[Area::compare_order()](ee4ab08c8a/crates/egui/src/memory/mod.rs (L1174-L1183))
is not a total ordering. If three layers A, B, and C have the same
`order` field but only A and B are present in `order_map`, then `A==C`
and `B==C` but `A!=C`. This can cause a panic in the stdlib sort
function, and does in [my
app](https://github.com/HactarCE/Hyperspeedcube/tree/v2.0) although it's
very difficult to reproduce.
* [x] I have self-reviewed this PR and run `./scripts/check.sh`
* [x] I have followed the instructions in the PR template
Current WebP loader assumes all WebP images to be RGBA, which is the
case if the image is animated (that's what `image` crate assumes at
least). Static images can instead choose to exclude its alpha channel,
though it seems to be more of a default choice to include it, even if
it's not being utilized. Currently, loading a static RGB WebP image will
cause a panic when `ColorImage::from_rgba_unmultiplied` gets called in
the loader
```
thread 'main' panicked at /home/aely/.cargo/git/checkouts/egui-226fc7cdd51201c1/f87219d/crates/epaint/src/image.rs:97:9:
assertion `left == right` failed
left: 29184
right: 21888
```
Previously, `Harness::run` just called `Harness::step` 3 times. If that
wasn't enough, tests would often call run multiple times so all
animations would finish properly.
Also, I introduced `HarnessBuilder::with_step_dt` to customize with how
big of a dt each frame is called. I set the default to 1.0 / 6.0 (~6fps)
so we don't waste cpu in tests waiting on animations.
`HarnessBuilder::max_steps` allows us to control how many steps
`Harness::run` should run before panicing.
The default is 6, so we run for up to 1.0 logical seconds (six frames at
6 fps), which should be enough to finish most animations.
Turns out a lot of snapshots where rendered before fully shown and had a
light opacity, those are now fixed.
* [x] I have followed the instructions in the PR template
<!--
Please read the "Making a PR" section of
[`CONTRIBUTING.md`](https://github.com/emilk/egui/blob/master/CONTRIBUTING.md)
before opening a Pull Request!
* Keep your PR:s small and focused.
* The PR title is what ends up in the changelog, so make it descriptive!
* If applicable, add a screenshot or gif.
* If it is a non-trivial addition, consider adding a demo for it to
`egui_demo_lib`, or a new example.
* Do NOT open PR:s from your `master` branch, as that makes it hard for
maintainers to test and add commits to your PR.
* Remember to run `cargo fmt` and `cargo clippy`.
* Open the PR as a draft until you have self-reviewed it and run
`./scripts/check.sh`.
* When you have addressed a PR comment, mark it as resolved.
Please be patient! I will review your PR, but my time is limited!
-->
Closes <https://github.com/emilk/egui/issues/3862>.
Factoring the `bool` members of `Response` into a bitfield, the size of
`Response` is now 96 bytes (down from 104).
I gave `Sense` the same treatment, however this has no effects on
`Response` due to padding. I've decided not to pursue `PointerState`, as
it is quite large (_many_ members that are sized and aligned to
multiples of 8 bytes), so I don't expect any noticeable benefit from
making handful of `bool`s slightly leaner.
In any case, the changes to `Sense` are already quite a bit more
intrusive than those to `Response`.
The previous implementation overloaded the names of the attributes
`click` and `drag` with similarly named methods that _construct_ `Sense`
with the corresponding flag set. Now, that the attributes can no longer
be accessed directly, I had to introduce methods with new names
(`senses_click()`, `senses_drag()` and `is_focusable()`). I don't think
this is the cleanest solution: the old methods are essentially redundant
now that the named constants like `Sense::CLICK` exist. I did however
not want to needlessly break backwards compatibility.
I am happy to revert it (or go further 🙂) if there are concerns.
[ x ] I have ~~followed~~ _read_ the instructions in the PR template
Unfortunately i had several issues:
- Some snapshot-tests didn't run successfully on osx. diff shows errors
around fonts or missing menu items)
- cargo clippy doesn't run successfully (egui_kittest cannot find `wgpu`
and `image`)
- ./scripts/check.sh had other issues on my system (env: python: No such
file or directory), even if python3 can be called via python in my shell
Is there a system independent, standard way to run these tools (e.g. via
Docker?)
I submit the pr anyway, because there changes are very simple and
shouldn't cause issues.
A user of my Windows application reported a papercut where the
application restores its size on next load, but does not restore its
maximized state. This PR fixes that.
To test, I patched https://github.com/emilk/eframe_template to use my
local code since I knew that template saves/restores window data.
Testing methodology was to simply `cargo run`, maximize the application,
then close the application. `cargo run` again and the application should
start maximized.
Closes#1517.
* [x] I have followed the instructions in the PR template
* * This is mostly true, I had difficulties running `./scripts/check.sh`
for some reason. Possibly a bad Python version?
* Part of https://github.com/emilk/egui/issues/4019
`Frame` now includes the width of the stroke as part of its size. From
the new docs:
### `Frame` docs
The total (outer) size of a frame is `content_size + inner_margin +
2*stroke.width + outer_margin`.
Everything within the stroke is filled with the fill color (if any).
```text
+-----------------^-------------------------------------- -+
| | outer_margin |
| +------------v----^------------------------------+ |
| | | stroke width | |
| | +------------v---^---------------------+ | |
| | | | inner_margin | | |
| | | +-----------v----------------+ | | |
| | | | ^ | | | |
| | | | | | | | |
| | | |<------ content_size ------>| | | |
| | | | | | | | |
| | | | v | | | |
| | | +------- content_rect -------+ | | |
| | | | | |
| | +-------------fill_rect ---------------+ | |
| | | |
| +----------------- widget_rect ------------------+ |
| |
+---------------------- outer_rect ------------------------+
```
The four rectangles, from inside to outside, are:
* `content_rect`: the rectangle that is made available to the inner
[`Ui`] or widget.
* `fill_rect`: the rectangle that is filled with the fill color (inside
the stroke, if any).
* `widget_rect`: is the interactive part of the widget (what sense
clicks etc).
* `outer_rect`: what is allocated in the outer [`Ui`], and is what is
returned by [`Response::rect`].
### Notes
This required rewriting a lot of the layout code for `egui::Window`,
which was a massive pain. But now the window margin and stroke width is
properly accounted for everywhere.
Adds `Marginf` to fill the previous niche.
This is all in a pursuit to shrink the sizes of often-used structs, to
improve performance (less cache misses, less memcpy:s, etc).
* On the path towards https://github.com/emilk/egui/issues/4019
* Part of https://github.com/emilk/egui/issues/4019
As part of the work on adding a custom `Border` to everything, I want to
make sure that the size of `RectShape`, `Frame` and the future `Border`
is kept small (for performance reasons).
This PR changes the storage of the corner radius of rectangles from four
`f32` (one for each corner) into four `u8`. This mean the corner radius
can only be an integer in the range 0-255 (in ui points). This should be
enough for most people.
If you want to manipulate rounding using `f32`, there is a new
`Roundingf` to fill that niche.
This adds `WidgetType::Image` and correctly sets it in the Image widget.
This allows us to query for images in kittest tests and tells accesskit
that a node is an image.
It also adds `Image::alt_text` to set a text that will be shown if the
image fails to load and will be read via screen readers. This also
allows us to query images by label in kittest.
* [x] I have followed the instructions in the PR template
---------
Co-authored-by: Emil Ernerfeldt <emil.ernerfeldt@gmail.com>
Adds support for animated WebP images. Used the already existing GIF
implementation as a template for most of it.
* [x] I have followed the instructions in the PR template
---------
Co-authored-by: Emil Ernerfeldt <emil.ernerfeldt@gmail.com>
<!--
Please read the "Making a PR" section of
[`CONTRIBUTING.md`](https://github.com/emilk/egui/blob/master/CONTRIBUTING.md)
before opening a Pull Request!
* Keep your PR:s small and focused.
* The PR title is what ends up in the changelog, so make it descriptive!
* If applicable, add a screenshot or gif.
* If it is a non-trivial addition, consider adding a demo for it to
`egui_demo_lib`, or a new example.
* Do NOT open PR:s from your `master` branch, as that makes it hard for
maintainers to test and add commits to your PR.
* Remember to run `cargo fmt` and `cargo clippy`.
* Open the PR as a draft until you have self-reviewed it and run
`./scripts/check.sh`.
* When you have addressed a PR comment, mark it as resolved.
Please be patient! I will review your PR, but my time is limited!
-->
* Images with capitalized extensions do not load because the list of
extensions they are checked against is lowercase. The image extension is
now converted to lowercase before comparing
* [x ] I have followed the instructions in the PR template
This makes it easier to hit the corners.
Previously the corner response-area was covered by the response-areas of
the edges.
* Related to https://github.com/emilk/egui/pull/5523
An interactive widget should only be marked hovered if a click/drag
would start an interaction with it.
egui 0.30 introduced a feature where a thin interactive widget could be
hit even if it was partially behind a larger interactive widget.
Unfortunately, this introduced a bug where the top widget would still be
marked as hovered, even though a click would go through to the thin
widget below.
This bug was most notacible when trying to reisize a window by dragging
its corner, which often would result in dragging one of its sides
instead.
This PR fixes this bug.
* Merge this first: https://github.com/emilk/egui/pull/5517
This aligns all rectangles and (horizontal or vertical) line segments to
the physical pixel grid in the `epaint::Tessellator`, making these
shapes appear crisp everywhere.
* Closes https://github.com/emilk/egui/issues/5164
* Closes https://github.com/emilk/egui/issues/3667
This undoes a lot of the explicit, egui-side aligning added in:
* https://github.com/emilk/egui/pull/4943
The new approach has several benefits over the old one:
* It is done automatically by epaint, so it is applied to everything (no
longer opt-in)
* It is applied after any layer transforms (so it always works)
* It makes line segments crisper on high-DPI screens
* All filled rectangles now has sides that end on pixel boundaries
<!--
Please read the "Making a PR" section of
[`CONTRIBUTING.md`](https://github.com/emilk/egui/blob/master/CONTRIBUTING.md)
before opening a Pull Request!
* Keep your PR:s small and focused.
* The PR title is what ends up in the changelog, so make it descriptive!
* If applicable, add a screenshot or gif.
* If it is a non-trivial addition, consider adding a demo for it to
`egui_demo_lib`, or a new example.
* Do NOT open PR:s from your `master` branch, as that makes it hard for
maintainers to test and add commits to your PR.
* Remember to run `cargo fmt` and `cargo clippy`.
* Open the PR as a draft until you have self-reviewed it and run
`./scripts/check.sh`.
* When you have addressed a PR comment, mark it as resolved.
Please be patient! I will review your PR, but my time is limited!
-->
Title. This would have helped me debug bugs quicker.
* [x] I have followed the instructions in the PR template
Was a duplicate article in the sentence. Already has "the"
<!--
Please read the "Making a PR" section of
[`CONTRIBUTING.md`](https://github.com/emilk/egui/blob/master/CONTRIBUTING.md)
before opening a Pull Request!
* Keep your PR:s small and focused.
* The PR title is what ends up in the changelog, so make it descriptive!
* If applicable, add a screenshot or gif.
* If it is a non-trivial addition, consider adding a demo for it to
`egui_demo_lib`, or a new example.
* Do NOT open PR:s from your `master` branch, as that makes it hard for
maintainers to test and add commits to your PR.
* Remember to run `cargo fmt` and `cargo clippy`.
* Open the PR as a draft until you have self-reviewed it and run
`./scripts/check.sh`.
* When you have addressed a PR comment, mark it as resolved.
Please be patient! I will review your PR, but my time is limited!
-->
* Closes <https://github.com/emilk/egui/issues/THE_RELEVANT_ISSUE>
* [ ] I have followed the instructions in the PR template
Native is already delayed by a frame because it calls
`handle_viewport_output` -> `egui_winit::process_viewport_commands`
after drawing. On web however, we process input including viewport
commands separately from drawing.
This adds an arbitrary frame delay mechanism for web and then uses this
with 1 frame delay always
When there are multiple layers (e.g. with custom transforms) close to
each other, the hit test code used to only consider widgets in the layer
directly under the mouse. This can make it difficult to hit thin widgets
just on the outside of a transform layer.
This PR fixes that.
It also prioritizes thin widgets, so that if there is both a thin widget
and a thick widget under the mouse cursor, you will always hit the thin
widgets, even if the thin widgets is layered behind the thick one. This
makes it easier to hit thin resize-handles.
In theory this should allow us to make `resize_grab_radius_side` and
`resize_grab_radius_corner` smaller in a future PR, if we want to.
Andrew Farkas
Aely
lucasmerlin
Pol Welter
Markus Ineichen
Lander Brandt
Emil Ernerfeldt
YgorSouza
Tom Maffia
Ryan Bluth
Jochen Görtler
Onè
Andreas Reich