For embedded images the path is empty but we unconditionally used it to create a TextureCacheKey,
which clashes.
Instead, preserve and store the ImageCacheKey in the ImageInner variants.
This includes the cache of decoded images, the HTMLImage element support
and the SVG rendering adapter.
The objective is that Image holds an ImageInner, which is not a path
anymore that the backend has to process, but instead always either
decoded image data, a pointer to a static texture or an SVG tree that
can be rendered to the desired size.
```
warning: unused return value of `alloc::boxed::Box::<T>::from_raw` that must be used
Warning: --> internal/core/properties.rs:382:9
|
382 | Box::from_raw(_self as *mut BindingHolder<B>);
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
= note: call `drop(from_raw(ptr))` if you intend to drop the `Box`
```
Just do what the note says
The as_any() method returning a `&mut dyn Any` meant that the type has to be
'static. The class solution of returning &()
doesn't work because it has to be a mutable ref. Therefore just return an Option.
Now that we store the layer textures as Rc<Texture>, before rendering a
layer we can take a clone of a possibly existing layer texture and if if
the layer's tracker is dirty, we can re-use the existing layer texture
if it's the same size.
This avoids unnecessary texture allocations.
... via two ways bindings
Otherwise we could be in a case where we have an unbounded number of "shared" property
which might not be shared anymore if the original property goes out of scope
Add flags that enable the Button to be used as a Toggle, e.g. for use in toolbars or similar places.
Co-authored-by: Simon Hausmann <hausmann@gmail.com>
I want to track component structure changes in the window without
generating more code. So use a more generic name for the init_*_items
functions, so that I can add the functionality I need in there.
Also add a register_component to PlatformWindow and call that.
I want a more generic name as I want to do to track component structure
changes in addition to resource freeing and I do not want to add another
call into the generated code.
The borrow logic in ItemCache<T>::get_or_update_cache_entry requires us
to take the per-item cache dirty tracker out of the cache entry, so that
we can drop the ref and allow the update_fn to access the cache for
other items during evaluation.
However in order for the cache to work by not re-evaluating the
update_fn next time if nothing rendering related change, we must put the
tracker back after the check/evaluation, otherwise take() will always
return none.
This is a regression from commit
6cbf2c0609 that changed the call from
get_or_update() - which puts the tracker back - to
get_or_update_cache_entry(), which doesn't. Since the former is now
unused, this commit also removes it.
This ensures that the partial renderer updates correctly and this also
fixes a panic in the printer demo on the stm32h7 when opening the format
combobox in the scan page:
We would attempt to render the word "JPEG" with a font size of 12 pixels.
The letter J has a height of 12 pixels, like the font ascender reported, but it's
y coordinate is negative.
That would also result in a negative text item relative origin for the glyph (target
rect in physical pixels),
which fails to convert to an usize and panics.
When concatenating two fragment, we should account for the trailing whitespace
in the previous fragment in the total width.
This fixes the "l" of "Ink level" not beoing drawn properly in the printer
demo with the partial renderer if the Text item (which is too small as a result
of this bug) does not overlap the dirty region, but part of the text still
need to be drawn outside before it overflows
When computing the dirty regions, the call to geometry() might end up
querying a value that's computed by the layout. For the layout
computation we might end up calling ensure_updated() on a repeater,
which then might destroy components. The destruction involves freeing
cached dirty rectangles (via free_graphics_resources), and that also
requires a write access to the cache. So after retrieving the property
tracker for the dirty geometry, drop access to the cache before calling
geometry().
When using repeaters - like in the slide puzzle - and during renderer a component
gets deleted, we call free_graphics_resources and try to free
the dirty rectangle list in the partial renderer cache. At that point the cache is
already mutably borrowed, which causes a panic.
As remedy, apply the mutable borrow more fine grained and not right when calling
render().
Implement basic accessibility (a11y) support, using the Qt backend.
_This should get us started, but accessibility support is an additional way to interact with UIs that is very different from the "graphical way" most users will interact with the UI. No single PR will "make a toolkit accessibility", this needs to be an ongoing effort!_
Parts of this PR:
* Add functions to access a11y-related properties to Component
* Add helper functions to Item struct
* Handle accessible- properties in the compiler
* Add documentation, add description, enforce some basic rules
* Make the Text element accessible by default
* Don't optimize away accessibility property in the LLR
* Ensure that accessibility property are marked as used
* Add some accessibility properties to the native style widgets
* Support for bool and integer `accessible` properties
* Implement basic support for accessibility
* Make basic widgets accessible by default
* Make slider focus-able and interactable with keyboard
* Tell a11y layer about value changes
* Generate QAccessible constants using bindgen
* Don't expose the `accessible` properties when using the MCU backend: There is no backend to make use of them
* Handle focus change based on keyboard focus of the window
* Report accessible widgets at correct positions
* Allow for (virtual) focus delegation at the a11y level
* Calculate value step size dynamically
* Make sure to not send notifications to a11y backend about dead objects
