements.
A common Example is g_list_model_get_n_items() being the getter for
GtkWhateverListModel::n-items.
But also property implementations that don't use override_property()
(usually because they have a different default) are handled by this.
The test doesn't hold 2 references, it holds only one.
The reason one unref can cause a leak is that some backends - like X11 -
only destroy the surface once the DestroyNotify event from the X server
has come in.
We want to support GLES 2, so make sure we test that support.
Also force-disable common extensions we don't explicitly check for and
don't want to accidentally use.
This one exhaustively tests reusing the same model as a child model for
many nodes.
This tracks that multiple items-changed signals emitted at the same time
(or multiple handlers for one such signal) doesn't put the treelistmodel
in an inconsistent state while it is handling all of them.
I'm not sure this (ab)use of treelistmodel should be officially
supported, but it works today, so let's test it to see if we can keep it
working.
In constrast to our other tests, these use
textures that are big enough to force slicing
with setting GSK_MAX_TEXTURE_SIZE, which we
will use in the following commits to improve
test coverage.
That way, we can return the item even after the row is removed. This is
particularly relevant in ListItemFactory::unbind callbacks because they
often use gtk_tree_list_row_get_item() and user code never tracks
changes to this property.
A side effect of this is that the item will survive until the row gets
destroyed, but that's what users expect anyway, so we can live with it.
Related: #5646
In certain scenarios, address the issue where gnome.compile_resources
fails to transmit the present source directory. This is most notably
visible with MSBuild.
The GL renderer was creating sripes for nodes that were scaled in
particular ways, probably due to rounding errors.
This testsuite focuses on one of those stripes to make sure they are
gone.
This test fails if we naively create fullscale
intermediate offscreens. This was fixed in the
previous commits.
This tests the fixes in 22ba6b1f33 (for
cairo) and 3a0152b65f (for GL).
The previous code would include CSS padding/margin/border in the
measurement and that is wrong.
Until commit a96c75ff02 this was not actually visible, but afterwards
listitems were allocated 16px too wide.
Test included
The design patterns using statusbar are no longer popular,
and it is pretty easy to make a statusbar yourself with boxes
and labels, if you need one. The only thing special about
GtkStatusbar was its window resize handle, but that has
been gone for a long time.
The unaligned-offscreen and upside-down-label-3d tests are failing after
upgrading our CI images, seemingly because of some font rendering issue
that is hard to track. Let's use the "failing" testsuite mechanism that
we also use for the reftests.
These tests can be run manually, but are not suitable for use as an
acceptance test, so let's not make frameworks like Debian's autopkgtest
run these when they run ginsttest-runner in the most obvious way.
Signed-off-by: Simon McVittie <smcv@debian.org>
There are two possible interpretations of "expected failure": either
the test *must* fail (exactly the inverse of an ordinary test, with
success becoming failure and failure becoming success), or the test
*may* fail (with success intended, but failure possible in some
environments). Autotools had the second interpretation, which seems
more useful in practice, but Meson has the first.
Instead of using should_fail, we can put the tests in one of two new
suites: "flaky" is intended for tests that succeed or fail unpredictably
according to the test environment or chance, while "failing" is for
tests that ought to succeed but currently never do as a result of a
bug or missing functionality. With a sufficiently new version of Meson,
the flaky and failing tests are not run by default, but can be requested
by running a setup that does not exclude them, with a command like:
meson test --setup=x11_unstable --suite=flaky --suite=failing
As a bonus, now that we're setting up setups and their excluded suites
programmatically, the gsk-compare-broadway tests are also excluded by
default when running the test setup for a non-broadway backend.
When running the tests in CI, --suite=gtk overrides the default
exclude_suites, so we have to specify --no-suite=flaky and
--no-suite=failing explicitly.
This arrangement is inspired by GNOME/glib!2987, which was contributed
by Marco Trevisan.
Signed-off-by: Simon McVittie <smcv@debian.org>
Add a test that runs make-pot. This will only pass
if you've updated po/POTFILES.in and .skip after
moving source files around.
Unfortunately, it won't catch new source files that
are missing.
These are being replaced by GtkFileDialog.
This commit only moves the headers for GtkFileChooserWidget and
GtkFileChooserDialog to deprecated/, and keeps the implementations
in gtk/, since they will eventually be salvaged into a private
GtkFileChooserWindow.
These are being replaced by GtkColorDialog
and GtkColorDialogButton.
This commit only moves the headers for GtkColorChooserWidget
and GtkColorChooserDialog to deprecated/, and keeps the
implementations in gtk/, since they will eventually be
salvaged into a private GtkColorChooserWindow.
