We normally wait for the surface ::size-changed signal,
but then we only queue a resize, which is not sufficient
on map - we need to have an allocation by the time start
moving focus around in show. So, call check_resize()
right away.
Rename gtk_window_move_resize to gtk_window_do_resize
(no moving involved anymore), rename configure_notify_received
to need_resize and replace gtk_window_guess_default_size
by gtk_window_compute_default_size.
We should not emit configure events before we
are realized - size changes at this point are
not relevant.
This gets rid of a mysterious emission of
GdkSurface::size-changed (52, 52), that is happening
when GtkWindow sets the shadow_width before the window
is mapped.
Surpisingly, when a monitored file gets atomically replaced
with g_file_set_contents(), the file monitor reports a
DELETED event for the temp file, which is not the file
I'm monitoring.
Just ignore DELETED events.
Fixes: #3036
G_DEFINE_DYNAMIC_TYPE declares these functions as static, and
-Werror=redundant-decls won't let us redeclare them. This is the
equivalent of 72c72d0b, but for a different backend.
Signed-off-by: Simon McVittie <smcv@debian.org>
Currently there is no way to alter the offset of the popup when positioning
with GdkPopupLayout. This makes using the popup difficult for scenarios
like completion windows where you may need to offset the window by a given
amount for aligning text.
gtk_popover_set_offset() allows setting these values and are analagous to
the function of the same name for GdkPopupLayout.
When we invalidate a y_range using the common pattern of y==0 and
old_height==new_height, we are generally invalidating the entire buffer.
This short-circuits that case to just invalidate the buffer in a faster
and more complete form. The problem here appears to be that we can't
always calculate the ranges properly to invalidate because validation
has not run far enough.
The colors demo was leaking a reference to its model.
This was showing up as crash when you manage to close
both the demo window and the main window while the
sorting is ongoing.
Drop the homegrown highlighting code, and just use highlight
to produce Pango markup.
When using an external highlighter, we can also highlight css,
xml, headers, at least.
When asking for a password, the message string is split on primary
and secondary if it contains a newline character. However, the newline
character is currently part of both strings, which creates weird
spacing between the GtkLabels. I suppose this is bug, which was not
visible as in most cases (if not all) the message string hasn't
contained the new line characters so far. But we are going to change
that now, see GNOME/gvfs!82. Let's drop the new line character similarly
as it is done when asking for a question, or showing processes in order
to fix the weird spacing.
We might break the loop early, e.g. if we're unmapped before the round
trip finishes, and to avoid the callback to write to invalid stack
memory, destroy the callback so it won't be invoked.
Fixes: #3026
We were not removing the pending_update idle
in dispose, which is at least suspicious, if
not deadly. Move the idle cleanup code into
unset_widget(), which we are already calling
in dispose().
As things currently stand, we get events for focus changes
before the widget is allocated, and try to scroll in response.
Therefore, leaving n_columns at 0 until size-allocate leads
to plenty of division-by-zero. Just set it to 1 initially
to avoid that. This is a workaround for #3025.
The testsvg test uses a method in librsvg that was introduced in
2.46.0. The test is now skipped if the librsvg version is too old.
(It was previously already skipped if librsvg wasn't found.)
The regular paths just emit ::end, which isn't the whole thing.
This gives an opportunity to gestures that are stolen the sequence
to clean themselves up.
The win toolchain has problems linking meson subprojects to libgtk.
To work around that build glib/pango from git and install them first
as long as the installed version is too old at least.
Fixes#3002
GtkAspectFrame code assumes that its child will be in
self->child, but that is only the case if we arrange
for <child> in ui files to end up calling
gtk_aspect_frame_set_child(). Therefore, implement
GtkBuildable.
Fixes: #3020
We were playing fast-and-loose with private GIO data
when showing settings bindings in the property editor,
and this was causing crashes.
We can show this information again if GIO ever gets
api to introspect it.
Fixes: #3015
GTK will not up front know how to correctly calculate a size, since it
will not be able to reliably predict the constraints that may exist
where it will be mapped.
Thus, to handle this, calculate the size of the toplevel by having GDK
emitting a signal called 'compute-size' that will contain information
needed for computing a toplevel window size.
This signal may be emitted at any time, e.g. during
gdk_toplevel_present(), or spontaneously if constraints change.
This also drops the max size from the toplevel layout, while moving the
min size from the toplevel layout struct to the struct passed via the
signal,
This needs changes to a test case where we make sure we process
GDK_CONFIGURE etc, which means we also needs to show the window and
process all pending events in the test-focus-chain test case.
gdk_gl_context_has_framebuffer_blit() and gdk_gl_context_has_frame_terminator()
were only used by by GDK/Win32, and they do not provide performance advantages
in GTK master, so clean up the code a bit by dropping them.
Check whether we really have x11 and wayland enabled before we try to setup the
tests to use these respective GDK backends, and only attempt to setup tests
running with the Broadway backend if it has been enabled.
Also, add a setup for running tests with the GDK-Win32 backend on Windows, for
builds that target Windows.
Use gdk_surface_get_geometry() to get the correct x and y coordinates of the
window that we are resizing, so that the window does not reposition itself
automatically at the top-left corner at resizing as we to used hard-code the x
and y coordinates to 0.
By doing so, we ensure that resizes of windows will work on Vulkan renderer, by
first calling gdk_win32_surface_handle_queued_move_resize() before we proceed
as usual
Use the shared function that was added in the previous commit, to simplify
things.
Also make gdk_win32_surface_get_queued_window_rect() and
gdk_win32_surface_apply_queued_move_resize() back into static functions, since
they are now used only by the code in gdksurface-win32.c
Since we need to deal with queued moves and resizes in the Cairo, GL and Vulkan
draw contexts, and the logic involved in all three of these are largely
similar, add a function gdk_win32_surface_handle_queued_move_resize() that will
handle this, which will be shared between these three types of draw contexts.
Move gdk_win32_surface_get_queued_window_rect() and
gdk_win32_surface_apply_queued_move_resize() to gdksurface-win32.c, since these
functions are not only used for Cairo draw contexts, but is also used for GL
draw contexts, and will be used for Vulkan draw contexts.
Don't get the default display when we compute the Aerosnap region, but instead
get it from the underlying GdkSurface that we are using for the computation.
Also, don't unref the monitors that we obtain from the display in the wrong
place, which was why we had crashes whenever we triggered AeroSnap code (and we
are actually not supposed to do that as they are owned by the GdkDisplay that
is owned by the GdkSurface we are using), and this will eliminate lots of
criticals that are spewed as a result.
gtk-doc assumes Docbook4, with <ulink> and so on.
Without this, all the links in markdown are converted
to <link xlink:href=...> and then lost in the docbook->html
conversion.
We can't use the ::destroy signal anymore; use
a weak ref instead, and make ensure the entry
stays around long enough for us to finish the
cleanup.
Fixes: #3004
No need for an intermediate box widget here;
we can just use a box layout. As a side-effect,
this fixes the theme to apply to the selection.
Fixes: #3005
ss00 doesn't exist, and we use xxxx as placeholder
for 'default' choices in alternatives. Add a warning
in case we run across invalid OpenType feature tags
in fonts.
Fixes: #2962
If we have a y==-1 then we are generally invalidating the whole textview.
For this case, we can just discard the entire GtkTextLineDisplay cache.
Fixes#2975
We don't want the completions to pop up after we call
gtk_file_chooser_set_current_name(). This used to be
handled by gtk_entry_set_text() blocking the completion
signal handler. We don't have that anymore, so block
popup completion around the call to gtk_editable_set_text()
instead.
Fixes: #2995
We can only insert tags in the buffer if they come
from the same GtkTextTagTable as the buffer uses.
If that is not the case, paste the text without tags.
Fixes: #2991
While we guarantee that the widget that a controller
is attached to stays around while it is handling an event,
the same is not true for the root that the widget belongs
to. In corner cases (such as clicking "Close" in the
fallback window menu), it may already be gone.
Avoid a critical in that case.
Fixes: #2998
We don't want to select on focus-in when the focus
comes from a child. The case where this does harm
is when you activate copy or paste actions from the
context menu. We close the menu before triggering the
action, and if that causes the text in the label to
be selected, unexpected things happen, since the action
applies to the current selection.
This is the equivalent of cd9f5733b3 for GtkLabel.
The code used to do
if (parent_class->clicked)
parent_class->clicked (...)
That is pointless because the parent_class never changes,
so there' no need for that if and commit 415946eb0f
took it out. Unfortunately, p arent_class->clicked is NULL
though, so the whole call needs to go.
This check used to read if (grab || device_type != GDK_DEVICE_TYPE_PHYSICAL),
the grab check was only reserved to physical devices, which the current
pointer device definitely doesn't act like. So the condition was "fixed" the
wrong way around, and the latter check is now moot, so the condition should
really go away. We always want to check the new toplevel under the pointer
here.
Fixes: https://gitlab.gnome.org/GNOME/gtk/-/issues/2970
This allows us to use DPI_AWARENESS_CONTEXT_PER_MONITOR_AWARE_V2 for the
DPI awareness mode, which will help us to better support use cases with
multiple monitors. This is actualy a more advaned version of the
current PROCESS_PER_MONITOR_DPI_AWARE via using SetProcessDpiAwareness().
Note that this is not enabled by default, but also enabled via using
GDK_WIN32_PER_MONITOR_HIDPI, as in the PROCESS_PER_MONITOR_DPI_AWARE
case.
Note also, that appliation compatibility settings and DPI-awareness
manifests takes precedence over this API call, as before.
Since we have now made the Win32 GL contexts share the global context as
the other backends have, we are more ready to use the GL renderer by
default on Windows as well.
Note that currently we can only enable this when not running on
OpenGL/ES as the OpenGL/ES shaders are not ready at this point, and the
OpenGL/ES support that we have from libANGLE does not support full
desktop OpenGL operations.
Like the other backends, we ought to create our WGL/EGL GL contexts like
the following:
"Create a global GL context that connects all GL contexts on a display
and lets us share textures between them."
Use the new GTK_CROSSING_ACTIVE crossing events to update
the im context focus when the window becomes active or
inactive. IBus requires this, since it has only a single,
global focus location.
Fixes: #2390
Emit crossing events when the active window changes.
We don't want to emit GTK_CROSSING_FOCUS events, since
every window has its own focus location (focus does not
jump from window to window), so we use the new
GTK_CROSSING_ACTIVE type of crossing event for this.
Document the different crossing event kinds that we use,
and add GTK_CROSSING_ACTIVE. We are going to use it in
the future when the active window changes.
If GLES support is enabled on Windows, force GLES mode if we are running
on a ARM64 version of Windows (i.e. Windows 10 for ARM).
This is required as ARM64 versions of Windows only provide a software
implementation of OpenGL 1.1/1.2, which is not enough for our purposes.
Thus, we could make instead use the GLES support provided via Google's
libANGLE (which emulates OpenGL/ES 3 with Direct3D 9/11), so that we
can run GtkGLArea programs under OpenGL/ES in ARM64 versions of Windows.
Note that eventually we could update the libepoxy build files for Windows
to not check nor enable WGL when building for ARM64 Windows, as the WGL
items do not work, although they do build.
We need to use GL_BGRA instead of GL_RGBA when doing glReadPixels() on
EGL on Windows (ANGLE) so that the red and blue bits won't be displayed
inverted.
Also fix the logic where we determine whether to bit blit or redraw
everything.
Some implementations of the ES 1.00 shader (such as Google's ANGLE) do
not like the 'f' suffix for floats, so just drop it, as it should be
harmless to drop.
This is for adding a EGL-based renderer which is done via the ANGLE
project, which translate EGL calls to Direct3D 9/11. This is done as a
possible solution to issue #105, especially for cases where the needed
full GL extensions to map OpenGL to Direct3D is unavailable or
unreliable, or when the OpenGL implementation from the graphics drivers
are problematic.
To enable this, do the following:
-Build ANGLE and ensure the ANGLE libEGL.dll and libGLESv2.dll are
available. A sufficiently-recent ANGLE is needed for things to
work correctly--note that the copy of ANGLE that is included in
qtbase-5.10.1 is sufficient. ANGLE is licensed under a BSD 3-clause
license.
-Build libepoxy on Windows with EGL support enabled.
-Currently, prior to running GTK+ programs, the GDK_DEBUG envvar needs
to be set with gl-gles as at least one of the flags.
Known issues:
-Only OpenGL ES 3 is supported, ANGLE's ES 2 does not support the needed
extensions, notably GL_OES_vertex_array_object, but its ES 3 support is
sufficient.
-There is no autodetection or fallback mechanism to enable using
EGL/Angle automatically yet. There are no plans to do this in this
commit.
...EGL support needs to be explicitly enabled during the build of
libepoxy on Windows as it is not enabled by default on Windows.
With this, we can add an EGL renderer for Windows that make use of
Google's libANGLE, which is a library that translates OpenGL/ES calls
to Direct3D 9/11, which will provide better hardware compatibility
on Windows and would act as one of the foundations to resolve issue #105.
Set the accessible role for GtkLinkButton to button.
We don't use the 'link' role since ARIA says "if it
behaves like a button, use 'button'".
Update docs and add a test.
This changes should not be neccessary, since
GtkLinkButton derives from GtkButton, see #2965.
It's not a portable API, so remove it. The corresponding backend
specific functions are still available, if they were implemented, e.g.
gdk_macos_monitor_get_workarea() and gdk_x11_monitor_get_workarea().
pandoc insists on using the xlink namespace for hrefs,
and the namespace setup doesn't carry over xi:includes.
My first fix was to tell pandoc to generate standalone
docbook documents, which makes it insert the xlink
namespace. But it also makes it wrap all sections and
chapters in articles, and that messes up our toc structure.
So, patch things up differently by stripping the xlink:
from hrefs via regex.
Yay for XML!
Make GdkEvents hold a single GdkDevice. This device is closer to
the logical device conceptually, although it must be sufficient for
device checks (i.e. GdkInputSource), which makes it similar to the
physical devices.
Make the logical devices have a more accurate GdkInputSource where
needed, and conflate the event devices altogether.
Besides the implicit x/y assumptions, devices don't have axes. Those
are actually provided by the GdkDeviceTool driving the device, and
different tools may have different axes.
It does not make sense to offer this API that can change beneath
someone's feet, we now have gdk_device_tool_get_axes() which is static
to the tool.
Use the label accessible role for GtkLabel. ARIA has some
ominous wording about it going way, but while we have it,
GtkLabel is the obvious candidate for carrying it.
Update the documentation and add a test.
Sysprof has moved to a new ABI which removes GLib from the capture library
so that GLib itself can link against sysprof-capture.
This bumps the library ABI so we can keep things coordinated between all
the new tracing layers in the stack.
In some cases we explicitly want to unset an accessible attribute; for
instance, an accessible property is gated on a widget property, and if
the widget property gets unset, the accessible property should be reset.
We're currently overloading NULL to mean both "this value is undefined,
and should be reset to its default" and "the value collection failed".
Let's do error reporting right, by using GError to mean "the collection
failed, for this specific reason"; then, we can use a NULL return value
to signal that the accessible attribute should be reset to its default
value.
This is only relevant for pointer-sized attribute values: strings,
references, and reference lists; numeric, boolean, tristate, and token
values either cannot be undefined, or have a specific "undefined" value.
Looking at the xf86-input-wacom driver code, this is not even a thing
anymore. Drop this device type, in modern days there's
GDK_DEVICE_TOOL_TYPE_MOUSE for this.
Show a tab for accessibility information.
This shows the role and the accessible attributes
(states, properties, relations).
For now, changing the values is not possible, and
we only show the explicitly set values. In the future,
we want to show the attributes that are relevant for
the role, regardless of whether they are set or not,
and allow changing some of the attributes (the ones
that are not fully managed by GTK itself).
We're stopping activity mode when finalizing, which will change
accessible state; this will create a GtkATContext, and since GtkWidget
drops its GtkATContext on dispose(), we're going to end up leaking it on
the floor:
```
2,007 (64 direct, 1,943 indirect) bytes in 1 blocks are definitely lost in loss record 36,242 of 36,944
at 0x483977F: malloc (vg_replace_malloc.c:307)
by 0x5222105: g_malloc (gmem.c:106)
by 0x523E222: g_slice_alloc (gslice.c:1025)
by 0x523E261: g_slice_alloc0 (gslice.c:1051)
by 0x534B398: g_type_create_instance (gtype.c:1849)
by 0x53302EE: g_object_new_internal (gobject.c:1937)
by 0x53312AF: g_object_new_valist (gobject.c:2262)
by 0x532FEE8: g_object_new (gobject.c:1780)
by 0x4B3F942: gtk_test_at_context_new (gtktestatcontext.c:107)
by 0x491CC50: gtk_at_context_create (gtkatcontext.c:380)
by 0x4BFEDA0: gtk_widget_accessible_get_at_context (gtkwidget.c:8127)
by 0x4906079: gtk_accessible_get_at_context (gtkaccessible.c:83)
by 0x490618F: gtk_accessible_update_state (gtkaccessible.c:137)
by 0x4ACBA6D: gtk_progress_bar_act_mode_leave (gtkprogressbar.c:690)
by 0x4ACB4F8: gtk_progress_bar_finalize (gtkprogressbar.c:564)
```
We're also unparenting widgets and changing styles, which is another
potential source of leaks and side effects.
Track what we really need to send for inset shadows, which are used
as a border replacement in many cases.
Fishbowl says I can draw around 200-300 more switches per frame like
this too.
We must initialize the properties in init, since they
are expected to be there ab initio, and we can't call
gtk_accessible_update_property from finalize - it causes
us to recreate an at context and badness follows.
When converting DisplayLink frame presentation times, we need to take into
account the arch-specific types. This tracks changes in GNOME/GLib!1566 so
that precision is not lost.
Accessible values storing references and reference lists can be unset by
using NULL to mean "undefined"; since we cannot return the NULL value,
we need to intercept it when collecting a value, and replace it with an
undefined GtkAccessibleValue, which is also the value used as the
default for GtkAccessibleRelation values that store a reference or a
reference list.
Fixes: #2955
Make both gtk_grid_view_new and gtk_grid_view_new_with_factory
take a model as first argument, and make all arguments
allow-none and transfer full.
Update all callers.
Make both gtk_list_view_new and gtk_list_view_new_with_factory
take a model as first argument, and make all arguments
allow-none and transfer full.
Update all callers.
A dropdown without a model is useless, so accept a model
and expression in the constructor. Allow them to be NULL,
but consume them if given. This makes chained constructors
convenient without breaking language bindings.
Drop gtk_drop_down_set_from_strings() and instead add
gtk_drop_down_new_from_strings().
Update all users.
We don't pay attention to item-type anymore, so
drop the item-type property and the _for_item_type()
constructor, and allow passing NULL to the regular
constructor.
We don't make this constructor transfer-full, since
the selection filter model is not a wrapping model
like the others. It is more like fork than a wrap.
This is for consistency with other wrapping list constructors.
We want them all to be transfer full, allow-none.
Also make the constructor return GtkMultiSelection *.
Update all callers.
Make gtk_tree_list_model_new() take the root model
as first argument, and make it transfer full, for
consistency with other wrapping list constructors.
Update all callers.
Still missing here: Make the model property writable,
and allow passing NULL in the constructor.
We don't need as many functions to print out the property, relation, and
state of an accessible. Additionally, we should allow comparing the
accessible attributes with an expected value, and print out the real
accessible value if they do not match.
Some widgets have different roles after they are constructed, so we need
to allow changing the role defined by the class. We should still avoid
setting a role after the GtkATContext has been created.
This is a bit unfortunate, since the aria modelling
doesn't quite agree with ours, so we have to listen
for the togglebutton property change, and we inherit
the pressed state from the togglebutton accessible.
Only GtkWidget should use GTK_ACCESSIBLE_ROLE_WIDGET as its default
accessible role; the default for GtkAccessible and GtkATContext should
be GTK_ACCESSIBLE_ROLE_NONE, meaning "an element whose implicit native
role semantics will not be mapped to the accessibility API", according
to the WAI-ARIA specification.
We want to test the accessibility API, as well as the implementation
inside each widget. For that, we should expose an API that lets us
verify that a GtkAccessible has a given role, as well as a given
property.
The API follows the pattern of other GTest API:
- a macro to assert that a condition is respected
- a function that prints out the error message in case of failure
While we have split the various attributes for convenience, there's no
reason why we should have specialised data types for the attributes
container object.
Reduce the amount of subclassing, by handling collection of fundamental
types directly from the generic code paths. We now handle boolean,
tristate, integer, number, string, and relation values in the generic
code path; if an attribute supports the "undefined" value, we return the
undefined value singleton.
Drop roles and properties that were deprecated in WAI-ARIA 1.1, and add
new roles and properties defined in WAI-ARIA 1.2 and later.
We also split the relationship properties into their own enumeration, so
we can keep the GtkAccessibleProperty type more compact.
It's pointless, we can use an explicit value of `-1` everywhere.
Additionally, it complicates all code that uses the state enumeration as
an array index, since now we need to guard against a negative offset.
Some widgets have different accessible roles depending on some
parameter, so we cannot set the role at class init time. For those
widgets, we add an "accessible-role" property to GtkAccessible, and we
allow setting it (only) at construction time.
Each widget type has an accessible role associated to its class, as
roles cannot change during the life time of a widget instance.
Each widget is also responsible for creating an ATContext, to proxy
state changes to the underlying accessibility infrastructure.
An Accessible implementation must create an ATContext object. UI
elements are supposed to interact with the GtkAccessible API, but we
expose GtkATContext to allow patterns like delegation.
The ATContext type is meant to be used as the base class for
implementations of the assistive technology API—the actual mechanism
needed to communicate to components like the screen reader, or any other
AT.
Every time the widget state changes, the ATContext is meant to broadcast
the state change; and every time the AT queries the state of a UI
element, the ATContext is meant to provide that information.
We also have a "test" ATContext implementation, which is meant to be
used to write tests to verify that changes are propagated without
requiring a whole desktop session.
All accessible properties and states may have one of the following
types:
- true/false
- true/false/undefined
- true/false/mixed/undefined
- reference (to another UI element)
- reference list
- integer
- number (real numerical value)
- string
- token (one of a limited set of allowed values)
- token list
See: https://www.w3.org/WAI/PF/aria/states_and_properties#propcharacteristic_value
The GtkAccessibleValue is a simple reference counted type that can be
"subclassed" to implement each value type.
This initial commit adds GtkAccessibleValue and the basic subclasses for
plain boolean, tristate (true/false/undefined), and token types,
including statically allocated values that can be shared instead of
allocated.
GtkAccessible is an interface for accessible UI elements.
Currently, it doesn't do much except exist as a type; in the future, it
will be the entry point for all accessible state in GTK.
The list of roles is taken from the WAI-ARIA 1.2 specification:
https://w3c.github.io/aria/
Some of these roles do not make entirely sense from a GTK application
perspective, but we can remove them before finalizing the API.
To build a better world sometimes means having to tear the old one down.
-- Alexander Pierce, "Captain America: The Winter Soldier"
ATK served us well for nearly 20 years, but the world has changed, and
GTK has changed with it. Now ATK is mostly a hindrance towards improving
the accessibility stack:
- it maps to a very specific implementation, AT-SPI, which is Linux and
Unix specific
- it requires implementing the same functionality in three different
layers of the stack: AT-SPI, ATK, and GTK
- only GTK uses it; every other Linux and Unix toolkit and application
talks to AT-SPI directly, including assistive technologies
Sadly, we cannot incrementally port GTK to a new accessibility stack;
since ATK insulates us entirely from the underlying implementation, we
cannot replace it piecemeal. Instead, we're going to remove everything
and then incrementally build on a clean slate:
- add an "accessible" interface, implemented by GTK objects directly,
which describe the accessible role and state changes for every UI
element
- add an "assistive technology context" to proxy a native accessibility
API, and assign it to every widget
- implement the AT context depending on the platform
For more information, see: https://gitlab.gnome.org/GNOME/gtk/-/issues/2833
We are using floats for rgb, and we don't need more precision
for hsl colors either. We use hsl for computing color expressions
like shade(), lighter() and darker(), which are not precisely
specified anyway.
This commit updates the one test where the output changes a
tiny bit due to this.
Setting a width request is not quite enough, since
gtk_widget_set_size_request() only queues a resize
when the widget is visible. Explicitly force one
here. Without this, the popup sometimes shows up
too small.
This is truly a russian doll of documentation formats:
a string containing <> inside an xml fragment in an |[ ]|
gtk-doc example in markdown in a doc comment.
Sadly, something gets escaping wrong, so the <> end up
literally in the docbook and mess up the last step of
our document formatting, even after turning them into
entities.
Work around this with an extra level of entities that
really shouldn't be necessary.
This flag causes pandoc to emit a proper doctype
declaration and, crucially, namespace declarations
for the xlink namespace that it insists on using
for href attributes. Without this, putting external
links in md documents doesn't survive the journey
through xml.
Commit 0145809a94 replace the response-requested
signal with an action, but didn't actually remove the emission
of that no-longer-existing signal.
Fixes: #2942
Add a table mapping event signals to their event controller
replacements, and a table mapping former GtkContainer
subclasses to their gtk_container_add replacement.
SSave the missing keys as a bitset and iterate over that bitset in the
step function.
Solves the problem with a large UI block at the beginning of a sort
operation when all the keys were generated, in particular when key
generation was slow.
Benchmarks for maximum time taken by a single main loop callback:
initial sort with complex GFileInfo keys
old new
32,000 items 137ms 3ms
128,000 items 520ms 31ms
initial sort with string keys
old new
32,000 items 187ms 1ms
128,000 items 804ms 3ms
When updating a (partially) sorted model, take the known runs for the
existing sort and apply them to the new sort. That way, we don't have to
check the whole model again.
Benchmarks:
appending half the items to a model of strings
old new
512,000 items 437ms 389ms
1,024,000 items 1006ms 914ms
appending 10% of the items to a model of strings
old new
512,000 items 206ms 132ms
1,024,000 items 438ms 301ms
appending 1 item to a model of strings
old new
64,000 items 1.8ms 0.00ms
512,000 items --- 0.01ms
Previously, the sort was not stable when items were added/removed while
sorting or the sort algorithm was changed.
Now the sort looks at the item position (via the key's location in the
keys array) to make sure each comparison stays stable with respect to
this position.
This massively speeds up sorting with expensive sort functions that it's
the most worthwhile optimization of this whole branch.
It's slower for simple sort functions though.
It's also quite a lot slower when the model doesn't support sort keys
(like GtkCustomSorter), but all the other sorters do support keys.
Of course, this depends on the number of items in the model - the number
of comparisons scales O(N * log N) while the overhead for key handling
scales O(N).
So as the log N part grows, generating keys gets more and more
beneficial.
Benchmarks:
initial sort of a GFileInfo model with display-name keys
items keys
8,000 items 715ms 50ms
64,000 items --- 554ms
initial sort of a GFileInfo model with complex keys
items keys
64,000 items 340ms 295ms
128,000 items 641ms 605ms
removing half a GFileInfo model with display-name keys
(no comparisons, just key freeing overhead of a complex sorter)
items keys
512,000 items 14ms 21ms
2,048,000 items 40ms 62ms
removing half a GFileInfo model with complex keys
(no comparisons, just key freeing overhead of a complex sorter)
items keys
512,000 items 90ms 237ms
2,048,000 items 247ms 601ms
GtkSortKeys is an immutable struct that can be used to manage "sort
keys" for items.
Sort keys are memory that is created specifically for sorting. Because
sorting involves lots of comparisons, it's a good idea to prepare the
data relevant for sorting in advance and sort on that data.
In measurements with a PropertyExpression on a string sorter, it's about
??? faster
Instead of one item keeping the item + its position and sorting that
list, keep the items in 1 array and put the positions into a 2nd array.
This is generally slower while sorting, but allows multiple improvements:
1. We can replace items with keys
This allows avoiding multiple slow lookups when using complex
comparisons
2. We can keep multiple position arrays
This allows doing a sorting in the background without actually
emitting items-changed() until the array is completely sorted.
3. The main list tracks the items in the original model
So only a single memmove() is necessary there, while the old version
had to upgrade the position in every item.
Benchmarks:
sorting a model of simple strings
old new
256,000 items 256ms 268ms
512,000 items 569ms 638ms
sorting a model of file trees, directories first, by size
old new
64,000 items 350ms 364ms
128,000 items 667ms 691ms
removing half the model
old new
512,000 items 24ms 15ms
1,024,000 items 49ms 25ms
1. Run step() for a while to avoid very short steps
This way, we batch items-changed() emissions.
2. Track the change region accurately
This way, we can avoid invalidating the whole list if our step just
touched a small part of a huge list.
As this is a merge sort, this is a common occurence when we're buys
merging chunks: The rest of the model outside those chunks isn't
changed.
Note that the tracking is accurate: It determines the minimum change
region in the model.
This will be important, because the testsuite is going to test this.
... and use it in the SortListModel
Setting runs allows declaring already sorted regions so the sort does
not attempt to sort them again.
This massively speeds up partial inserts where we can reuse the sorted
model as a run and only resort the newly inserted parts.
Benchmarks:
appending half the model
qsort timsort
128,000 items 94ms 69ms
256,000 items 202ms 143ms
512,000 items 488ms 328ms
appending 1 item
qsort timsort
8,000 items 1.5ms 0.0ms
16,000 items 3.1ms 0.0ms
...
512,000 items --- 1.8ms
Simply replace the old qsort() call with a timsort() call.
This is ultimately relevant because timsort is a LOT faster in merging
to already sorted lists (think items-chaged adding some items) or
reversing an existing list (think columnview sort order changes).
Benchmarks:
initially sorting the model
qsort timsort
128,000 items 124ms 111ms
256,000 items 264ms 250ms
The model now tracks the original positions on top of just the items so that
it can remove items in an items-changed emission.
It now takes twice as much memory but removes items much faster.
Benchmarks:
Removing 50% of a model:
before after
250,000 items 135ms 10ms
500,000 items 300ms 25ms
Removing 1 item:
4,000 items 2.2ms 0ms
8,000 items 4.6ms 0ms
500,000 items --- 0.01ms
This is the dumbest possible sortmodel using an array:
Just grab all the items, put them in the array, qsort() the array.
Some benchmarks (setting a new model):
125,000 items - old: 549ms
new: 115ms
250,000 items - new: 250ms
This performance can not be kept for simple additions and removals
though.
This is fairly substantial rewrite of the GDK backend for quartz and
renamed to macOS to allow for a greenfield implementation.
Many things have come across from the quartz implementation fairly
intact such as the eventloop integration design and discovery of
event windows from the NSEvent.
However much has been changed to fit in with the new GDK design and
how removal of child GdkWindow have been completely eliminated.
Furthermore, the new GdkPopup allows for regular NSWindow to be used
to provide popovers unlike the previous implementation.
The object design more closely follows the ideal for a GDK backend.
Views have been broken out into subclasses so that we can support
multiple GSK renderer paths such as GL and Cairo (and Metal in the
future). However mixed mode GL and Cairo will not be supported. Currently
only the Cairo renderer has been implemented.
A new frame clock implementation using CVDisplayLink provides more
accurate information about when to draw drawing the next frame. Some
testing will need to be done here to understand the power implications
of this.
This implementation has also gained edge snapping for CSD windows. Some
work was also done to ensure that CSD windows have opaque regions
registered with the display server.
** This is still very much a work-in-progress **
Some outstanding work that needs to be done:
- Finish a GL context for macOS and alternate NSView for GL rendering
(possibly using speciailized CALayer for OpenGL).
- Input rework to ensure that we don't loose remapping of keys that was
dropped from GDK during GTK 4 development.
- Make sure input methods continue to work.
- Drag-n-Drop is still very much a work in progress
- High resolution input scrolling needs various work in GDK to land
first before we can plumb that to NSEvent.
- gtk/ has a number of things based on GDK_WINDOWING_QUARTZ that need
to be updated to use the macOS backend.
But this is good enough to start playing with and breaking things which
is what I'd like to see.
This was preventing any sort of building on macOS, even though the quartz
backend is currently non-functional. Fixing this is a pre-requisite to
getting a new macOS backend compiling.
Run the gdkkeysyms-update.pl script to pick up several
new keysyms:
GDK_dead_lowline
GDK_dead_aboveverticalline
GDK_dead_belowverticalline
GDK_dead_longsolidusoverlay
GDK_Keyboard
GDK_WWAN
GDK_RFKill
GDK_AudioPreset
Changing the selection in the object tree is
not a useful action if we are already in the
object details. Most likely, a user who picks
an object wants to inspect its details, so
just always show them.
Fixes: #1876
Bring back the actions tab; we don't receive
changes anymore, since GtkActionMuxer lost
the GActionGroup signals for this, and the
action observer machinery has no way to listen
for all changes.
Instead of implementing the GActionGroup interface
and using its signals for propagating changes up
and down the muxer hierarchy, use the GtkActionObserver
mechanism. This cuts down on the signal emission
overhead.
We should not rely on GtkWindow to have global
"activate-default" key bindings that happen to
fall back to activating the focus widget. This is
unreliable, since the bubbling up from the button
to the toplevel may run across other widgets that
may want to use Enter for their own purpose, and
then the button loses out. By adding our own
key bindings, the button gets to handle it before
its ancestors.
This fixes check buttons in the inspector property
list not reacting to Enter despite having focus.
If we don't, an ancestor (such a GtkListItemWidget)
may interpret the click as "I should grab focus!",
and still our focus away. This was causing hard-to-focus
entries in the property list in the inspector.
We were hiding the inspector when the window
is closed, but that has the side-effect of
keeping references to application windows,
so we would keep them artificially alive,
which can have side-effects.
So, make the inspector go away when closed.
This fixes the widget factory rendering too much.
In the widget-factory, we generally have a pretty small update area (two
spinners and a progressbar). We take the extents of that as a update
area and inital clip.
However, the first clip node we see is from the toplevel window, which
essentially increases the clip again to almost the entire window.
Fix that by ignoring such cases.
Porting code from GTK 3 without the ability to subclass GtkTreeView
directly can cause an extreme amount of pain on application developers.
It can also complicate performance when it comes to dealing with
encapsulation as the outer widget would also encapsulate the GtkScrollable
implementation from GtkTreeView, typically through GtkViewport.
Fixes#2936
That way, demo windows can be maximized and multiple demos can run at
once.
It's especially useful when using --run because the main window is
invisible then.
* GDK_ARRAY_BY_VALUE
#define this to get GArray-like behavior
* gdk_array_splice (v, 0, 0, NULL, 25)
Adding items but passing NULL as the items will zero() them.
* gdk_array_set_size()
A nicer way to call gdk_array_splice()
* constify getters
This is a scary idea where you #define a bunch of preprocessor values
and then #include "gdkarrayimpl.c" and end up with a dynamic array for
that data type.
See https://en.wikipedia.org/wiki/X_Macro for what's going on.
What are the advantages over using GArray or GPtrArray?
* It's typesafe
Because it works like C++ templates, we can use the actual type of
the object instead of having to use gpointer.
* It's one less indirection
instead of 2 indirections via self->array->data, this array is
embedded, so self->array is the actual data, and just one indirection
away. This is pretty irrelevant in general, but can be very noticable
in tight loops.
* It's all inline
Because the whole API is defined as static inline functions, the
compiler has full access to everything and can (and does) optimize
out unnecessary calls, thereby speeding up some operations quite
significantly, when full optimizations are enabled.
* It has more features
In particular preallocation allows for avoiding malloc() calls, which
can again speed up tight loops a lot.
But there's also splice(), which is very useful when used with
listmodels.
This feature was previously only supported on DBus compositors, such as
Mutter, this adds support for other compositors such as all of those
based on wlroots.
This implementation prefers the idle-inhibit Wayland protocol to the
DBus version if it is available, since the inhibitor is per-surface
instead of global it allows a finer control over which displays get
dimmed for instance. For every case not supported by this protocol, a
fallback to the DBus version is used.
It can’t do anything if the GtkWindow isn’t passed, which might warrant
some documentation change to encourage users to not use NULL for this
argument.
This has been tested on Sway.
Fixes#2202.
This uses the idle-inhibit protocol from wayland-protocols, to attach an
inhibitor to the GdkSurface. The inhibit function can be called as many
times as the user wants, but the uninhibit function MUST be called as
many times to unset the idle inhibition.
This has been tested on Sway.
When this code was ported from gtk_container_get_children
to the dom api, we inadvertendly inverted the order of the
list in one place. With the dom api, we can just avoid
reversing lists altogether, so do that.
Fixes: #2928
Not all compositors support _NET_WM_FRAME_DRAWN. In cases
where the compositor doesn't support _NET_WM_FRAME_DRAWN we don't
need to do all the fancy damage tracking and fence watching.
Furthermore, if the compositor doesn't support _NET_WM_FRAME_DRAWN,
it's possible that one frame will start before the previous frame has
made it through the pipeline, leading to a blown assertion.
This commit side-steps the unnecessary code and associated assertion
when _NET_WM_FRAME_DRAWN isn't supported.
Fixes: https://gitlab.gnome.org/GNOME/gtk/-/issues/2927
The press gesture does not need to set claimed state just for n-press
tracking, it will however reset all other click gestures on each press
if the event gets propagated all the way up.
Fixes tracking of multi-press in gestures in widgets that are children
of the window handle. E.g. the headerbar listview in the "File browser"
gtk4-demo demo.
When the search entry disappears on Escape,
reset the search string to "", so we don't
end up with a filtered list and no obvious
way to remove the filtering.
Instead of an array of arrays, let's use an array of dictionaries; it's
easier to add optional keys without requiring to remember where to put
empty arrays.
The leak sanitizer causes on average 3-4 tests
to segfault during a testsuite run. Disable it
for now to see if we can get a successful
testsuite run with asan alone.
Run our testsuite under asan. We do this in a
separate build, since we need to turn off
introspection to make the build succeed.
As Michael Catanzaro pointed out, this requires a
privileged runner in order to use kernel apis.
When running the testsuite with the address sanitizer,
many of our dependencies cause it to report cause it
to report memory leaks, causing tests to fail.
Therefore, point the leak sanitizer at a list of
suppressions. The list is kept in the lsan.supp
file in git.
Add libasan and libubsan to fedora-base:v20,
and build fedora:20 on top of it.
This is so we can build and run the tests with
the address and undefined behavior sanitizers.
The gtk_gesture_group() call is not a commutative operation, it
takes two gestures, maybe detaches the first one from its current
group, and adds it to the same group than the second gesture.
With the flipped argument order here, GtkRange was actually detaching
the same gesture in order to group it with a second one two times, so
the desired effect to group all 3 gestures was not achieved.
Fixes autoscroll as the drag gesture is now actually grouped with the
click one, so drag offsets can be accessed from the autoscroll
timeout.
If we create an implicit grab on a surface, leave the surface, and
release the button, we would get 2 XI_Leave events, one with mode
XINotifyNormal when the pointer leaves the surface, and another with
mode XINotifyUngrab when the button is released.
Meanwhile, the upper layers rely on crossing events being paired,
and particularly in no crossing event being sent until the implicit
grab is dismissed (either by releasing it, or via more pervasive
grabs).
Ignoring the set of XINotifyNormal events while an implicit grab
is active adapts the X11 backend to this behavior. If the grab were
released or taken away by another grab, a crossing event with one
of the other XINotify*Grab/XINotify*Ungrab will be generated.
Fixes: https://gitlab.gnome.org/GNOME/gtk/-/issues/2879
Check correctly that the captured motion events are emitted towards the
content or one of the scrollbars, in order to have it set the expected
"over" state depending on whether the drag begins from the scrolledwindow
content or one of the scrollbars.
Fixes: https://gitlab.gnome.org/GNOME/gtk/-/issues/2879
This signal does not work on native file choosers,
and it exposes internals of the widget that should
not be public. And it is just not very interesting.
This signal does not work on native file choosers,
and it exposes internals of the widget that should
not be public. And it is just not very interesting.
This signal does not work on native file choosers,
and it exposes internals of the widget that should
not be public. And it is just not very interesting.
Add a GtkDirectoryList:monitored property, and
keep a file monitor if it is set to TRUE. To ensure
that the list reflects reality, we reload the directory
when monitoring is turned on after the fact. This means
that turning monitoring is expensive, while turning it
off is cheap, so we default to monitoring being on.
The stack wasn't updating its visible-child when
the stack pages visible property changes. This
showed up in the inspector, when showing the details
for a list model item.
Make mime-type rules hold multiple types. Store the
content types, so we don't have to do mime-type -> content-type
conversion in the match function. Store content types for the
pixbuf-formats rule as well, so we can avoid memory allocation
in the match function altogether.
char ** arrays are null-terminated everywhere, so make sure they are in
splice(), too.
Also fix the argument to be a const char * const * like in the
constructor.
Simplify all view model APIs and always return G_TYPE_OBJECT as the
item-type for every model.
It turns out nobody uses item-type anyway.
So instead of adding lots of APIs, forcing people to think about it and
trying to figure out how to handle filter or map models that modify item
types, just having an easy life is a better approach.
All the models need to be able to deal with any type of object going
through anyway.
Verify that the selection filter changes mirror
the selection changes of the underlying model,
as expected. These tests verify the fixes in
the previous commit.
When the position is 0, we can't check for unchanged
elements below with gtk_bitset_size_in_range. And
we don't need to, either.
And be careful when translating [start,length]
intervals to [first,last] ones. Off-by-one errors
lurk everywhere.
GtkTextLayout is private now and therefore we can drop all of
the indirection through the class vtable. Instead, just call the
implementations directly and remove the unused vtable entries
for default signal handlers.
Rename the DataList object to TreeData, in preparation
for adding a ListData object for list models. While
we are touching it, modernize it a bit (drop the Private
struct, use a layout manager, etc).
Since commit 972134abe4 a frame getting
drawn has three states (with the vendor nvidia driver at least):
1. drawn by gtk waiting on the GPU
2. drawn by GPU waiting on the compositor
3. drawn by compositor
Those three states are encoded in two flags: frame_pending and
frame_still_painting.
frame_pending means step 1 is done, but step 2 and 3 are still
in progress. frame_still_painting means step 2 is still in progress.
After step 1 is finished the surface is frozen until step 3 is finished.
When the compositor notifies gtk it's done with step 3, with a
_NET_WM_FRAME_DRAWN client message, the toolkit thaws the surface to
allow the next frame to proceed.
The compositor sometimes sends gtk a _NET_WM_FRAME_DRAWN client message
between steps 1 and 2. This message should be ignored because it's not
a reply to the current frame.
Unfortunately, gtk currently assumes if it gets a _NET_WM_FRAME_DRAWN
client message while waiting for step 2 that it's actually at step 3,
and proceeds to draw a new frame while the existing frame is still
pending, leading to a blown assertion.
This commit addresses the problem by ignoring _NET_WM_FRAME_DRAWN
client messages from the compositor unless actually expecting one.
Fixes: #2902
Since commit 972134abe4 we now call
glClientWaitSync for the vendor nvidia driver, to know when a frame
is ready for the compositor to process.
If a surface is hidden while a frame is still being rendered by the GPU,
the surface will never produce the damage event the code relies on to
trigger the call to glClientWaitSync. This leaves the fence dangling,
and the next time the surface is shown, it will start a fresh frame
and blow an assertion since the fence from the last frame is still
hanging around.
This commit ensures a frame gets fully wrapped up before hiding a
surface.
This makes the inspector lock up when used with any production
size list model, and blocks access to properties of the model
itself. Instead, we'll make the model available as an object
and add a data tab for list model contents, like we already
do for tree models.
In particular, it will NULL-ified the current global context if this is
the finalized one, avoiding dangling invalid pointers.
Would have been a cherry-pick from branch gtk-3-24 of commit
b592ded80a, but files moved.
Add a Help item to the gear menu that opens the
node-format.md file in a new window. This could
be improved if we could parse markdown and apply
tags, similar to how we can load pango markup.
One of the widget-factory focus tests is flaky in ci,
perhaps due to font changes causing size computations
to go slightly differently.
Drop this for now.
Include docstrings and format the list of supported
values better.
Also, add the same warning we have for GTK_DEBUG when
the environment variable is ignored.
Even if `gtk_expression_watch()` will do the same, we're calling public
API, so we should perform a check at the point of use, to ensure that
warnings are easily debuggable.
The pkg-config variables have been added in GLib 2.62.0. Let's fallback
to default names for these tools in such case (`pkg-config` still
returns a 0 return value, but with empty output for absent variables).
Some distributions are renaming Glib/GIO utilities for multi-arch
reasons so pkg-config variables have been added to find the correct name
of a tool. GTK+ should use these variables instead of searching in PATH.
See glib#1796.
Use GTK_DEBUG=builder-objects to make GtkBuilder warn
if a named object from a ui files doesn't get claimed
by gtk_builder_get_object(). This is useful for finding
dead wood in .ui files.
Always keep the order:
- [value]
- [marks.top]
- [marks.bottom]
- trough
Which makes sense given the rendering order. Slider should be drawn
after the marks.
Makes it possible to simply remove the custom snapshot implementations
in scale and range. And Adwaita does not depend on the node order
anyway.
Take ordering of cursor_position and selection_bound
into account when copying text to the clipboard, and
ensure that both orders work the same.
Fixes: #2898
We don't want to select on focus-in when the focus
comes from a child. The case where this does harm
is when you activate copy or paste actions from the
context menu. We close the menu before triggering the
action, and if that causes the text in the entry to
be selected, unexpected things happen, since the action
applies to the current selection.
Fixes: #2869
This reverts commit 67c2665028.
The splicing we do here has the important side-effect
of shifting the preedit attributes to the right position.
Without it, we end up always underlining the first chars
in the entry, regardless where the preedit happens.
This makes sure that we do actual key input right
in the middle between all the capture and bubble
event controllers, and are not dependent on the
ordering of those controllers.
The bug that triggered this change was that the
shortcut for activation (Enter) was getting triggered
before the key input, causing Ctrl-Shift-u hex
to stop working, since it never received the enter
to commit the sequence.
The gesture should claim the sequence after triggering uncancellable
actions, like pasting, showing a menu or selecting words/lines. A
single first button press initiating a drag does not trigger
anything yet, so it should avoid claiming the sequence.
The gesture should be accepted whenever it triggers uncancellable
actions in the widget. This means it should be accepted if the
click does result in toggling the switch.
This leaves the pan gesture room to handle dragging the handle.
Fixes: https://gitlab.gnome.org/GNOME/gtk/-/issues/2895
... and do the right things:
nothing: selection = rubberband
ctrl: selection = selection OR rubberband
shift: selection = selection AND (NOT rubberband)
ctrl+shift: selection = selection XOR rubberband
(not sure this one makes sense, but toggling is fun)
Instead of storing the active items as we go, compute the affected items
whenever the rubberband changes and in particular when the rubberband
ends.
That way, the rubberband is guaranteed to select a rectangle even
after scrolling very far.
This is achieved by having a get_items_in_rect() vfunc that selects all
the items in the rubberbanded rectangle and returns them as a bitset.
The rubberband is now handled on the list coordinate system.
When starting the rubberband, we track the item under the pointer and
follow it when it is moving.
This may lead to the rubberband start position changing position and
while this may be confusing, it alerts users to the fact that something
crazy is going on.
In particular, track which items remain in ::items-changed
signal emissions.
But the main use case is sorting, which causes items-changed(0, n, n)
to be emitted.
We only want to send grab-notify to widgets that might have been
interacting with devices via events. Instead of going through all
widgets in all toplevels, we have the window/pointer focus information,
so we can just traverse the widget stacks for every involved foci.
Move away from grab_notify, and use the set_child_focus() vfunc to track
child widgets being set the keyboard focus. This is not 1:1, but seems
good enough at the moment.
Fix various issues that prevented inline completion
and inline selection from working reliably. We were
passing byte counts to gtk_editable_select_region in
one place, but that function expects char counts.
We were listening for GtkEditable::insert-text on
the GtkText widget, but that does not emit those signals,
so listen for GtkEntryBuffer::inserted-text instead.
Finally, we were not clearing the stored completion_prefix
enough, leading to situations where the stored prefix
does not match the text in the entry anymore.
In 99.9% of all cases, these are just NULL, NULL.
So just do away with these arguments, people can
use the setters for the rare cases where they want
the scrolled window to use a different adjustment.
While it's worth thinking about bringing the "windows can be dragged
with open popovers" behavior back, this does not kick in anymore, nor
should be the way to handle this given all the autoclose surface
semantic changes.
This kind of transient state sets the expectative that events update
devices, while it's more accurate to say that devices generate events.
It does not make to expose this function anymore.
This got stuck in ancient times when widgets were windowed, so the devices
in a window to know the devices in that widget would pan out. We do only
want here the devices that are inside the widget, not spread over the
surface, so rewrite this helper function to poke the toplevel foci, and
look they are contained inside the widget.
Crossing events are now detached from widget state, all tricky consequences
from getting multiple crossing events are now somewhat moot. Resort to sending
all generated crossing events, and drop this barely (ever?) used API.
When a gesture (group) claims a sequence, all other gesture groups
in the same widget should get cancelled. Not just previously claimed
ones, that shouldn't happen actually.
This is a list model holding strings, initialized
from a char **. String lists are buildable as well,
and that replaces the buildable support in GktDropDowns.
These sources are using GtkListStore apis,
but were replying on indirect includes to
get the header. Make this explicit, to prepare
for GtkEntryCompletion losing its tree view
dependencies.
Commit a0f6ff101e made sure that a
context was bound before calling glClientWaitSync, but it doesn't
check that the context shares objects with the context that created
the fence.
This commit does a little more validation before deciding the current
context is good enough.
Since commit 972134abe4 we now call
glClientWaitSync for the vendor nvidia driver, to know when a frame
is ready for the compositor to process.
glClientWaitSync can be called regardless of which context is currently
bound, but if no context is bound at all, it returns 0 without
doing anything.
This commit checks for that edge case, and ensures a context gets
made current in the event no context is already current, before calling
glClientWaitSync.
This api has not really been kept up with current
user experiences in popups, and we're better off
just dropping it and letting people do their own
popups if they need custom UI.
Use gtk_widget_prepend_controller to supersede entry keynav
while the popup is open. This fixes selecting completions
with the keyboard - the Enter keypress was ending up
triggering GtkText::activate instead of inserting the
selected completion into the entry.
Add a variant of gtk_widget_add_controller that
inserts the controller at the beginning, instead
of the end. This will be used in entry completion
to make sure the entry completion key event handling
supersedes the entry one while the popup is open.
Keep this private for now, until we determine if
it needs to be public api.
We were adding event controllers at the end, but
announcing a change at the beginning, in
gtk_widget_add_controller. Fix that by emitting
::items-changed for the position where we actually
inserted the controller.
When given a 0 timeout, glClientWaitSync is only supposed to return one
of three possible values:
- GL_ALREADY_SIGNALED - fence fired
- GL_WAIT_FAILED - there was an error
- GL_TIMEOUT_EXPIRED - fence hasn't fired yet
In addition, it can also return GL_CONDITION_SATISFIED if a non-zero
timeout is passed, and the fence fires while waiting on the timeout.
Since commit 972134abe4 we now call
glClientWaitSync (with a 0 timeout), but one user is reporting it's
returning some value that's not one of the above four.
This commit changes the g_assert to a g_error so we can see what
value is getting returned.
May help with https://gitlab.gnome.org/GNOME/gtk/-/issues/2858
According to [1], '_timezone' is already used for a global variable in the
time.h system header that is supplied by Microsoft, so using that for our
variable name when we are including time.h either directly or indirectly
will cause trouble.
This renames such variables to '_tz' to avoid that
[1]: https://docs.microsoft.com/en-us/cpp/c-runtime-library/daylight-dstbias-timezone-and-tzname
Due to an oversight, when multi filters were split into
any and every, any ended up with the listmodel and buildable
implementations, and every didn't get any.
Move the implementations up.
Do it via the new size setting on the color list.
Also add an option to set the maximum size of the color list that it'll
be filled to so rudimentary performance testing is possible.
Hint: Do not select 16777216.
This reinstates commit c16848c2bb683f7ddf51571aa5951f8a4a1ea622.
This showed up as a test failure when we get NULL as
selected item and then try to unref it. Luckily
get_selected_item is transfer none, so we don't have
to worry about it.
gtk_weather_info_new was consuming the info, but not
the timestamp, which confused me. Make it not consume
either, and be more explicit about where the unrefs
happen.
We are currently not robust against model changes or
widget invalidations, so we can actually end up in
situations where we run out of items here. Handle
the failure a bit more gracefully, by returning NULL.
This is good enough to make scrolling work okish most
of the time. We still need a proper fix to handle
other situations.
Use a drop motion controller to autoscroll horizontally
while a drag operation is hovering over the list. The
vertical scrolling is handled by the listview.
Break out an update_autoscroll() function that can
be used for other things than rubberbanding. It will
be used for autoscroll during DND in the future.
All widgets cache their render node already. Just allocate the
last_visible_child always at 0/0 and then move its rendernode around
during snapshot.
Fixes#2678
We already know that a widget will have literally 1 node, not more.
Avoid doing the GtkSnapshot state stack dance and just append a new
transform node instead.
Seems to give me around 400 more icons in the fishbowl
First, almost all widgets are either visible and mapped, or visible and
unmapped because they are !child_visible. In both cases, we do care
about the size. If they are indeed invisible, gtk_widget_measure() will
take care of that by returning all 0.
And gtk_widget_unparent() already queues a resize on the parent if
necessary, so that is unneeded as well.
Juneteenth (a portmanteau of June and nineteenth)[2] (also known as
Freedom Day,[3] Jubilee Day,[4] and Liberation Day,[5]) is an unofficial
American holiday celebrated annually on the 19th of June in the United
States.
-- https://en.wikipedia.org/wiki/Juneteenth
The GtkIMMultiContext uses a delegate pattern to proxy an existing input
method context. Let's not use loaded terms like "slave" when we have
perfectly fungible terms like "delegate".
Libcloudproviders has a single header entry point, so we shouldn't
include a sub-header.
Additionally, the include path provided by the pkg-config file is:
-I${includedir}/cloudproviders
So the include directive should be:
#include <cloudproviders.h>
The fact that it worked until now was an accident caused by the blanket:
-I${includedir}
we get for free; it broke the build when using libcloudproviders as a
subproject.
GtkSingleSelection does not give us a notification when
autoselect kicks in to select the first item. This was
barely noticeable, since the notebook ends up correctly
populated anyway, but the window title is not.
Currently we have to hunt all the `image` keys when updating the image
in use for a job. We should centralise them all at the top of the file,
so they are easy to update in one simple change.
This patch implements the openFiles delegate which is required
to open files which are associated with an application via the
Finder or via open on the command line. The patch has been
proposed by jessevdk@gmail.com.
See: https://gitlab.gnome.org/GNOME/gtk/-/issues/463
I tested the patch with the GNU pspp application on MacOS with
the quartz backend.
Shift-Tabbing was getting stuck in GtkSearchEntry and
GtkPasswordEntry, since they grab the focus to a child
of theirs. Copy the same fix that we are already using
in GtkEntry.
Fixes: #2842
If we leave the valign to be the default fill, then
the images pick up baseline alignment from the environment,
which can occasionally lead to misplaced -/+ icons in
spinbuttons, for example in the "Page Setup" tab of
the print dialog.
Fix scheduling of the frame clock when we don't receive "frame drawn"
messages from the compositor.
If we received "frame drawn" events recently, then the "smooth frame
time" would be in sync with the vsync time. When we don't receive frame
drawn events, the "smooth frame time" is simply incremented by constant
multiples of the refresh interval. In both cases we can use this smooth
time as the basis for scheduling the next clock cycle.
By only using the "smooth frame time" as a basis we also benefit from
more consistent scheduling cadence. If, for example, we got "frame
drawn" events, then didn't receive them for a few frames, we would still
be in sync when we start receiving these events again.
When an animation is started while the application is idle, that often
happens as a result of some external event. This can be an input event,
an expired timer, data arriving over the network etc. The result is that
the first animation clock cycle could be scheduled at some random time,
as opposed to follow up cycles which are usually scheduled right after a
vsync.
Since the frame time we report to the application is correlated to the
time when the frame clock was scheduled to run, this can result in
uneven times reported in the first few animation frames. In order to fix
that, we measure the phase of the first clock cycle - i.e. the offset
between the first cycle and the preceding vsync. Once we start receiving
"frame drawn" signals, the cadence of the frame clock scheduling becomes
tied to the vsync. In order to maintain the regularity of the reported
frame times, we adjust subsequent reported frame times with the
aforementioned phase.
Use better matching format modifiers/specifiers, initialise some things
which in theory wont be written to because of getters using g_return_if_fail(),
a cast, and gsize as input for malloc because gsize!=glong on 64bit Windows.
We lost this when GtkSpinButton was first ported
to the new editable regime, and then the GtkBoxLayout.
Bring it back, but without text measurement, by overriding
width-chars for the GtkText inside, and only do it if
GtkSpinButton::width-chars is unset (ie -1).
Also adjust the documentation slightly to point out
how auto-sizing can be turned off.
GTK 4.0 was currently using GL_EXT_framebuffer_object, which is
deprecated as the ARB version has been merged into OpenGL 3.0 as well as
OpenGL ES 2.0, and provides laxer requirements.
This is a port of !2076 for 4.x.
This gesture handles both individual touch events and touchpad gesture
events, and was checking the touchpad phase in generic code paths. This
is dubious since event methods error out on the wrong GdkEventTypes.
Check the touchpad gesture phase within the branch handling touchpad
events, and make it clear which is the gesture phase of all that we are
ignoring.
Fixes: https://gitlab.gnome.org/GNOME/gtk/-/issues/2825
This makes 'gtk_file_chooser_get_filter' work for the
portal native file chooser by handling the corresponding
'current_filter' argument in the response retrieved via
D-Bus.
In order to try to map the retrieved 'current_filter' to one
of the existing list of filters, use the retrieved filter's name,
similar to how xdg-desktop-portal-gtk does it when evaluating the
'current_filter' input parameter in 'options'.)
Note: This depends on the following merge/pull requests
which fix the filter handling in gtk for native file choosers
and introduce the 'current_filter' handling for FileChooser portal.
* https://gitlab.gnome.org/GNOME/gtk/-/merge_requests/1959
* https://github.com/flatpak/xdg-desktop-portal/pull/493
* https://github.com/flatpak/xdg-desktop-portal-gtk/pull/311
This fixes#1820 for desktop portal case.
Fixes: #1820
'gtk_file_chooser_get_filter' did not work for GtkFileChooserNative,
since the previous way did not properly handle the delegate dialog,
s.a. commit a136cbae8f
("filechoosernative: forward current_filter to delegate dialog",
2018-11-29) for details, wich basiscally fixed the same thing for
the 'gtk_file_chooser_set_filter' case.
This fixes#1820 for the fallback dialog. A solution for the portal
one (which also requires changes to xdg-desktop-portal and
xdg-desktop-portal-gtk as well) will be suggested in a subsequent step.
Bug: #1820
Scroll events can have history too, so make a
getter that works for both. This drops the
gdk_scroll_event_get_history getter that was
added a few commits earlier, since we now
store scroll history in the same way as
motion history.
Update the docs, and all callers.
There is really no need to store 128 doubles as axes,
ever. We can do just fine with 10. At the same time,
add a GdkAxisFlags member, so we can interpret the
values without having to go chasing the right device
for this information.
Only return one accumulated scroll event per frame.
Compress them by adding up the deltas.
Still missing: a way to capture history, like
we do for motion events.
Fixes: #2800
This is not just about consistency with other functions.
It is about avoiding reentrancy problems.
GtkListBase first doing an unselect_all() will then force the
SelectionModel to consider a state where all items are unselected
(and potentially deciding to autoselect one) and then cause a
"selection-changed" emission that unselects all items and potentially
updates all the list item widgets in the GtkListBase to the unselected
state.
After this, GtkListBase selects new items, but to the SelectionModel and
the list item widgets this looks like an enitrely new operation and
there is no way to associate it with the previous state, so the
SelectionModel cannot undo any previous actions it took when
unselecting.
And all listitem widgets will now think they were just selected and
start running animations about selecting.
Replace a previous fix with a more correct one: Update the
selected state from the model instead of reusing the old state, the
model might have updated the selected state.
When translating coordinates from an event, we need to
take the surface transform into account. This makes
double-clicking on editable cells in treeviews work
again.
Fixes: #2831
This is a selection model that stores the selection
state in a boolean property of the items, and thus
persists across reordering and similar changes.
Fixes: #2826
The selection model will only emit one of items-changed
or selection-changed, so when we handle an items-changed,
we must assume that selection state may have changed for
any of the newly added items.
This fixes lingering selection state in the visible range
if you change the sorting of the colors demo.
Fixes: #2827
Move the selection changes to button release, to
avoid conflict with the drag gesture for rubberbanding.
This avoids peeking at the parent, and is generally
nicer.
Among other things, you can now shift-click to select
a range in the colors demo in gtk4-demo.
When I changed things to only collect the set
in the stop() function, I overlooked that this
has the side-effect of only handling items which
are backed by a widget at the time stop() is called.
If we make a big rubberband and autoscroll down too
far, we loose the items that go out the visible range
at the top. Fix that by maintaining the set as we go.
It's finally unused.
Accessible types should either watch properties they are interested in
directly, or should have (private) API to allow widgets to update the
accessible state directly.
Now that we don't have any additional subclasses of GtkEntryAccessible
in GTK, we can drop all the conditional fluff in the base class.
We still need to subscribe to the global notify signal, because of the
sheer amount of properties watched by GtkEntryAccessible.
GtkPasswordEntryAccessible is not a GtkEntryAccessible any more, so it
will need a proper implementation of various interfaces and
functionality in order to work like any other entry.
We're already listening to the adjustment property on the spin button,
there's no need to reset the adjustment on widget set/unset, since the
accessible instance is always tied to the same widget.
Drop the GtkWidgetAccessibleClass.notify_gtk and the
AtkObjectClass.initialize overrides: they don't do anything relevant.
Instead, have GtkProgressBar update the accessible state when the
fraction changes.
Do not use a generic "notify" signal handler.
Additionally, clean up the GtkIconViewAccessible implementation to bring
it up with modern idiomatic GObject.
It feels slightly wrong to have GtkOrientable operate on widgets, but at
least what happens when an orientable widget changes orientation should
be part of GtkWidget.
This will allow to add more state changes without accessing widget state
from the outside of gtkwidget.c.
We expect widgets to use their own derived GtkWidgetAccessible type,
these days, and given that we hard code the default accessible type of a
GtkWidget to GtkWidgetAccessible, and that we enforce the dependency of
the type passed to gtk_widget_class_set_accessible_type(), the registry
code path is clearly unused.
The tooltip handling in GtkWidget is "special":
- the string is stored inside the qdata instead of the private
instance data
- the accessors call g_object_set() and g_object_get(), and the
logic is all inside the property implementation, instead of
being the other way around
- the getters return a copy of the string
- the setters don't really notify all the involved properties
The GtkWidgetAccessible uses the (escaped) tooltip text as a source for
the accessible object description, which means it has to store the
tooltip inside the object qdata, and update its copy at construction and
property notification time.
We can simplify this whole circus by making the tooltip properties (text
and markup) more idiomatic:
- notify all side-effect properties
- return a constant string from the getter
- if tooltip-text is set:
- store the text as is
- escape the markup and store it separately for the markup getter
- if tooltip-markup is set:
- store the markup as is
- parse the markup and store it separately for the text getter
The part of the testtooltips interactive test that checks that the
getters are doing the right thing is now part of the gtk testsuite, so
we ensure we don't regress in behaviour.
When exclusive is TRUE, we would not always emit a
::selection-changed signal that covers all the items
that were unselected.
This commit includes a test.
When exclusive is TRUE, we would not always emit a
::selection-changed signal that covers all the items
that were unselected.
This commit includes a test.
Tell reordered columns to reorder their cells to
the new position. This is necessary to get things
like separators right. The visible symptom of this
problem was the lack of the right border when the
last column is reorder to another position, since
the title widget was still the last in its container,
so :last-child applied.
When we are given a for_size as width for the whole
column view, we need to distribute it over the columns
as gtk_column_view_allocate_columns would, in order
to find out which for_size to give to each cell.
This is a bit recursive, but works. Since we are
doing this recursion for every row, we should consider
adding a cache for those distributed widths.
With the vendor provided Nvidia driver there is a small window of time
after drawing to a GL surface before the updates to that surface
can be used by the compositor.
Drawing is already coordinated with the compositor through the frame
synchronization protocol detailed here:
https://fishsoup.net/misc/wm-spec-synchronization.html
Unfortunately, at the moment, GdkX11Surface tells the compositor the
frame is ready immediately after drawing to the surface, not later,
when it's consumable by the compositor.
This commit defers announcing the frame as ready until it's consumable
by the compositor. It does this by listening for the X server to announce
damage events associated with the frame drawing. It tries to find the
right damage event by waiting until fence placed at buffer swap time
signals.
This commit moves some of the end frame sync counter handling
code to subroutines.
It's a minor readability win, but the main motivation is to
make it easier in a subsequent commit to defer updating the
sync counter until a more appropriate time.
commit 14bf58ec5d dropped support
for using the DAMAGE extension since there was no code that
needed it.
We're going to need it again, however, to address an NVidia
vendor driver issue.
This commit does the plumbing to add it back.
The columnview overrides the width of its content,
so if a label is not willing to wrap its content
below a certain width, it will just get cut off.
Avoid that by not setting width-chars on the wrapping
content.
When the application does not receive "frame drawn" signals we schedule
the clock to run more or less at intervals equal to the last known
refresh interval. In order to minimize clock skew we have to aim for
exact intervals.
We try to step the frame clock in whole refresh_interval steps, but to
avoid drift and rounding issues we additionally try to converge it to
be synced to the physical vblank (actually the time we get the
frame-drawn message from the compositor, but these are tied together).
However, the convergence to vsync only really makes sense if the new
frame_time actually is tied to the vsync. It may very well be that
some other kind of event (say a network or mouse event) triggered
the redraw, and not a vsync presentation.
We used to assume that all frames that are close in time (< 4 frames
apart) were regular and thus tied to the vsync, but there is really no
guarantee of that. Even non regular times could be rapid.
This commit changes the code to only do the convergence-to-real-time
if the cause of the clock cycle was a thaw (i.e. last frame drawn and
animating). Paint cycles for any other kind of reason are always
scheduled an integer number of frames after the last cycle that was
caused by a thaw.
When we get to a paint cycle we now know if this was caused by a
thaw, which typically means last frame was drawn, or some other event.
In the first case the time of the cycle is tied to the vblank in some
sense, and in the others it is essentially random. We can use this
information to compute better frame times. (Will be done in later
commits.)
Visual Studio does not allow decorating functions with '__declspec (dllexport)'
if a prototype exists and is not decorated with '__declspec (dllexport)' as
well, so we cannot just decorate g_io_module_[load|unload|query] in the various
module sources with G_MODULE_EXPORT because the prototypes of these functions
have been marked with _GLIB_EXTERN, which equates to 'extern' unless overridden
Fix this by overriding _GLIB_EXTERN with the appropriate visibility flag, as we
have used to define _GDK_EXTERN. Unfortunately, we can't just use _GDK_EXTERN
G_MODULE_EXPORT as they may have not been defined yet for our use
Do this across the board for all modules, even if they are not buildable on
Visual Studio nor Windows, for consistency's sake.
The included fribidi header is not used in gdkkeys-wayland.c and already
included in gdk.c which causes linker issues due to the header defining
a global variable.
If you add a widget to a parent, this will invalidate the css nodes
for parent/siblings. Afterwards, if the parent is mapped, we will
realize the new child. This calls gtk_widget_update_alpha() which
needs the css opacity, so it revalidates the css.
Thus, for each widget_add (while visible) will trigger a full
revalidation of each sibling. If you add N children to a parent that
leads to O(N^2) revalidations.
To demo this I changed gtk-demo to always double the count
(independent of the fps) and print the time it took. Here is the
results (after a bit):
Setting fishbowl count=256 took 3,4 msec
Setting fishbowl count=512 took 10,1 msec
Setting fishbowl count=1024 took 34,1 msec
Setting fishbowl count=2048 took 126,3 msec
Setting fishbowl count=4096 took 480,3 msec
Setting fishbowl count=8192 took 1892,7 msec
Setting fishbowl count=16384 took 7751,0 msec
Setting fishbowl count=32768 took 38097,7 msec
Setting fishbowl count=65536 took 191987,7 msec
To fix this we drop gtk_widget_update_alpha() and just
calculate it when needed (which is only in a single place).
It was really only necessary because we previously set
the alpha on the surface.
With this fix the above becomes:
Setting fishbowl count=256 took 1,0 msec
Setting fishbowl count=512 took 1,9 msec
Setting fishbowl count=1024 took 3,7 msec
Setting fishbowl count=2048 took 7,4 msec
Setting fishbowl count=4096 took 18,1 msec
Setting fishbowl count=8192 took 31,0 msec
Setting fishbowl count=16384 took 66,3 msec
Setting fishbowl count=32768 took 126,7 msec
Setting fishbowl count=65536 took 244,6 msec
Setting fishbowl count=131072 took 492,2 msec
Setting fishbowl count=262144 took 984,3 msec
The new names are
GtkListView - listview row
GtkGridView - gridview child
GtkColumView - columnview header
columnview listview row
Adwaita css has been updated to preserve
existing styles.
Fixes: #2818
This was done in a weird way where we always call reftest_uninhibit_snapshot()
on paint, and then re-inhibited it if it wasn't inhibited. To make this
work it also started with an extra inhibit.
This is very contorted and based on how this historically worked. This
changes it to just do:
if (inhibit_count > 0)
return;
And keep inhibit_count at its initial zero value unless it is actually
inhibited.
In https://gitlab.gnome.org/GNOME/gtk/-/merge_requests/2027 i was getting
Bail out! ERROR:../testsuite/reftests/reftest-snapshot.c:212:reftest_uninhibit_snapshot: assertion failed: (inhibit_count > 0)
In (for example the box-shadow-changes-modify-clip reftest. I can reproduce this (on master) with:
```
$ xvfb-run -a -s "-screen 0 1024x768x24" meson test --suite gtk:reftest "reftest box-shadow-changes-modify-clip.ui"
...
1/1 gtk:reftest / reftest box-shadow-changes-modify-clip.ui ERROR 0.77s
``
Fix this by re-inhibiting if we didn't draw anything, or we will get an assert the next paint.
gtk_css_node_ensure_style() recurses over previous siblings to ensure
these have a style before its following sibling. As seen in
https://gitlab.gnome.org/GNOME/gtk/-/merge_requests/2027 this can
cause us to stack overflow and crash if we have a lot of children.
And even if we don't have *that* many children its still somewhat
bad to have stack depths of the same magnitude as the number of
children, both for performance reasons and debuggability.
This is a demo that measures performance, so keep the child
info in a hashtable instead of a list. This means adding or removing
a child is not O(n-children).
Rename the show-separators property to show-row-separators,
and add a matching show-column-separators property. It is
implemented by setting the .column-separators style class
on the column view.
This is an expected feature with rubberband selection:
as you get close to the edge while doing rubberband
selection, the list scrolls to extend your selection.
Implement the typical rubberband selection, including
autoscroll. This is only useful with multiselection,
and not very compatible with single-click-activate.
Therefore, it is not enabled by default, and needs
to be turned on explicitly.
Autoscroll when the pointer gets close to the
edge during column resizing or reordering. This
is similar to what the treeview does, but it is
implemented using a tick callback, and has
variable speed.
Allow rearranging columns by dragging, in the same
way the treeview does.
We add the "dnd" style class to the header while
it is dragged, and we move the header of the dragged
column to the end of its parents children, so that
it gets drawn on top.
Add helper functions that let us temporarily give
a different allocation to headers. These will be
used to implement interactive column reordering
in GtkColumnView.
The listview inside always thinks it gets its full size,
and updates its horizontal adjustment accordingly.
So keep our own adjustment, and update it when allocating.
Tweak the behavior slightly. We don't show
a scrollbar as long as we have at least
min-size available, but we still give the
entire size to the child, up to nat-size.
This matches how viewports handle scroll-minimum.
stat/fstat aren't compatible with GStatBuf on Windows in all cases.
Since we don't really need the file descriptor anyway here replace it with
a g_stat call.
This fixes a compiler warning with 64bit mingw.
And lets us remove this scary ifdeffery.
Since it's a type with sub-classes, we need to use GTypeInstance (at the
very least), otherwise we won't be able to address each sub-class as
such.
This is similar to how GskRenderNode and GdkEvent are handled, with the
added difficulty that GtkExpression is meant to be used in properties,
in order to be deserialised by GtkBuilder. This requires adding a
GParamSpec sub-class that we can match on from within GtkBuilder,
alongside some convenience API for storing a GtkExpression inside a
GValue.
We're printing out the file we're testing once we succeed, but it's hard
to know which file caused a failure. Let's add a g_test_message()
directive so we can look in our logs.
The print backends do some complicated dispose handling
where the implementations call gtk_print_backend_destroy().
Our tests (in particular, the templates test) trigger
situations where we use print backends after dispose,
and they can't handle the printers listmodel being
NULL at that time. So just remove the printers in
dispose, keep the empty liststore until finalize.
Scroll events do not have a position, so they shouldn't implement the
GdkEventClass.get_position() virtual function; nor they should have an x
and y fields that never get updated.
When we run the frameclock RUN_FLUSH_IDLE idle before the paint,
then gdk_frame_clock_flush_idle() sets
```
priv->phase = GDK_FRAME_CLOCK_PHASE_BEFORE_PAINT
```
at the end if there is a paint comming.
But, before doing the paint cycle it may handle other X events, and
during that time the phase is set to BEFORE_PAINT. This means that the
current check on whether we're inside a paint is wrong:
```
if (priv->phase != GDK_FRAME_CLOCK_PHASE_NONE &&
priv->phase != GDK_FRAME_CLOCK_PHASE_FLUSH_EVENTS)
return priv->smoothed_frame_time_base;
```
This caused us to sometimes use this smoothed_frame_time_base even
though we previously reported a later value during PHASE_NONE, thus
being non-monotonic.
We can't just additionally check for the BEGIN_PAINT phase though,
becasue if we are in the paint loop actually doing that phase we
should use the time base. Instead we check for `!(BEFORE_PAINT &&
in_paint_idle)`.
There is no agreement that a coverflow widget is
appropriate for GTK 4.
It would be ok as a demo if it could live in gtk-demo,
but that requires us to make GtkListBase public first.
The demo is also somewhat rough and needs more work
to look plausible.
Drop GtkCoverFlow and the related demo for now.
We require a C compiler supporting C99 now. The main purpose of
these fallbacks was for MSVC. From what I can see this is now all supported
by MSVC 2015+ anyway.
The only other change this includes is to replace isnanf() with the
(type infering) C99 isnan() macro, because MSVC doesn't provide isnanf().
This is a somewhat large commit that:
- Adds GtkColumnViewSorter
This is a special-purpose, private sorter implementation which sorts
according to multiple sorters, allowing each individual sorter to be
inverted. This will be used with clickable column view headers.
- Adds a read-only GtkColumnView::sorter property
The GtkColumnView creates a GtkColumnViewSorter at startup that it uses
for this property.
- Adds a writable GtkColumnViewColumn::sorter property
This allows defining per-column sorters. Whenever an application sets a
sorter for a column, the header becomes clickable and whenever
a header is clicked, that column's sorter is prepended to the list of
sorters, unless it is already the first sorter, in which case we invert
its order. No column can be in the list more than once.
Add a mode to GtkListItemWidget that activates on
single click and selects on hover. Make
GtkListItemManager set this on its items
when its own 'property' of the same name is set.
The port is kind of evil, in that it stores either a PangoFontFamily or a
PangoFontFace in the list, depending on if the fontchooser is configured
to select fonts or faces.
It also does not cache the font description anymore, so more calls to
pango_font_describe() may happen.
If both of these issues turn out problematic, the fontchooser would need
to resurrect GtkDelayedFontDescription again and put objects of that
type through the model.
These changes depend on Pango 1.46's introduction of listmodels and
various new getters, so the dependency has been upgraded.
This uses a custom GtkColumnViewTitle widget. So far that widget is
pretty boring, but that will change once we added
resizing, reordering, dnd, sorting, hiding/showing of columns or
whatever UIs we want.
The ColumnView now allocates column widths first and then the individual
rows use the new layout manager which looks at the column allocations to
allocate their children.
It's a GtkListItemWidget subclass that tracks the column it belongs to
and allows the column to track it.
We also use this subclass to implement sizing support so columns share
the same size and get resized in sync.
It's private, no APIs, we don't talk about it. But we will start using
it very soon, so we can do size request caching in columns and avoid
sizegroups...
We only create them in root/unroot (they should be created in
appear/disappear, but that vfunc doesn't exist yet), that way we can
avoid expensive work while the widget isn't used for anything.
This way, we can ensure it's always there when we need it (before the
item gets created) and gone when we don't (if some GC language holds on
to the item after we've destroyed the widget).
Instead of 6 vfuncs, we now have 3 and rely on the factory keeping track
of what it needs to do.
We're doing lots of dancing from one object to another here, but this
will hopefully get simpler with further commits.
This splits GtkListItem into 2 parts:
1. GtkListItem
This is purely a GObject with public API for developers who want to
populate lists. There is no chance to cause conflict with GtkWidget
properties that the list implementation assumed control over and
defines a clear boundary.
2. GtkListItemWidget
The widget part of the listitem. This is not only fully in control of
the list machinery, the machinery can also use different widget
implementations for different list widgets like I inted to for
GtkColumnView.
The demo shows creating ones own listmodel and using it to fill a grid.
I am totally getting the hang of React btw:
500 lines of logic with no UI code and 100 lines of GtkBuilder XML and
I get a sweet UI.
The widget mostly works out of the box, but some tweaking may be
necessary (in particular in the theme) and the gtk-demo changes might
require removing before this is production-ready.
This reverts commit 6a164ab306dad9096bde736c907494c71086d3c4.
The function was awkward and we now have only one caller again, so we
can fold it back into it.
Nothing really changes, because both ListView and GridView still keep
self->item_manager around, but it's set up to point at the base's item
manager.
This way we can slowly move things to GtkListBase that need the item
manager (like trackers).
- Handle anchor as align + top/bottom
This fixes behavior for cells that are higher than the view
- Add gtk_list_view_adjustment_is_flipped()
This should fix RTL handling of horizontal lists
- Fix scrolling
This should make scrolling more reliable, particularly on short lists
that are only a few pages long.
I couldn't come up with a better way to automatically inherit the scope
in the builder list item factory that didn't involve a magic
incantation in the XML file. And I do not want developers to know magic
incantations to do a thing that should pretty much always be done.
This implements all the keybindings from GtkTreeView that can be
supported.
It does not implement expand-all, because supporting that means
causing the TreeListModel to emit lots of create_model vfuncs which in
turn would cause many items-changed signal which in turn would cause
many signal handlers to run which in turn would make "expand-all" very
reentrant, and I'm uneasy about supporting that.
For the mouse, just add a click gesture to the expander icon that toggles
expanded state.
Focus in the listitem now works like this:
1. If any child can take focus, do not ever attempt
to take focus.
2. Otherwise, if this item is selectable or activatable,
allow focusing this widget.
This makes sure every item in a list is focusable for
activation and selection handling, but no useless widgets
get focused and moving focus is as fast as possible.
It's quite a bit faster now, but the code is also a bit more awkward.
Pain points:
- GtkTreeListModel cannot be created in UI files because it needs
a CreateModelFunc.
Using a signal for this doesn't work because autoexpand wants to
expand the model before the signal handler is connected.
- The list item factory usage is still awkward. It's bearable here
because the list items are very simple, but still.
This is a container widget that takes over all the duties of tree
expanding and collapsing.
It has to be a container so it can capture keybindings while focus is
inside the listitem.
So far, this widget does not allow interacting with it, but it shows the
expander arrow in its correct state.
Also, testlistview uses this widget now instead of implementing
expanding itself.
Implement measuring and allocating items - which makes the items appear
when drawing and allows interacting with the items.
However, the gridview still does not allow any user interaction
(including scrolling).
Due to the many different ways to set factories, it makes sense to
expose them as custom objects.
This makes the actual APIs for the list widgets simpler, because they
can just have a regular "factory" property.
As a convenience function, gtk_list_view_new_with_factory() was added
to make this whole approach easy to use from C.
Shift-clicking to extend selections now also works, imitating the
behavior of normal clicking and Windows Explorer (but not treeview):
1. We track the last selected item (normally, not via extend-clicking).
2. When shift-selecting, we modify the range from the last selected item
to this item the same way we modify the regular item when not using
shift:
2a. If Ctrl is not pressed, we select the range and unselect everything
else.
2b. If Ctrl is pressed, we make the range have the same selection state
as the last selected item:
- If the last selected item is selected, select the range.
- If the last selected item is not selected, unselect the range.
Make sure the APIs follow a predictable path:
setup
bind
rebind/update (0-N times)
unbind
teardown
This is the first step towards providing multiple different factories.
... and replace the anchor tracking with a tracker.
Trackers track an item through the list across changes and ensure that
this item (and potentially siblings before/after it) are always backed
by a GtkListItem and that if the item gets removed a replacement gets
chosen.
This is now used for tracking the anchor but can also be used to add
trackers for the cursor later.
Remove a bunch of API from the headers that isn't used anymore and then
refactor code to not call it anymore.
In particular, get rid of GtkListItemManagerChange and replace it with a
GHashTable.
This is implemented by using actions, which are a neat trick to get to
allow the ListItem to call functions on the ListView without actually
needing to be aware of it.
This way, newly displayed rows don't play an unselect animation (text
fading in) when they are unselected, but the row was previously used for
a selected item.
Instead of just destroying all items and then recreating them (or even
hide()ing and then show()ing them again (or even even repositioning
them in the widget tree)), just try to reust them in the order they are.
This works surprisingly well when scrolling and most/all widgets
just moved.
We reorder widgets start to end, so when reusing a list item, we
correctly know the previous sibling for that list item, but not the
next sibling yet. We just know the widget it should ultimately be in
front of.
So we can do a more correct guess of the list item's place in the widget
tree if we think about where to place an item like this.
Actually using this change will come in the next commit.
Previously, we were recreating all widgets every time the list item was
rebound, which caused a lot of extra work every time we scrolled.
Now we keep the widgets around and only set their properties again when
the item changes.
This is the big one.
The listview only allocates 200 rows around the visible row now.
Everything else is kept in ListRow instances with row->widget == NULL.
For rows without a widget, we assign the median height of the child
widgets as the row's height and then do all calculations as if there
were widgets that had requested that height (like setting adjustment
values or reacting to adjustment value changes).
When the view is scrolled, we bind the 200 rows to the new visible area,
so that the part of the listview that can be seen is always allocated.
The anchor is now a tuple of { listitem, align }.
Using the actual list item allows keeping the anchor across changes
in position (ie when lists get resorted) while still being able to fall
back to positions (list items store their position) when an item gets
removed.
The align value is in the range [0..1] and defines where in the visible
area to do the alignment.
0.0 means to align the top of the row with the top of the visible area,
1.0 aligns the bottom of the widget with the visible area and 0.5 keeps
the center of the widget at the center of the visible area.
It works conceptually the same as percentages in CSS background-position
(where the background area and the background image's size are matched
the same way) or CSS transform-origin.
We now don't let the functions create widgets for the item from the
listmodel, instead we hand out a GtkListItem for them to add a widget
to.
GtkListItems are created in advance and can only be filled in by the
binding code by gtk_container_add()ing a widget.
However, they are GObjects, so they can provide properties that the
binding code can make use of - either via notify signals or GBinding.
Also refactor the whole list item management yet again.
Now, list item APIs doesn't have bind/unbind functions anymore, but only
property setters.
The item factory is the only one doing the binding.
As before, the item manager manages when items need to be bound.
Require that items created with the manager get destroyed via the
manager.
To that purpose, renamed create_list_item() to acquire_list_item() and
add a matching release_list_item() function.
This way, the manager can in the future keep track of all items and
cache information about them.
It's all stubs for now, but here's the basic ideas about what
this object is supposed to do:
(1) It's supposed to be handling all the child GtkWidgets that are
used by the listview, so that the listview can concern
itself with how many items it needs and where to put them.
(2) It's meant to do the caching of widgets that are not (currently)
used.
(3) It's meant to track items that remain in the model across
items-changed emissions and just change position.
(2) It's code that can be shared between listview and potential
other widgets like a GridView.
It's also free to assume that the number of items it's supposed to
manage doesn't grow too much, so it's free to use O(N) algorithms.
This is mostly for dealing with proper anchoring and can be used to
check that things don't scroll or that selection and focus handling
properly works.
For comparison purposes, a ListBox is provided next to it.
The thing we're actually doing is create and maintain a widget for every
row. That's it.
Also add a testcase using this. The testcase quickly allocates too many
rows though and then becomes unresponsive though. You have been warned.
Thisis the abstraction I intend to use for creating widgets and binding
them to the item out of the listview.
For now this is a very dumb wrapper around the functions that exist in
the API.
But it leaves the freedom to turn this into public API, make an
interface out of it and most of all write different implementations, in
particular one that uses GtkBuilder.
<lookup>foo</lookup>
is now short for
<lookup>
<constant>foo</constant>
</lookup>
ie it looks up the object with the given name so it can then do a
property lookup with it.
This is the most common operation, so it's a nice shortcut.
A constant without a type is assumed to be an object. This is the most
common case and allows
<constant>foo</constant>
without requiring updates to the type whenever the foo object changes.
The sort of the sortlistmodel is now stable with respect to the original
list model.
That means that if the sorter compares items as equal, the model
will make sure those items keep the order they were in in the original
model.
Or in other words: The model guarantees a total order based on the
item's position in the original model.
We need to keep this data around for changes in future commits where we
make the sorting stable.
An important part of the new data handling is that the unsorted list
needs to always be dealt with before the sorted list - upon creation we
rely on the unsorted iter and upon destruction, the sorted sequence
frees the entry leaving the unsorted sequence pointer invalid.
This change does not do any behavioral changes.
This is an enum that we're gonna use soon and it's worth introducing as a
separate commit.
The intention is to have meaningful names for return values in
comparison functions.
Use a weak ref to invalidate bindings. Make sure that this happens
before creating any watches, so that notifies from the
watched expression about changes will not trigger set_property() calls
during dispose()/finalize().
Invalidating also ensures that the watches aren't removed, which can
trigger warnings if the watches are watching the object itself, and the
weak refs cannot be removed anymore.
Users provide a search filter and an expression that evaluates the items
to a string and then the filter goes and matches those strings to the
search term.
GtkExpressions allow looking up values from objects.
There are a few simple expressions, but the main one is the closure
expression that just calls a user-provided closure.
We were applying the pango version requirements inconsistently,
leading to different pango variables being taking from system
pango vs the subproject at times. Thankfully, meson detects
this and complains, so we can fix it.
We soon want to rely on the list model apis in
pango 1.45. This commit also fixes a mixup where
using pango as a submodule would break the build
when pangoft2 is required.
We want to deliver crossing events to controllers
with scope same-native as long as at least one of
the targets is on the same native. As a new approach,
treat out-of-scope targets like NULL, and deliver
crossing events as long as one of the targets is
not NULL.
We soon want to rely on the list model apis in
pango 1.45. This commit also fixes a mixup where
using pango as a submodule would break the build
when pangoft2 is required.
Without a way to create events, there is no point
in allowing gdk_display_put_event to be used from
the outside. And little good can come out of using
the other apis, so just make them all private.
A call to frame gdk_frame_clock_get_frame_time() outside of the paint
cycle could report an un-error-corrected frame time, and later a
corrected value could be earlier than the previously reported value.
We now always store the latest reported time so we can ensure
monotonicity.
Actually inhibit snapshotting of frames from reftest_inhibit_snapshot.
We were not ignoring the case where inhibit_count > 0, and then disconnected
the callback meaning we only ever got the first snapshot.
In commit c6901a8b, the frame clock reported time was changed from
simply reporting the time we ran the frame clock cycle to reporting a
smoothed value that increased by the frame interval each time it was
called.
However, this change caused some problems, such as:
https://gitlab.gnome.org/GNOME/gtk/-/merge_requests/1415https://gitlab.gnome.org/GNOME/gtk/-/merge_requests/1416https://gitlab.gnome.org/GNOME/gtk/-/merge_requests/1482
I think a lot of this is caused by the fact that we just overwrote the
old frame time with the smoothed, monotonous timestamp, breaking
some things that relied on knowing the actual time something happened.
This is a new approach to doing the smoothing that is more explicit.
The "frame_time" we store is the actual time we ran the update cycle,
and then we separately compute and store the derived smoothed time and
its period, allowing us to easily return a smoothed time at any time
by rounding the time difference to an integer number of frames.
The initial frame_time can be somewhat arbitrary, as it depends on the
first cycle which is not driven by the frame clock. But follow-up
cycles are typically tied to the the compositor sending the drawn
signal. It may happen that the initial frame is exactly in the middle
between two frames where jitter causes us to randomly round in
different directions when rounding to nearest frame. To fix this we
additionally do a quadratic convergence towards the "real" time,
during presentation driven clock cycles (i.e. when the frame times are
small).
On my X11 + nvidia setup gnome-shell doesn't report presentation times.
However it does report refresh rate. We were mostly using this in our
calculation except when computing predicted presentation time, were
it fell back on the default 60Hz.
Iterate the shortcuts we found in order, not in
reverse. Otherwise, we always end up activating
the last_selected one, since it is last in the
list.
This broke in fb9b54d4b2 when a list was
turned into an array.
Make Alt-e the mnemionic for both the Edit menu
and the Select button on page 2. This shows that
mnemonic cycling doesn't currently work, we always
open the menu.
We don't want wrapping labels to cause tooltips to
have excessive height, so we need to set a reasonable
value for width-chars, without forcing short tooltips
into a full line length. Also be careful to respect
preexisting line breaks (we have such examples in
widget factory).
Split the fedora image into fedora-base (containing just
the packages) and fedora (adding the user setup), and add
a fedora-docs image that is adding pandoc on top of the
fedora-base image.
With this, the expand_content_files list has been
completely converted to markdown files. Whats left
in content_files is man pages, and a few special
cases.
Arrange for files named section-foo.md to be turned
into docbook sections, while others get turned into
chapters. This is necessary to allow including such
content in chapters, since chapters in docbook don't
nest.
As part of the conversion, give up on including
sources files from the examples directory, and
instead include the content directly. All include
mechanisms add complications. They were already
complicated with xml, and markdown is not making
things easier.
We already did that for fragments, and if you
make changes to these example sources, you
probably need to revise the surrounding text
anyway.
Use pandoc to convert freestanding markdown files to docbook for
inclusion in the generated docs, and use bits and pieces of
gtk-doc code to continue expanding typical gtk-doc abbreviations.
The new tool for markdown -> docbook is a python script called
gtk-markdown-to-docbook.
The markdown dialect is specified via a list of pandoc extension
in gtk-markdown-to-docbook. It includes header annocations,
definition lists and tables, among other things.
This commit converts the 3 overview chapters (drawing, input
handling and actions) and the migration guide to markdown
syntax. Other files that are still listed in content_files
can be converted later.
This commit adds a pandoc dependency.
Drop apis and code related to appmenus from
both GtkApplication and GtkApplicationWindow.
We still keep the menubar support, since it
is needed for system integration on OS X.
Fixes: #2731
We are using floating point for coordinates
everywhere now, so be consistent here.
This commit also changes the implementation of
gtk_fixed_get_child_position to work with
non-translation child transforms.
gtk_fixed_get_child_position does not work if children
have transforms that are not just 2D translations.
Use gtk_widget_translate_coordinates instead.
We are passing the event to the tooltip handle_event
function at the very end. Unfortunately, the target_widget
may have already died at that point. We prevent that
by taking a ref during propagage_event, but the tooltip
code was outside of that. Keep a ref until the very
end ot prevent crashes.
Trash monitor queries info from gvfsd-trash after each file monitor
change which can be problematic when too many changes happen in
a short time. Let's rate limit the number of queries...
Fixes: #1010
The expected behavior is that we trigger a keyboard-driven
interactive move or resize operation. But that doesn't work
with common compositors like mutter or weston, so lets not
expose non-working menuitems.
It is possible that the target widget is already
unparented at the time that we call the tooltips
handle_event function. Quietly return in that case,
no need to emit a critical.
When displaying accelerators, differentiate keypad
symbols with a 'KP' prefix. Fixing a 17 year old bug.
Update expected output in accelerator tests.
Fixes: #227
We were leaking the builder in the css blendmodes demo,
by creating a ref cycle. This was showing up as
the list entry not going back to upright after
closing the window.
We were leaking the builder in the revealers demo,
by creating a ref cycle. This was showing up as
the list entry not going back to upright after
closing the window.
We were leaking the builder in the cursors demo,
by creating a ref cycle. This was showing up as
the list entry not going back to upright after
closing the window.
We were leaking the builder in the builder demo,
by creating a ref cycle. This was showing up as
the list entry not going back to upright after
closing the window.
Get the native transform only once, for all overlays. Unfortunately we
have to undo this for the updates overlay since that one gets values
in surface coordinates.
Don't call into the backends when the input region
or shadow width don't actually change. This avoid
distracting calls in debug logs, and just generally
is the right thing to do.
This is not used anymore now that surfaces are always toplevel in the
semantics of GdkWindow where child windows were available. We can drop
that and simplify the vfunc just a bit more.
Fixes#2765
Provide the minimal info necessary. Improves apparent responsiveness
(since we don't visibly clear and repopulate the list) and saves doing
file stat/reads on every file in the result set.
The filechooser tries to figure out whether it got results by poking
the model, but all files might go through the async GFileInfo querying
state.
Make all search engines (and the composite one) just notify about this
fact, so the file chooser can behave appropriately without waiting for
the async operations to finish.
On X11, shortcuts inhibition is emulated using a grab on the keyboard.
So if another widget ungrabs the keyboard behind our back (for example
when a popup window is dismissed) that effectively disables the effects
of the shortcut inhibition on the surface and we need to update the
shortcut inhibition status accordingly.
Check for "grab-broken" events on the surface and clear existing
shortcuts inhibition for the matching seat, so that the client can be
notified and may decide to re-enable shortcut inhibition if desired.
We pass the GdkEvent as a pointer, because the autogenerated marshallers
don't know how to handle GTypeInstance-derived classes.
Since the GValue box that we use in the marshaller passes the GdkEvent
instance as is, we also need to acquire a reference before invoking the
closure, and release it afterwards, to ensure that the GdkEvent instance
survices the invocation.
There are a few more places where we were forgetting
to apply the surface->native transform. With these
changes, tooltips are positioned correctly when
the toplevel has padding applied.
Fixes: #1619
Just because we take a ref on a surface does not
guarantee that it is still usable a second later.
Check if its been destroyed in the meantime.
This is breaking the template tests in ci, since
there is no client behind the Broadway server.
The assumption is that the source device in events
is a slave device, so create pointer and keyboard
devices and use them in events.
This fixes the seat test on Broadway.
Respect that cairo won't create image surfaces larger
than 32767 x 32767.
This makes the one reftest pass that specifically checks
this condition, treeview-crash-too-wide.
The test setup mechanism in meson is not flexible
enough to let us run different suites depending
on setup, so just pass in explicitly which suites
we want to skip, depending on the backend.
If we don't destroy the surface, it leaks.
GDK backends keep an extra reference on the
surface for the external resources associated
with it, and only drop it in destroy().
GDK backends are expected to keep a references on
their surfaces as long as they are associated with
external resources, and drop it in destroy().
This showed up as criticals in the shortcuts test
which manually creates and destroys surfaces.
We want the test names in the junit xml to be
unique across all the tests in a job, so we need
to include the backend in the test name.
And we also want to see the used backend in
the html report.
If the tablet gets removed/freed while there are pad events in flight,
we leave a dangling pointer from the pad to the tablet, which may
lead to invalid reads/writes when handling the pad event(s).
Add test setups that set the GDK_BACKEND and
TEST_OUTPUT_SUBDIR environment variables.
This lets use run
meson test --setup x11 --suite reftest
meson test --setup wayland --suite reftest
and the output will be nicely separated.
We still need to do compositor / display server
setup from the outside.
meson seems somewhat weak when it comes to handling
test output. We need to get the output from different
test runs into different locations, and the only
way to communicate from a test setup with the actual
test code seems the environment, so use that.
Make all tests that produce output in files respect
a TEST_OUTPUT_SUBDIR environment variable which specifies
the name of a subdirectory to use. This is combined
with the existing --output argument, which specifies
a per-test location.
Affected tests are reftests, css performance tests
and gsk compare tests.
If you run weston with the headless backend, you get a Wayland
display with no seat, which is just fine by the protocol.
gdk_display_get_default_seat() returns NULL in this case. Various
widgets assume that we always have a seat with a keyboard and a
pointer, since that is what X guarantees. Make things survive
without that, so we can run the testsuite under a headless
Wayland compositor.
Once upon a time, there was a function called gdk_event_get_scroll_deltas().
It returned %TRUE when an event had scroll deltas and that was used as the
condition to decide whether to push scroll deltas to the scroll history,
even when the both deltas are 0 for the stop event at the end of scrolling.
When GtkScrolledWindow kinetic scrolling code was adapted for
GtkEventControllerScroll, it was replaced with a (dx != 0 && dy != 0)
check. This prevented the stop event from getting into the history, and
instead allowed non-smooth scrolling to affect the history as they have
synthetic deltas with one of the values being -1 or 1 and the other on 0.
Instead, check the direction as we already have it as a local variable.
This needs to return a boolean, also it should not call end because
that will be called anyway by Gtk+ after cancel, and this was causing
warnings due to the opacity being unset with no dragged widget set.
In the gtk-demo drag-and-drop demo i can't drag anything, all I get
is:
(gtk4-demo:358993): Gdk-CRITICAL **: 09:36:19.617: Surface 0x7e1bb0 has not been mapped in GdkSeatGrabPrepareFunc
This is because GdkX11Drag.ipc_surface is not considered mapped, even
though we called gdk_x11_surface_show() on it, because the
GDK_SURFACE_STATE_WITHDRAWN flag is still set.
I added calls to gdk_synthesize_surface_state() to match what
e.g. show_popup() and gdk_x11_toplevel_present() does.
The flowbox demo is otherwise less than useful,
if /usr/share/gnome/backgrounds isn't present.
At the same time, give the scale in the listbox
some function.
The api contract for size_allocate() vfuncs is
that they must allocate all the children that are
going to be snapshotted in snapshot(). The flowbox
size_allocate() was just bailing out when the children
request a size of 0x0, leading to an assertion in
snapshot() vfunc later. Just allocate all children
a size of 0x0 in this case.
The previous fix broke the case where we're Shift-Tabbing
from a listboxrow child to the row itself. This was causing
the widget-factory2.tab-backward test to fail. Fix it, by
grabbing the focus to the row explicitly.
If a row has content that is focus-on-click, and is set
to focus-on-click itself, then the row steals the focus
fromt he content, since it uses focus-on-click on button
release, as opposed to button press. Avoid that by
refusing to take focus if it is already on some
descendent of the row.
This was showing up in the widget-factory listbox on
page 2, where clicking on the spinbutton would briefly
put the focus on the spinbutton, only to lose it to
the row.
Change which rows in the listbox on page 2 are activatable,
and trigger a dialog. It did not really make sense that this
would happen when clicking on the spin button to focus it.
This was not working in the case that the existing child
is not a scrollable. It showed up as crashes of the
scrolling benchmark in gtk4-demo when switching examples.
The previous code was unreffing the window twice, which caused problems
during dialog destruction. Move to g_list_store_find instead of
iterating manually.
ref() the window before and unref() after. g_list_store_remove will
actually unref() the window, since the toplevel_list owns its own
reference.
Fixes#2741Fixes#2742
If the inner clip intersects with the corners of the outer clip, we
potentially need a texture. We should add more fine-grained checks for
this in the future though.
Test case included.
I thought I could get away with just unrealizing the
window, but it turns out that gtk_window_hide() is the
place where we remove grabs when a modal dialog goes
away, so we ended up with stuck grabs.
Nested main loops are bad, as they introduce layers of complexity caused
by the potential re-entrancy in the case of multiple event sources, like
IPC, threads, etc. Additionally, the programming model they provide—stop
the world while spinning a new loop—does not conform to the event-driven
model employed by GTK.
Replace it with an explicit nested main loop, as we need to block the
signal handler currently being emitted depending on the response of the
overwrite confirmation dialog.
This is a bit of a hack, and the only reason we need it is that the
print dialog will load the last used path as the output file name, when
printing to a file; this means that, in theory, it would be possible to
press Print without selecting a file, and accidentally overwriting an
existing file.
It would be much simpler if we did not store the last used path, and
always explicitly asked the user to select a file; this would avoid
destructive actions, and would allow us to rely on the overwrite
confirmation dialog right inside the file chooser.
Either use the "response" signal for dialogs that are already modal, or
use an explicit nested loop for tests that rely on the response id being
available in sequence.
Nested main loops are bad, as they introduce layers of complexity caused
by the potential re-entrancy in the case of multiple event sources, like
IPC, threads, etc. Additionally, the programming model they provide—stop
the world while spinning a new loop—does not conform to the event-driven
model employed by GTK.
Let keyboard/pointer paths handle their own events, and find the
current focus. The event will be propagated through instead of
being just emitted on the toplevel.
This makes it handled throughout all the gestures that want to
know about it.
Tracking of those broke sometime along the gdk cleanups, so we
started missing some GDK_GRAB_BROKEN events from being emitted
(eg. after a button press/implicit grab triggers an active grab).
Implicit grabs are only added if there's no prior grab (either
implicit through other button presses, or explicit), in order to
keep accounting correct, make those prevail.
It is unreliable to use the widget dom api to locate
action widgets. For example in a headerbar, they might
be deeper in the hierarchy, with boxes in between.
Therefore, make GtkDialog keep a list of action widgets,
and use that when operating on action widgets.
When GtkBox stops being a container, GtkShortcutsSection
will also no longer be a container. So, stop overriding
container vfuncs, and instead add a buildable implementation
that does the right thing.
When GtkBox stops being a container, GtkShortcutsGroup
will also no longer be a container. So, stop overriding
container vfuncs, and instead add a buildable implementation
that does the right thing.
When moving from gtk_container_forall to the widget dom
api, we are now iterating over all children of the listbox,
including headers, separators, etc. So, skip everything
that is not a listboxrow, to make the tests work again.
This test was relying on gtk_container_forall returning
the visual (ie sorted) order of children, while iterating
with the widget dom api gives the insertion order.
Instead of using gtk_container_forall, use
gtk_list_box_row_get_index to reconstruct the visual
order.
This commit is porting GtkPaned to be derived
from GtkWidget instead of GtkContainer, while adding
start-child and end-child properties. The existing
properties are renamed to follow the start/end naming
scheme, and we add proper getters and setters.
Update all users.
See #2719
We want to remove GtkBin and GtkContainer as they don't
provide much useful functionality anymore. This requires
us to move get_request_mode and compute_expand down.
See #2719
We were stepping on our own toes, by first setting
up a save entry and telling the filechooserwidget
about it, and then nuking it by setting a title
on the headerbar. The filechooserwidget wasn't
ready for the entry to die without anybody telling
it.
This fixes a crash when using the filechooser for
print-to-file in the print dialog.
To avoid making this mistake again, add a static assertion that the
enum is in sync with gtk_license_info, and use the length of
gtk_license_info for the precondition check.
Signed-off-by: Simon McVittie <smcv@debian.org>
Resolves: #2734
Don't call gtk_widget_destroy; instead implement
gtk_window_destroy outselves by removing the window
from the toplevel lista and dropping the reference that
GTK holds.
gtk_window_set_transient_for does not ref its parent either. This is
important because a child widget of the parent might be the one calling
this function.
This was showing up as widget-factory not existing on close after
opening the file chooser.
It is a bad idea to have such cycles in the first place,
and there is no need to let them linger past unroot.
This stop the treeview depending on run_dispose to get
freed, a simple unparent will work now.
Redo this series of examples from 2013, and adapt it to modern
way of doing things. The biggest differences are that we use
a headerbar right from the start, and don't mention the app
menu.
Fixes: #2730
One of the treeview tests was calling gtk_widget_destroy
on a child instead of the toplevel, which leaks the toplevel
unnecessarily. Plus, we're moving towards allowing destroy
only on toplevels.
The css tree and the widget tree are not in sync, so we need
to explicitly set the parent of the css node before inserting
the widget, or else we end up with critical warnings and a
non-working popover.
This can be seen in the print dialog, when moving the focus
to the printer list.
Conditionally check whether the Vulkan headers version defines
VK_RESULT_RANGE_SIZE, and avoid using it for version >=140. The
following comming in Vulkan-Headers has removed the enum value:
0c5351f5e9 (diff-4febd94c0666d59030d8b1dd20c72403)
Add back a property that determines whether an individual
widget will accept focus or not. :can-focus prevents the
focus from ever entering the entire widget hierarchy
below a widget, and :focusable just determines if grabbing
the focus to the widget itself will succeed.
See #2686
Add template tests that show the complex dialogs before
destroying them. This reveals that we are leaking in
several of them. These leaks don't show up if the
dialogs are destroyed right away, as the existing
tests do.
Disable the two failing tests for now:
/template/GtkFileChooserDialog/show
/template/GtkPrintUnixDialog/show
In the presence of attached children, the css tree and the
widget tree are not in sync, so we need to explicitly set
the parent of the css node before inserting the widget, or
else we end up with critical warnings and a non-working
menu.
This can be seen in testtextview.
If we use "text" for the children that are attached to the
text view, they end up rendering the same background as the
main text content, causing that content to be covered up.
Fixes: #2729
Making them GObjects is unnecessary. This enables further optimizations
down the road. The only place we use them in is gtkcssanimatedstyle.c
after all.
It turns out that we have a ref leak at the very
core of our dom model :( gtk_widget_insert_before/after
leak a reference if the widget was already under
the same parent. This is something that GtkBox
frequently does. It shows up e.g. when packing
widgets at the end in a headerbar.
Since GtkWindowHandle and GtkHeaderBar do it now, it can be removed from
GtkWindow, along with GTK_WINDOW_REGION_TITLE which at this point doesn't
differ from GTK_WINDOW_REGION_CONTENT.
Closes https://gitlab.gnome.org/GNOME/gtk/-/issues/2689, since
GtkWindowHandle doesn't do that anymore.
Do also the async file info query for remote files when they
are not bookmarked, because otherwise "None" will be shown as
file name (and fallback text generic icon will be used).
The remote file was already browsed by the file chooser
instance when selecting it, so querying the display name
again should be using gio cache and not be slow.
Even if it's were slow it's better than showing 'None'
which makes it seem as if nothing was selected.
Fixes#1966
2020-04-19 16:59:33 -04:00
2227 changed files with 274548 additions and 164162 deletions
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
