Especially on new setups, it is handy for `headless-browser --run-tests`
to "just work". Anyone using ladybird.py will have LADYBIRD_SOURCE_DIR
set in their environment already.
The streams AO that we were calling to close the stream would assert
if it was not in a readable state. This version of close is exported
publicly in the streams specification, and properly handles this
situation.
Fixes a crash in the imported test, and happens to fix some others!
This header file is never included by any other headers, and including
it would require adding the fontconfig include paths to the swift
interop header generation, which is not desirable.
We haven't built a universal binary in over 11 months. The name is
confusing, and actually breaks esvu on macOS. The fact that nobody
has complained suggests that this is not a common use case..
On my Linux machine with 32 cores, ninja actually defaults to 34 jobs.
By defaulting ourselves to multiprocessing.cpu_count(), we actually
decrease the number of jobs used.
We don't need to pick a host compiler every time ladybird.py is invoked.
We only need to do so when configuring the Build directory.
This shaves about 10ms off an invocation of `ladybird.py build` on my
machine, going from ~170ms to ~160ms.
Not cleaning these up by rejecting or resolving the promise causes
the main thread to try to reject them at EventLoop::exit() time.
If the RenderThread has already been destroyed by then, we get into
use-after-free territory and segfault.
We already had all necessary things for pseudo elements support in place
except ability to save transition properties in Animatable. This commit
adds the missing part.
The way we echo the cache keys for output variables strips these quotes.
So when we save the caches at the end of CI, the keys are sans quotes.
This patch adds an extra cache restore key without quotes to allow jobs
to fetch their own caches again. This will become moot once all runners
are able to use the Blacksmith cache action.
In WindowProxy.[[Get]] it's not guaranteed that the current principal
global object has an associated document at the moment. This may happen
if a script is continuing to execute while a navigation has been
initiated.
Because of that, we can't blindly dereference the active document
pointer, so this patch adds a null check.
By default, we want `ladybird.py build` to build everything. We were
previously defaulting to only building the Ladybird target.
We now only fall back to the Ladybird target for commands that run a
program. So `ladybird.py run` will build and run Ladybird.
As opposed to just running subprocess.check_call, our `run_command`
utility handles e.g. ctrl+c to avoid spamming the terminal with
KeyboardInterrupt stack traces.
This will allow us to re-use this logic from within other python
scripts. The find_compiler.sh script still exists, as it is used by
some other bash scripts. The pick_host_compiler() function will now
execute find_compiler.py and store its result in $CC and $CXX.
Note that the python script supports Windows.
This is the default python version on macOS, so let's support it since
it is trivial for now. Using "str | None" as a type annotation is only
supported in python 3.10 or later.
This will be needed by other scripts.
To do so, this patch gives up on the importlib method of importing
packages. I tried extracting this helper to e.g. __init__.py, but the
python runtime was unable to find the imported symbols.
When recording the display list for a stacking context, the following
operations (relevant to this bug) happened:
* push a stacking context
* as part of that push a None-value to the scroll frame id stack
* apply filters
* apply masking
* paint recursively
This meant that mask-images were always recorded without scroll frame
id, causing them to be painted without any scroll offset. As a result
mask-images would break as soon as the website using them was scrolled.
Instead, push to the scroll frame id stack later to solve the problem:
* push a stacking context
* apply filters
* apply masking
* push a None-value to the scroll frame id stack
* paint recursively
When serializing CSS declarations we now support combining multiple
properties into a single shorthand property in some cases.
This comes with a healthy dose of FIXMEs, including work to be done
around supporting:
- Nested shorthands (e.g. background, border, etc)
- Shorthands which aren't represented by the ShorthandStyleValue type
- Subproperties pending substitution
This gains us a bunch of new test passes, both for WPT and in-tree
This exposed a few bugs which caused the following tests to behave
incorrectly:
- `tab-size-text-wrap.html`: This previously relied on a bug where we
incorrectly treated `white-space: pre` as allowing text wrapping. The
fix here is to implement the text-wrap CSS shorthand property.
- `execCommand-preserveWhitespace.html`: We don't correctly serialize
shorthand properties. This is covered by an existing FIXME in
`CSSStyleProperties::serialized()`
- `white-space-shorthand.html`: The last 5 subtests here fail as we
don't correctly handle shorthand properties in
`CSSStyleProperties::remove_property()`. This is covered by an
existing FIXME in said function.
This commit enables building and testing js.exe for windows. Needed
libraries are built in CI, and tests for those which pass were added.
Tests for LibJS which don't require javascripttestrunner were added but
the rest need to wait for that to be ported to windows.
This commit allows building js.cpp on windows. The repl functionality is
ifdef'ed out. To decrease the number of ifdefs the code that runs the
repl on Linux was moved into one place. Some globals that are unused as
a result of that are markes maybe_unused. The following commit enables
building and testing js in cmake for windows.
This commit adds the minimal export macros needed to run js.exe on
windows. A followup commit is planned to move to explicit export
entirely.
A static_assert for the size of a struct is also ifdef'ed out as the
semantics around object layout and inheritance are different on MSVC abi
and the struct IteratorRecord ends up being 40 bytes not 32.
