It turns out that the previous handling of just providing a way to set
visuals just on toplevels was not sufficient. In particular it
complicated the various implementations of the tray icon specification.
This patch reintroduces gtk_widget_set_visual() which behaves very
similar to GTK2's gtk_widget_set_colormap().
A future commit will remove the gtk_window_set_visual() function.
When fitting a widget into its allocation, the second dimension
is always dependent on the first, so gtk_widget_get_preferred_size()
cannot be used directly (because we want the natural height for
the allocated width, not the natural height for the natural width,
which is generally a smaller height than the height-for-minimum-width
or height-for-allocated-width).
Added test to testadjustsize to ensure proper behaviour.
The function reverses the transform that GTK does before emitting a draw
event. So we can use it in "old" widgets to revert the coordinate system
properly.
It doesn't make sense to keep them separate as GtkSizeRequest requires a
GtkWidget and GtkWidget implements GtkSizeRequest, so you can never have
one without the other.
It also makes the code a lot easier because no casts are required when
calling functions.
Also, the names would translate to gtk_widget_get_width() and people
agreed that this would be a too generic name, so a "preferred" was added
to the names.
So this patch moves the functions:
gtk_size_request_get_request_mode() => gtk_widget_get_request_mode()
gtk_size_request_get_width() => gtk_widget_get_preferred_width()
gtk_size_request_get_height() => gtk_widget_get_preferred_height()
gtk_size_request_get_size() => gtk_widget_get_preferred_size()
gtk_size_request_get_width_for_height() =>
gtk_widget_get_preferred_width_for_height()
gtk_size_request_get_height_for_width() =>
gtk_widget_get_preferred_height_for_width()
... and moves the corresponding vfuncs to the GtkWidgetClass.
The patch also renames the implementations of the vfuncs in widgets to
include the word "preferrred".
Previously, we tried to move the context's origin to the widget's top
left location, no matter what window the paint was happening on. Now we
only do that for child windows of the widget's window and leave the
context untouched for windows that the widget has created outside its
own hierarchy. In those casses, we also don't clip the context to
the widget's allocation.
Includes fixes to GtkHandlebox for this effect and fixes all known
issues with it.
These semantics assume that gtk_widget_draw() should only draw the parts
of a widget that are inside child windows and not draw stuff that is
located in completely different GdkWindows. In the handlebox case, it
means that it should only draw the handle when it is attached, but not
when it isn't. We'll likely need a special draw function for the
detached handlebox if we want to draw it.
I've seen (and written) quite some widgets (and theme engines) that use
cairo_paint() to draw the background. So avoiding overdraw makes sense.
Also move all that setup into a _gtk_widget_draw_internal() function
that will be used by all functions that can be used by other functions
that draw widgets.
For now, the draw signal is emitted by default from the expose event, so
widgets can chose to either implement the expose event or the draw
function.
This is for the transitional period from expose event to draw signal.
Note that for now subclasses can only implement the draw function when
none of their parents implemented the expose event.
The functions are gtk_widget_get_allocated_width() and
gtk_widget_get_allocated_height().
They are currently identical to using width/height of
gtk_widget_get_allocation(), but are introduced for ease of use (no need
to use a custom struct) and to free people from having to think about
allocation.x and allocation.y (which is where the origin of the cairo
context in GtkWidget::draw is located).
This removes:
gtk_widget_set_default_colormap()
gtk_widget_get_default_colormap()
gtk_widget_get_default_visual()
Colormaps are gone, and the default visual is the system visual of the
default screen.
Add gtk_window_set_visual() and a "visual" property. This allows
changing the window visual to the rgba one and other awesome things
(like implementing the trayicon spec).
We now return the visual of the topmost widget in widget's stack that
has a window. If no such widget exists, but a GtkWindow is a parent, we
return its visual (note: GtkWindow Will gain support for setting visuals
soon). If a window doesn't exist, we return the system visual of the
default screen.
This change has multiple reasons:
- Colormaps are gone
Now visuals are the most important resource for creating GDK windows.
- Allow widgets to change visuals for themselves and their children
By walking the hierarchy, we ensure that child windows have the same
visual as their parents by default. But widgets can still select a
different visual in their realize handler when creating the GDK
window.
- Have a replacement for gtk_widget_set_colormap()
That function is going to die with colormaps, so a replacement was
needed. That replacement is going to be gdk_window_set_visual().
- Make a future transition to no-window GTK easy
Should we ever attempt a change to make all GTK widgets no-window, a
gtk_widget_set_visual() would be silly, as only widgets with windows
can have their own visuals. So only toplevels will gain the ability to
change it.
Replace gtk_widget_shape_combine_region() with
gtk_widget_shape_combine_mask() and
gtk_widget_input_shape_combine_region() with
gtk_widget_input_shape_combine_mask().
As GdkBitmap is going away, and the region equivalents already exist,
this seems like pretty much the default step to take.
Includes code to fix up the users.
h-align = START,END,CENTER,FILL
v-align = START,END,CENTER,FILL
margin-left,right,top,bottom
margin
These should obsolete all such similar properties on
layout containers, GtkMisc, GtkAlignment, GtkContainer::border-width
Margin is outside the size request.
If margin were not outside the set_size_request() it would not work the
same way as container-supplied (child property) padding.
Conceptually set_size_request() forces the value from the subclass
(the original unadjusted request) and then we go on to adjust
the request further by adding the margin.
Use these new methods to handle set_size_request (aka aux_info)
inside gtkwidget.c, instead of having external code mess with it.
The virtual functions can be used for other purposes in the
future. For example, GtkContainer::border_width could be
automatically implemented for all container subclasses.
Did not update uses in other files because the plan is to
get rid of those other uses anyhow. So don't want to make
this function available in the header.
This is better than peeking aux info then testing != NULL
in several ways:
- it returns const aux info so if we don't create we can't write
- it ensures that the default we assume if aux_info is NULL is
the same as the default we set if we've created the aux info
- it avoids typing in != NULL checks
In GtkBin and GtkWidget we tried to provide handy defaults that
call get_width if there's no get_width_for_height and
get_height for get_height_for_width.
However, they used the wrapper API on GtkSizeRequest instead of
chaining directly to the other method implementation.
This could result in all kinds of surprising behavior, for example,
get_width_for_height() would now already include the effects of set_size_request().
If nothing else it's inefficient. But it's just conceptually wrong,
because to chain to another implementation, we should call the other
implementation, not call a wrapper around the other implementation
(when we're already inside a previous invocation of the wrapper,
i.e. compute_size_for_orientation() ends up reinvoking itself
in the same orientation on the same object which it pretty
likely isn't intending to do)
https://bugzilla.gnome.org/show_bug.cgi?id=628829
gtk_widget_set_window() does not add a reference to the
owned window... this is an exception to most gtk_foo_set_bar()
functions where as it acts as if it were named gtk_widget_take_window().
Adding note to the docs to avoid people giving the window to the
widget and naturally proceeding to unref the window.
Now that we have a private data installed directly on
the GtkWidget instance it makes no sense to cache the size
requests on widget qdata. This change will generally make
GTK+ memory less fragmented as well as significantly speed
up the size request process.
The keysyms create a lot of potential namespace conflicts for
C, and are especially problematic for introspection, where we take
constants into the namespace, so GDK_Display conflicts with GdkDisplay.
For C application compatiblity, add gdkkeysyms-compat.h which uses
the old names.
Just one user in GTK+ continues to use gdkkeysyms-compat.h, which is
the gtkimcontextsimple.c, since porting that requires porting more
custom Perl code.
