All fields are always initialized, so we don't need to initialize them
by default. This lets us send types over IPC that can't be
default-constructed, such as a Variant without Empty.
At some point, we stopped ever constructing invalid messages. This makes
that clearer, and will allow us to stop requiring that IPC arguments be
default-constructible.
Remember the query, so that if you're filtering for "color" on the
computed style tab, and switch to the resolved style tab, it's filtered
for "color" too.
This means we also can save looking up the filter text when a new node
is inspected.
If you have a search filter and then click on a different node, this now
applies the filter to the new node's properties, instead of showing all
of them.
These were previously skipped on macOS, due to hard to
fix inconsistencies in the exact produced pixels.
Turns out, this is not just a macOS thing. The same sort
of hard-to-spot slight pixel deviations are present on
arm64 Linux as well.
Explicitly link final targets with OpenSSL to ensure that the vcpkg
version is loaded instead of the system one.
Before this change we would inherit `libcrypto.so` and `libssl.so` from
other dependencies, like Qt, that do not have their RPATH rewritten.
This would cause the loader to prefer the system libraries over the
vcpkg ones causing all sorts of version mismatch issues.
The effectiveness of this change can be verified with
`readelf -d ./bin/Ladybird` showing `libcrypto.so` and `libssl.so` as
direct dependencies, before they would not appear. Additionally, `ldd`
will show `libcrypto.so` and `libssl.so` pointing to the vcpkg builds.
When drawing a table, some of the CSS properties must be moved from the
table grid box to an anonamyous table wrapper box. One of these
properties is `position`. `z-index` however is not. This leads to the
following behavior if a table has both `position` and `z-index`:
* The wrapper box has the `position`, but a `z-index` of `auto`.
* The grid box has the `z-index`, but `position: static`.
This effectively means that the `z-index property is ignored since it
has no effect on non-positioned elements. This behavior contradicts what
other browsers do and causes layout issues on websites.
To align Ladybird behavior with other browser this commit also moves the
`z-index` property to the wrapper box.
If the regex always matches the input, even if it's past the end, then
we need to stop execution of the regex when it's past the end. This
corresponds to step 13.a and prevents it from infinitely looping.
Reduced from: d98672060f/packages/react-i18n/src/utilities/money.ts (L10-L14)
The usage of color-scheme was introduced in commit:
ce5cd012b9
This reverts the Inspector to use self-selected colors, as the default
color-scheme colors do not look great in this window.
If the expansion of a custom property in variable expansion returns
tokens, then the custom property is not the initial guaranteed-invalid
value.
If it didn't return any tokens, then it is the initial
guaranteed-invalid value, and thus we should move on to the fallback
value.
Makes Shopify checkout show the background colours, borders, skeletons,
etc.
Our existing coalescing mechanism for input events didn't prevent
multiple mousemove/mousewheel events from being processed between paint
cycles. Since handling these events can trigger style & layout updates
solely for hit-testing purposes, we might end up doing work that won't
be observable by a user and could be avoided by shceduling input events
processing to happen right before painting the next frame.
Currently we create URLs such as 'about:blank' through the StringView
or ByteString constructor of URL. However, in order to elimate the
use of URL::is_valid, we need to get rid of these constructors as it
makes it way too easy to create an invalid URL.
It is very cumbersome to construct an 'about:blank' URL when using
URL::Parser::basic_parse. So instead of doing that, create some
helper functions which will create the 'about:XXX' URLs with the
correct properties set.
Conveniently, this is also a much faster way of creating these URLs
as it means we do not need to parse the URL and can set all of the
members up front.
The new test case crashes during bytecode generation due to
`emit_super_reference` not correctly generating the reference record
for the property access.
MediaQueryList will now remember if a state change occurred when
evaluating its match state. This memory can then be used by the document
later on when it's updating all queries, to ensure that we don't forget
to fire at least one change event.
This also required plumbing the system visibility state to initial
about:blank documents, since otherwise they would be stuck in "hidden"
state indefinitely and never evaluate their media queries.
Instead of checking all elements in a document for containment in
`:has()` invalidation set, we could narrow this down to ancestors and
ancestor siblings, like we already do for subject `:has()` invalidation.
This change brings great improvement on GitHub that has selectors with
non-subject `:has()` and sibling combinators (e.g., `.a:has(.b) ~ .c`)
which prior to this change meant style invalidation for whole document.
The Linux IPC uses SCM_RIGHTS to transfer fds to another process
(see TransportSocket::transfer, which calls LocalSocket::send_message).
File descriptors are handled separately from regular data.
On Windows handles are embedded in regular data. They are duplicated
in the sender process.
Socket handles need special code both on sender side (because they
require using WSADuplicateSocket instead of DuplicateHandle, see
TransportSocketWindows::duplicate_handles) and on receiver side
(because they require WSASocket, see FileWindows.cpp).
TransportSocketWindows::ReadResult::fds vector is always empty, it is
kept the same as Linux version to avoid OS #ifdefs in Connection.h/.cpp
and Web::HTML::MessagePort::read_from_transport. Separate handling of
fds permeates all IPC code, it doesn't make sense to #ifdef out all this
code on Windows. In other words, the Linux code is more generic -
it handles both regular data and fds. On Windows, we need only the
regular data portion of it, and we just use that.
Duplicating handles on Windows requires pid of target (receiver)
process (see TransportSocketWindows::m_peer_pid). This pid is received
during special TransportSocketWindows initialization, which is performed
only on Windows. It is handled in a separate PR #3179.
Note: ChatGPT and [stackoverflow](https://stackoverflow.com/questions/25429887/getting-pid-of-peer-socket-on-windows) suggest using GetExtendedTcpTable/GetTcpTable2
to get peer pid, but this doesn't work because [MIB_TCPROW2::dwOwningPid](https://learn.microsoft.com/en-us/windows/win32/api/tcpmib/ns-tcpmib-mib_tcprow2)
is "The PID of the process that issued a context bind for this TCP
connection.", so for both ends it will return the pid of the process
that called socketpair.
Co-Authored-By: Andrew Kaster <andrew@ladybird.org>
Also:
* Use SpecializationOf for Optional and Variant concepts
* Remove inline because it is implied by constexpr
* Reorder declarations from less to more specialized (in common sense)
It might be a good idea to do this on other platforms as well, but at
least on Windows, the command line for GenerateWindowOrWorkerInterfaces
becomes too large.
Before this commit, LibCore/System.h exposed only part of
System::stat API on Windows. Namely, users of Core::System::stat
had to #include <dirent.h> in order to check the return value of stat.
It is OK for low-level libs like LibCore/LibFileSystem, but
S_ISDIR is also used in LibWeb\Loader\GeneratedPagesLoader.cpp.
We want to avoid platform #ifdefs in LibWeb.
This introduces a new API in ImageDecoderPlugins that allow an image
decoder to return a CICP struct. Also, we use this API in
ImageDecoder::color_space() to create a color space corresponding to
these CICP.
The spec wants these keywords to appear in a particular order when
serialized, so let's just put them in that order during parsing.
This also fixes a bug where we didn't reject `font-variant-east-asian`
that contains `normal` alongside another value.
Also, rather than always parsing them as a StyleValueList, parse single
values on their own, and then support that in the to_font_variant_foo()
methods.
Without this, we'd happily parse `font-variant-caps: small-caps potato`
as just `small-caps` and ignore the fact that unused tokens were left
over.
This fix gets us some WPT subtest passes, and removes the need for a
bespoke parsing function for font-variant-caps.
Without this, getting a property's value from `element.style.foo` would
fail if `foo` is a shorthand property which has a longhand that is also
a shorthand. For example, `border` expands to `border-width` which
expands to `border-top-width`.
This is because we used `property()` to get a longhand's value, but this
returns nothing if the property is a shorthand.
This commit solves that by moving most of get_property_value() into a
separate method that returns a StyleProperty instead of a String, and
which calls itself recursively for shorthands. Also move the manual
shorthand construction out of ResolvedCSSStyleDeclaration so that all
CSSStyleDeclarations can use it.
This is a weird behaviour specific to `font` - it can reset some
properties that it never actually sets. As such, it didn't seem worth
adding this concept to the code generator, but just manually stuffing
the ShorthandStyleValue with them during parsing.
We can't simply walk the element tree and pass in `pseudo_element` each
time. Instead, we want to look at:
1. Element's pseudo-element
2. Element
3. Element's parents
This allows us to inspect its properties. To avoid wasted work, we only
compute and cache the properties if the originating element was, or is,
displaying as a list item.
This commit changes the strategy for updating inherited styles. Instead
of marking all potentially affected nodes during style invalidation, the
decision is now made on-the-fly during style recalculation. Child nodes
will only have their inherited styles recalculated if their parent's
properties have changed.
On Discord this allows to 1000x reduce number of nodes with recalculated
inherited style.
This was an old hack intended to make percentage sizes on flex items
before we had implemented the appropriate special behavior of definite
sizes in flex layout.
Removing it makes flex layout less magical and should not change
behavior in any observable way.
The spec tells us to treat `auto` as `fit-content` when determining
flex item cross sizes, so let's just do *that* instead of awkwardly
doing an uncacheable nested layout of the item.
This was the only instance of `LayoutState` nesting outside of intrinsic
sizing, so removing it is an important step towards simplifying layout.
Turns out it was a lot easier than expected.
We've long claimed to support this, but then silently ignored string
values, until 4cb2063577 which would
not-so-silently crash instead. (Oops)
So, actually pass the string value along and use it in the list marker.
As part of this, rename our `list-style-type` enum to
`counter-style-name-keyword`. This is an awkward name, attempting to be
spec-based. (The spec says `<counter-style>`, which is either a
`<counter-style-name>` or a function, and the `<counter-style-name>` is
a `<custom-ident>` that also has a few predefined values. So this is the
best I could come up with.)
Unfortunately only one WPT test for this passes - the others fail
because we produce a different layout when text is in `::before` than
when it's in `::marker`, and similar issues.
These are copied and modified from ByteString, with the addition of a
Case parameter so that we can construct them in lowercase instead of
having to them make a copy.
`invalidate_style()` already tries to avoid scheduling invalidation for
`:has()` by checking result of `may_have_has_selectors()`, but it might
still result in unnecessary work because `may_have_has_selectors()`
does not force building of rules cache. This change adds
`have_has_selectors()` that forces building of rules cache and is
invoked in `update_style()` to double-check whether we actually need to
process scheduled `:has()` invalidations.
This allows to skip ~100000 ancestor traversals on this WPT test:
https://wpt.live/html/select/options-length-too-large.html
It fixes a bug in which ImageDecoder and RequestServer
do not exit because their connections don't close.
This makes the shutdown behavior match the Linux version,
which receives FD_READ | FD_HANGUP on socket close, and
TransportSocket::read_as_much_as_possible_without_blocking calls
schedule_shutdown when read from a socket returns 0 bytes.
On Windows, we have to explicitly call WIN32 shutdown to receive
notification FD_CLOSE.
Windows flavor of non-blocking IO, overlapped IO, differs from that on
Linux. On Windows, the OS handles writing to overlapped buffer, while
on Linux user must do it manually.
Additionally, we can only have overlapped sockets because it is the
requirement to be able to wait on them - WSAEventSelect automatically
sets socket to nonblocking mode.
So we end up emulating Linux-nonblocking sockets with
Windows-nonblocking sockets.
Pending IO state (ERROR_IO_PENDING) must not escape read/write
functions. If that happens, all synchronization like WSAPoll and
WaitForMultipleObjects stops working (WaitForMultipleObjects stops
working because with overlapped IO you are supposed to wait on an event
in OVERLAPPED structure, while we are waiting on WSA Event, see
EventLoopImplementationWindows.cpp).
This is not a very pleasant fix, but matches a similar const_cast that
we do to return JS objects returned in a union. Ideally we would
'simply' remove the const from the value being visited in the variant,
but that opens up a whole can of worms where we are currently relying on
temporary lifetime extension so that interfaces can return a Variant of
GC::Ref's to JS::Objects.
We were previously assuming that dictionary members were always
required when being returned.
This is a bit of a weird case, because unlike _input_ dictionaries
which the spec marks as required, 'result' dictionaries do not seem to
be marked in spec IDL as required. This is still fine from the POV that
the spec is written as it states that we should only be putting the
values into the dictionary if the value exists.
We could do this through some metaprogramming constexpr type checks.
For example, if the type in our C++ representation was not an
Optional, we can skip the has_value check.
Instead of doing that, change the IDL of the result dictionaries to
annotate these members so that the IDL generator knows this
information up front. While all current cases have every single
member returned or not returned, it is conceivable that the spec
could have a situation that one member is always returned (and
should get marked as required), while the others are optionally
returned. Therefore, this new GenerateAsRequired attribute is
applied for each individual member.
This fixes a compile error of multiple variables of the same name within
the same scope for the URLPattern IDL, which has a dictionary return
type that contains multiple dictionaries of the same type. Conveniently,
this also makes the complicated generated code of the URLPattern
interface easier to read by adding some more structure :^)
This is the return value of a URLPattern after `exec` is called on it.
It conveys information about the named (or unammed) regex groups
matched for each component of the URL. For example,
```
let p = new URLPattern({ hostname: "{:subdomain.}*example.com" });
const result = pattern.exec({ hostname: "foo.bar.example.com" });
console.log(result.hostname.groups.subdomain);
```
Will log 'foo.bar'.
We don't yet support a proxy configuration, but we can still validate
the capability received from the WebDriver client. We should also fail
to create a WebDriver session if a proxy configuration is present.
We currently define our custom WebDriver capabilities with a dictionary
of the form:
"serenity:ladybird": {
"headless": true
}
This patch flattens the configuration, such that each Ladybird option
will be its own capability. This matches how Firefox configures their
own options with geckodriver. So we now have:
"ladybird:headless": true
The WebDriver spec now separately tracks an active HTTP session list,
which will contains all non-BiDi WebDriver sessions by default. There
may only be one active HTTP session at a time.
See: https://github.com/w3c/webdriver/commit/63a397f
Session management is a bit awkward right now in that the list of active
sessions is managed by Client, resulting in operations like closing a
session being split between several functions in Client and Session.
This patch moves all session management to the Session class. Closing a
session is now entirely in Session::close().
This will make managing a separate HTTP session list a bit simpler.
When a BackgroundAction completes, it resolves a Promise (stored on the
BackgroundAction object) with a reference to itself. The Promise will
never unset this resolved value, thus it will hold a strong reference to
the BackgroundAction until it is destroyed. But because the Promise is
owned by the BackgroundAction itself, we have a reference cycle, and
neither object can be destroyed.
The only user of BackgroundAction is the ImageDecoder process. The
consequence was that the ImageDecoder process would never release any
image data for successfully decoded images.
To fix this, instead of storing the promise on the class itself, we can
just create it as a local variable and pass it around.
We have to be careful to always destroy the jpeglib decompression struct
before returning from JPEGLoadingContext::decode. We were doing this in
jpeglib error handlers, but we have a couple of paths that bail from the
decoder via TRY. These paths were neither cleaning up memory nor setting
the image decoder to an error state.
So this patch sets up a scope guard to ensure we free the decompressor
upon exit from the function. And it delegates the responsibility of
setting the decoder state to the caller (of which there is only one),
to ensure all error paths result in an error state.
Executing scripts via WebDriver has a bit of awkwardness around dealing
with user dialogs that open during script execution. When this happens,
we must return control back to the client immediately with a null
response, while allowing the script to continue executing. When the
script completes, we must then ignore its result.
We've previously handled this by tracking a boolean for the ongoing
script execution, set to true when the script begins and false when it
ends (either via normal script completion or the above dialog handling).
However, this failed to handle the following scenario, running two
scripts in a row:
execute_script("alert('hi'); return 1;")
execute_script("return 2;")
The first script would execute and open a dialog, and thus return a null
response to the client while the script continued and the dialog remains
open. The second script would "handle any user prompts", which closes
the dialog. This would end the execution of the first script. But since
we're now executing a script again, the boolean flag is true, and we'd
return the result of the first script back to the client. The client
would then think this is the result of the second script.
So we now track script execution with a simple ID. If a script completes
whose execution ID is not the ID of the currently executing script, we
drop the result.
Tests have the glob run against the relative path of the test file.
Since this was never set for crash tests the '-f' argument to
headless browser would never match the global against any crash test.
Moves pseudo class matching helpers into Element methods, so they don't
have to be duplicated between SelectorEngine and function that checks if
element is included in invalidation set.
This was an old hack from before we understood how and when to resolve
percentages in flex layout. Removing it should not change anything,
but it does avoid a lot of redundant layout work on many pages.
The current implementation of `:has()` style invalidation is divided
into two cases:
- When used in subject position (e.g., `.a:has(.b)`).
- When in a non-subject position (e.g., `.a > .b:has(.c)`).
This change focuses on improving the first case. For non-subject usage,
we still perform a full tree traversal and invalidate all elements
affected by the `:has()` pseudo-class invalidation set.
We already optimize subject `:has()` invalidations by limiting
invalidated elements to ones that were tested against `has()` selectors
during selector matching. However, selectors like `div:has(.a)`
currently cause every div element in the document to be invalidated.
By modifying the invalidation traversal to consider only ancestor nodes
(and, optionally, their siblings), we can drastically reduce the number
of invalidated elements for broad selectors like the example above.
On Discord, when scrolling through message history, this change allows
to reduce number of invalidated elements from ~1k to ~5.
This fixes a crash in the included test that regressed in 0adf261,
and is hit by the following HTML:
```html
<body></body>
<script>
const frame = document.body.appendChild(document.createElement("iframe"));
frame.contentDocument.open();
const child = frame.contentDocument.createElement("html")
const html = frame.contentDocument.appendChild(child);
frame.contentDocument.close();
</script>
```
I am not 100% sure this is fully the correct fix and there are other
cases which would not work properly. But it's definitely an improvement
to make the confuisingly named 'insert_an_eof' function of the tokenizer
actually do something.
Previously, if the user made a find-in-page query, then cleared the
selection made by that query, subsequent queries would inadvertently
advance to the next match instead of reselecting the first match.
The implementation was removed with the migration to ANGLE. This
reimplements it. This is required by Stimulation Clicker on neal.fun,
which does not clear the framebuffer itself, instead relying on the
browser doing it.
These properties are always substrings of the RegExp input string,
and so we can store them as views and lazily construct strings if
they're actually accessed (which most of the time they aren't).
This avoids a bunch of unnecessary memory copying, saving roughly
2.1 seconds per iteration of Speedometer.
Required by the server-side rendering mode of React Router, used by
https://chatgpt.com/
Note that the imported tests do not have the worker variants to prevent
freezing on macOS.
ReadLoop requests require the chunks to be Uint8Array objects, however,
TextEncoderStream requires a String (Convertible) value. This is fixed
by implementing read_all_chunks as a loop of DefaultReader requests
instead, which is an identity transformation. This should be okay to
do, as stream chunk steps expect a JS::Value, and convert it to the
type they want.
Before this change, tasks associated with a destroyed document would get
stuck in the task queue forever, since document-associated tasks are not
allowed to run when their document isn't fully active (and destroyed
documents never become fully active again). This caused everything
captured by task callbacks to leak.
We now treat tasks for destroyed documents as runnable immediately,
which gets them out of the queue.
This fixes another massive GC leak on Speedometer.
Before this change, Agent held on to all of the live MutationObserver
objects via GC::Root. This prevented them from ever getting
garbage-collected.
Instead of roots, we now use a simple IntrusiveList and remove them
from it in the finalizer for MutationObserver.
This fixes a massive GC leak on Speedometer.
This matches the prototype attributes.
Used by https://chatgpt.com/, where it runs this code:
```js
CSS.supports('animation-timeline: --works')
```
If this returns false, it will attempt to polyfill Animation Timeline
and override CSS.supports to support Animation Timeline properties.
f7a3f78 made the layout tree invalidate only the inserted nodes
themselves, but it turned out that CSS containment invalidation relies
on the parent being invalidated as well.
Incorrect behavior of CreateProcess:
CreateProcess("a.exe", "b c", ...) ->
a.exe::main::argv == ["b", "c"] (wrong)
CreateProcess(0, "a.exe b c", ...) ->
a.exe::main::argv == ["a.exe", "b", "c"] (right)
https://learn.microsoft.com/en-us/cpp/cpp/main-function-command-line-args
"If you use both the first and second arguments (lpApplicationName and
lpCommandLine), argv[0] may not be the executable name."
This means first argument of CreateProcess should never be used.
-----------------------------------------------------------------------
Searching for executable in path is suppressed now by prepending "./"
Additional bonus of the new code: .exe extension does not need to be
specified.
There is no need for this invalidation because taking care of siblings
is already done by invalidation with `NodeInsertBefore` reason. Parent
element itself (without subtree) is always invalidated by
`Node::children_changed()` hook, so `:empty` pseudo-class invalidation
is already covered.
When checking whether an early return is possible because some ancestor
already has the whole subtree invalidation flag set, the check should
begin with the current node's parent rather than with the node itself.
Otherwise, if a node already has the whole subtree invalidation flag
set and is subsequently invalidated for the reason `NodeInsertBefore`
or `NodeRemove`, we will skip the sibling invalidation required for
these operations
This fix is required for optimizations in subsequent commits.
With this change, siblings of an inserted node are no longer invalidated
unless the insertion could potentially affect their style. By
"potentially affected," we mean elements that are evaluated against the
following selectors during matching:
- Sibling combinators (+ or ~)
- Pseudo-classes :first-child and :last-child
- Pseudo-classes :nth-child, :nth-last-child, :nth-of-type, and
:nth-last-of-type
This test appears to be flaky, similar to some other animation tests we
have. Let's disable it for now because changes in subsequent commits
cause it to fail more often, even though those changes are not related
to the test itself. It is likely another case where altered timings of
certain operations (specifically, style updates) cause the test to fail
because things are no longer executed in the same order as before.
This is not really a context, but more of a set of parameters for
creating a Parser. So, treat it as such: Rename it to ParsingParams,
and store its values and methods directly in the Parser instead of
keeping the ParsingContext around.
This has a nice side-effect of not including DOM/Document.h everywhere
that needs a Parser.
A few of these are only ever called with T=Token, so let's simplify them
a bit.
As a drive-by change: Also correct the "unnecessairy" typos and use
discard_a_token().
This file has been a pain to edit for a while, even with the previous
splits. So, I've divided it up into 3 parts:
- Parser.cpp has the "base" code. It's the algorithms and entry-points
defined in the Syntax spec.
- ValueParsing.cpp contains code for parsing single values, such as a
length, or a color, or a calculation.
- PropertyParsing.cpp contains code for parsing an entire property's
value. A few of these sit in a grey area between being a property's
value and a value in their own right, but the rule I've used is "is
this useful outside of a single property and its shorthands?"
This only moves code, with as few modifications as possible to make that
work. I did add explicit instantiations for the template implementations
as part of this, which revealed a few that are actually only compatible
with a single type, so I'll clear those up in a subsequent commit.
Lots of editorial spec bugs here, but these changes largely affect how
the unhandledPromptBehavior capability is handled. We also now set an
additional capability for the default User Agent string.
WebDriver script authors may now provide either:
* A user prompt handler configuration to be used for all prompt types.
* A set of per-prompt-type user prompt handlers.
This also paves the way for interaction with the beforeunload prompt,
though we do not yet support that feature in LibWeb.
See: https://github.com/w3c/webdriver/commit/43903d0
_O_OBTAIN_DIR flag makes _open use FILE_FLAG_BACKUP_SEMANTICS in
CreateFile call.
FILE_FLAG_BACKUP_SEMANTICS is required to open directory handles.
For ordinary files FILE_FLAG_BACKUP_SEMANTICS overrides file security
checks when the process has SE_BACKUP_NAME and SE_RESTORE_NAME
privileges, see https://learn.microsoft.com/en-us/windows/win32/api/fileapi/nf-fileapi-createfilea
icu 75.x and higher requires C++17. This change is pulled from an
abandoned PR to uprev vcpkg's version to 75. Presumably the flags
should be set upstream as well in their configure.ac
This catches errors that occur within async tests so that we fail faster
rather than timing out due to `done()` not being called.
We use `Promise.resolve()` because `f` isn't guaranteed to be an async
function.
At computed-value time, this is converted to whatever the parent's
computed value is. So it behaves a little like `inherit`, except that
an inherited start/end value uses the parent's start/end, which might
be different from the child's.
We've historically asserted that no "saturated" size values end up as
final metrics for boxes in layout. This always had a chance of producing
false positives, since you can trivially create extremely large boxes
with CSS.
The reason we had those assertions was to catch bugs in our own engine
code where we'd incorrectly end up with non-finite values in layout
algorithms. At this point, we've found and fixed all known bugs of that
nature, and what remains are a bunch of false positives on pages that
create very large scrollable areas, iframes etc.
So, let's change it! We now clamp content width and height of boxes to
17895700 pixels, apparently the same cap as Firefox uses.
There's also the issue of calc() being able to produce non-finite
values. Note that we don't clamp the result of calc() directly, but
instead just clamp values when assigning them to content sizes.
Fixes#645.
Fixes#1236.
Fixes#1249.
Fixes#1908.
Fixes#3057.
While keyword_to_foo() does return Optional<Foo>, in practice the
invalid keywords get rejected at parse-time, so we don't have to worry
about them here. This simplifies the user code quite a bit.
Used by chess.com, where it stores URLs to assets in CSS URL variables.
It then receives the value of them with getComputedStyle() and then
getPropertyValue(). With this, it trims off the url('') wrapper with a
simple slice(5, -2). Since we didn't preserve the opening quotation, it
would slice off the `h` in `https` instead of the quotation.
This fixes the very, _very_ slow loading of https://yzy-sply.com. The
`apply_style()` method also calls into this method recursively, so we
just need to call it once instead of once per node in the continuation
chain.
The promise job's fulfillment / rejection handlers may push an execution
context onto the VM, which will throw an internal error if our ad-hoc
call stack size limit has been reached. Thus, we cannot blindly VERIFY
that the result of invoking these handlers is non-abrupt.
This patch will propagate any internal error forward, and retains the
condition that any other error type is not thrown.
There are now no users of the MUST_OR_THROW_OOM macro. Let's rename this
macro to indicate it may be used to propagate any internal error (such
as the call stack limit error) in places that would otherwise crash due
to a MUST/VERIFY invocation.
Note there's no actual functional change here, as we weren't able to
ensure the internal error was an OOM error previously.
This is the current name for this property in CSS-Text-4. We don't
implement it, but at least our "missing property" message can be about
one we haven't implemented instead of one that's redundant. :^)
The previous code to determine the SourceDocument's lines was too naive:
the source text can contain other newline characters and sequences, and
the HTML/CSS/JS syntax highlighters would take those into account when
determining what line a token is on. This disagreement would cause
incorrect highlighting, or even crashes, if the source didn't solely use
`\n` for its newlines.
In order to have everyone agree on what a line is, this patch first
processes the source to replace all newlines with `\n`. The need to
copy the source like this is unfortunate, but viewing the source is a
rare enough action that this should not cause any noticeable
performance problems.
As the callers have a String, and we want a String, this also changes
the function parameters to keep the source as a String instead of
converting it to StringView and back.
Fixes https://github.com/LadybirdBrowser/ladybird/issues/3169
Useful when you need to check for more than one code point, some of
which are multiple bytes. (In which case StringView::contains() will
return wrong results.)
This change is mainly motivated by the fact that iterating in a loop
makes profiles easier to read and understand where time was spent in
traversal callback. Additionally, using a loop reduces function call
overhead and ensures constant stack usage.
This ad-hoc code informs the client of a potentially changed page title.
But because we always update the title element (either the SVG or HTML
title) the client was already informed, causing the code to run twice.
The Web::CSS::Parser's GradientParsing ignores color-stops if
it is only a single one. This change allows to have color-stops
with double positions against a single color.
Further, also allows for `linear-gradient(black)` and similar
other gradient functions
This was mainly a matter of deferring the wrapping of the button's
children until after its internal layout tree has been constructed.
That way we don't lose any pseudo elements spawned along the way.
Fixes#2397.
Fixes#2399.
Before this change, checking if fast selector matching could be used was
only enabled in style recalculation and hover invalidation. With this
change it's enabled for all callers of SelectorEngine::matches() by
default. This way APIs like `Element.matches()` and `querySelector()`
could take advantage of this optimization.
Before, if something went wrong with DNS lookup and there were unrelated
records (i.e. not A or AAAA) then we would still attempt to build a
resolve list. This resulted in curl errors related to the option itself
and displayed as "unknown network error" to the user.
This is "update document for history step application" but that's too
long for the commit title. :^)
No code changes, just adding more FIXME comments for the new steps.
(And indented step 7's substeps for clarity.)
Corresponds to https://github.com/whatwg/html/pull/10910
Before this change, `m_needs_repaint` was reset in
`Document::record_display_list()` only when the cached display list was
absent. This meant that if the last triggered repaint used the cached
display list, we would keep repainting indefinitely until the display
list was invalidated (We schedule a task that checks if repainting is
required 60/s).
This change also moves `m_needs_repaint` from Document to
TraversableNavigable as we only ever need to repaint a document that
belongs to traversable.
We just need to create the backend context once and let Skia handle the
context's state.
On my machine, this reduces the load time for https://tweakers.net from
7.5s to 3.5s.
This patch synchronizes changes from whatwg/fetch#1569 and
resolves a related FIXME: "Refactor this to the new version of the
spec introduced with whatwg/fetch@464326e.”
Allow wheel event to be consumed by a `overflow: scroll` box only if it
has content that overflows a scrollport.
This fixes the timing issue in the
`Text/input/scroll-window-using-wheel-event.html` test, where a `<body>`
element with `overflow: scroll` was incorrectly consuming wheel events
that should have propagated to the window.
Previous name for misleading because it checks if box could be scrolled
by user input event which is diffent from checking if box is scrollable.
For example box with `overflow: hidden` is scrollable but it can't be
scrolled by user input event.
Using https://github.com/whatwg/html/pull/9457
(with some changes made to catch up with the current spec)
to fix a spec bug and a crash when removing a visible popover.
I can't find of a way to trigger this codepath, and there is for sure no
correpsonding WPT test, but let's fix this up for whenever it does
become relevant.
Instead of adding a flag for the two callers that need a pop of the
execution context stack when invoking continue_async_execution inline
the pop of the execution context.
This makes the management of these stacks and surrounding VERIFY calls
much more obvious.
This has no functional difference as run_queued_promise jobs does
nothing when LibWeb is used as it has a different implementation of
enqueuing and running promise jobs. But this change makes it more
obvious that run_queued_promise jobs does nothing when there is an
embedder, and adjusts the comment to reflect what the code is
actually achieving.
This extends the optimization introduced in the previous commit to
also apply to the inserted steps for an option element. This makes
the test:
https://wpt.live/html/select/options-length-too-large.html
Go from not ever completing due to how slow it was to running, to
finishing in 800ms on my PC :^)
We can definitely expand on this a bunch more, but using the metadata
provided in the children change notification we are able to skip
runnning the expensive selectedness algorithm on the <select> element.
This removes children changed from appearing in the profile of:
https://wpt.live/html/select/options-length-too-large.html
Currently, this metadata is only provided on the insertion steps,
though I believe it would be useful to extend to the other cases
as well. This metadata can aid in making optimizations for these
steps by providing extra context into the type of change which
was made on the child.
This is still called _way_ too often, and we need to be much smarter
about when this needs to run. But we get two wins from this very
naive implementation:
1. There is a inner text setter nested within the selectedness
steps uses this list of elements. This saves us building
up a list of elements again within these steps.
2. More importantly, it caches the number of selected elements.
This already allows us to skip a minor amount of work iterating
over the children again. But in future commits, this will serve
as one of the criteria for skipping running the selectedness
algorithm altogether for certain cases, which is a very big win.
A example future idea might be to append to this list directly when
something like appendChild is run instead of iterating over all of
the children again. But that's left as future work.
The spec only refers to this property as something that is on a
Window object, and as far as I can tell this API is only ever
exposed on a Window object anyhow.
This is fundametally broken. A microtask only finishes after all
javascript has finished running. The selectedness of a select element
is observable by javascript, so any changes which are made as a result
of children changing associated with a select element should be made
through the children update steps and friends.
If a calculation was simplified down to a single numeric node, then most
of the time we can instead return a regular StyleValue, for example
`calc(2px + 3px)` would be simplified down to a `5px` LengthStyleValue.
This means that parse_calculated_value() can't return a
CalculatedStyleValue directly, and its callers all have to handle
non-calculated values as well as calculated ones.
This simplification is reflected in the new test results. Serialization
is not yet correct in all cases but we're closer than we were. :^)
Calc simplification eventually produces a single style-value as its
output. This extra context data will let us know whether a calculated
number should be treated as a `<number>` or an `<integer>`, so that for
example, `z-index: 12` and `z-index: calc(12)` both produce an
`IntegerStyleValue` containing 12.
The goal of this VERIFY was to ensure that we didn't mess up the logic
for calculating the correct type. However, it's now unable to do so
because it doesn't have access to the CalculationContext, which
determines what type percentages are. This makes it crash when running
the simplification algorithm. The benefits of this check are small, and
it meant doing extra work, so let's just remove it.
Calc simplification (which I'm working towards) involves repeatedly
deriving a new calculation tree from an existing one, and in many
cases, either the whole result or a portion of it will be identical to
that of the original. Using RefPtr lets us avoid making unnecessary
copies. As a bonus it will also make it easier to return either `this`
or a new node.
In future we could also cache commonly-used nodes, similar to how we do
so for 1px and 0px LengthStyleValues and various keywords.
To be properly compatible with calc(), the resolved() methods all need:
- A length resolution context
- To return an Optional, as the calculation might not be resolvable
A bonus of this is that we can get rid of the overloads of `resolved()`
as they now all behave the same way.
A downside is a scattering of `value_or()` wherever these are used. It
might be the case that all unresolvable calculations have been rejected
before this point, but I'm not confident, and so I'll leave it like
this for now.
Initially I added this to the existing CalculationContext, but in
reality, we have some data at parse-time and different data at
resolve-time, so it made more sense to keep those separate.
Instead of needing a variety of methods for resolving a Foo, depending
on whether we have a Layout::Node available, or a percentage basis, or
a length resolution context... put those in a
CalculationResolutionContext, and just pass that one thing to these
methods. This also removes the need for separate resolve_*_percentage()
methods, because we can just pass the percentage basis in to the regular
resolve_foo() method.
This also corrects the issue that *any* calculation may need to resolve
lengths, but we previously only passed a length resolution context to
specific types in some situations. Now, they can all have one available,
though it's up to the caller to provide it.
BFC roots behave differently in that their height is computed twice,
before and after inside layout, since automatic height depends on the
results of inside layout. Other formatting contexts only require the
"before" pass, and so we can treat their content sizes as definite
before proceeding with inside layout.
This makes https://play.tailwind.com/ look beautiful. :^)
The focus chain always consists of newly created GC::Root objects, so
the condition always produced `false`. The fix is to use GC::Root's
overloaded operator== method, which compares the pointers of the stored
type.
This fixes Figma dropdowns and context menus instantly disappearing
upon opening them. This is because they become focused when they insert
them. Because of this bug, it would fire blur events all the way up to
and including the window. Figma listens for the blur event on the
window, and when received, it will instantly hide dropdowns and context
menus. The intention behind this seems to be hiding them when the user
clicks off the browser window, or switches tab.
Several interfaces that return a high resolution time require that
time to be coarsened, in order to prevent timing attacks. This
implementation simply reduces the resolution of the returned timestamp
to the minimum values given in the specification. Further work may be
needed to make our implementation more robust to the kind of attacks
that this mechanism is designed to prevent.
Corresponds to https://github.com/whatwg/html/pull/10904
However, we don't implement most of what is changed in that PR, so this
simply moves the FIXME'd getContextAttributes() method from one place
to another.
...until Document::update_style(). This allows to avoid doing full
document DOM tree traversal on each Node::invalidate_style() call.
Fixes performance regression on wpt.fyi
It may happen that the scalars used by SECPxxxr1 turn out to be slightly
smaller than their actual size when serialized to `UnsignedBigInteger`,
especially for P521. Handle this case by serializing zeros instead of
failing.
Originally discovered as a flaky WPT test.
Now that we have implemented most of the element types, this dbgln
is a lot less useful for detecting issues on live sites, and is in
my experience more likely to log cases that are not a FIXME. It also
cleans up some of the log spam from running tests.
Prior to this change, we invalidated all elements in the document if it
used any selectors with :has(). This change aims to improve that by
applying a combination of techniques:
- Collect metadata for each element if it was matched against a selector
with :has() in the subject position. This is needed to invalidate all
elements that could be affected by selectors like `div:has(.a:empty)`
because they are not covered by the invalidation sets.
- Use invalidation sets to invalidate elements that are affected by
selectors with :has() in a non-subject position.
Selectors like `.a:has(.b) + .c` still cause whole-document invalidation
because invalidation sets cover only descendants, not siblings. As a
result, there is no performance improvement on github.com due to this
limitation. However, youtube.com and discord.com benefit from this
change.
...by replacing existing method to check if an element is affected by
invalidation property. It turned out there is no need to check if an
element is affected only by some specific property, so it's more
convenient to have a method that accepts the whole set.
...in hover style invalidation. Instead, pass down a flag that indicates
all subsequent nodes in tree traversal have to be marked for inherited
style update.
By using ancestor filters some selectors could be early rejected
skipping selector engine invocation. According to my measurements it's
30-80% hover selectors depending on the website.
Instead of allocating 3 vectors with size equal to the number of
elements potentially affected by hover:
- for the elements themselves
- for selector match state of each element before hovered node change
- for selector match state of each element after hovered node change
now we allocate none of them, but mark element for style recalculation
as we traverse the tree.
We have an optimization that allows us to invalidate only the style of
the element itself and mark descendants for inherited properties update
when the "style" attribute changes (unless there are any CSS rules that
use the "style" attribute, then we also invalidate all descendants that
might be affected by those rules). This optimization was not taking into
account that when the inline style has custom properties, we also need
to invalidate all descendants whose style might be affected by them.
This change fixes this bug by saving a flag in Element that indicates
whether its style depends on any custom properties and then invalidating
all descendants with this flag set when the "style" attribute changes.
Unlike font relative lengths invalidation, for elements that depend on
custom properties, we need to actually recompute the style, instead of
individual properties, because values without expanded custom properties
are gone after cascading, and it has to be done again.
The test added for this change is a version of an existing test we had
restructured such that it doesn't trigger aggressive style invalidation
caused by DOM structured changes until the last moment when test results
are printed.
Previously, we were generating timestamps relative to the current time
of the monotonic clock. We now generate timestamps relative to the
event loop's last render opportunity time, per the spec.
This means we only need to consider rules from the document and the
current shadow root, instead of the document and *every* shadow root.
Dramatically reduces the amount of rules processed on many pages.
Shaves 2.5 seconds of load time off of https://wpt.fyi/ :^)
When positioning floats against an edge, we are taking all current
relevant floats at that side into account to determine the Y offset at
which to place the new float. However, we were using the margin box
height instead of the absolute bottom position, which disregards the
current float's Y-position within the root, and we were setting the Y
offset to that height, instead of taking the new float's Y position
inside of the root into account.
The new code determines the lowest margin bottom value within the root
of the current floats, and adds the difference between that value and
the new float's Y position to the Y offset.
I see no good reason to keep them out of line, and having the methods
named differently than the property they're updating caused me some
confusion initially.
The URLPattern spec is intended to be implemented inside of LibURL, with
LibWeb only responsible for the IDL conversion layer, in a similar
manner to how URL is implemented.
This currently uses a non spec-compliant property on the Response
object, which represents the time that the Response was created.
Setting this value allows `Performance.timeOrigin` to return a
reasonable value.
Our `UnsignedBigInteger` implementation cannot handle numbers whose
size is not a multiple of 4. For this reason we need to carry the real
size around for P-521 support.
The code was printing one error message only, but multiple can be
generated in one call. Additionally, using this builtin produces
a much more descriptive output.
Fonts on Windows are stored only in %WINDIR%\Fonts and
%LOCALAPPDATA%\Microsoft\Windows\Fonts, see https://stackoverflow.com/a/67078786
And system_data_directories() is not implemented on Windows yet.
These are very hot functions in profiles, so let's avoid a potential
double dynamic_cast or virtual call. For consistency, port all of
these classes of function over to 'as_if' instead.
It is not cheap to do this, so only do it once within this function.
There is definitely some caching that we can do here, but this will
require some smart invalidation to detect _relevant_ changes in
the children.
Instead of traversing the entire DOM subtrees and marking nodes for
style update, this patch adds a new mechanism where we can mark a
subtree root as "entire subtree needs style update".
A new pass in Document::update_style() then takes care of coalescing
all these invalidations in a single traversal of the DOM.
This shaves *minutes* of loading time off of https://wpt.fyi/ subpages.
My previous attempt at resolving the continuation chain tried to deal
with `pointer-events: none` by repeatedly falling back to the parent
paintable until one was found that _would_ want to handle pointer
events. But since we were no longer performing hit-tests on those
paintables, false positives could pop up. This could happen for
out-of-flow block elements that did not overlap with their parent rects,
for example.
This approach works much better since it only handles the continuation
case that's relevant (the "middle" anonymous box) and it does so during
hit-testing instead of after, allowing all the other relevant logic to
come into play.
If Ctrl is pressed when a string is entered into the location bar and
that string doesn't contain a valid TLD, ".com" is added to that string.
Previously, only a small, fixed set of TLDs was checked. We now use the
public suffix list to determine if the given address contains a valid
TLD.
For all invalidation properties nested into nth-child argument list we
need to invalidate whole subtree to make sure style of sibling elements
will be recalculated.
We do not fire `beforeinput` events since other browsers do not seem to
do so either.
The spec asks us to check whether a command's action modified the DOM
tree. This means adding or removing nodes and attributes, or changing
character data anywhere in the tree. We have
`Document::dom_tree_version()` for node updates, but for character data
a new version number is introduced that allows us to easily keep track
of any text changes in the entire tree.
Both Chrome and Firefox return `true` whenever the value string provided
is an invalid color or the current color. Spec issue raised:
https://github.com/w3c/editing/issues/476
Instead of creating and passing around Vector<MatchingRule> inside
StyleComputer (internally, not exposed in API), we now use vectors
of pointers/references instead.
Note that we use pointers in case we want to quick_sort() the vectors.
Knocks 4 seconds of loading time off of https://wpt.fyi/
Instead, change the APIs from "has :foo" to "may have :foo" and return
true if we don't have a valid rule cache at the moment.
This allows us to defer the rebuilding of the rule cache until a later
time, for the cost of a wider invalidation at the moment.
Do note that if our rule cache is invalid, the whole document has
invalid style anyway! So this is actually always less work. :^)
Knocks ~1 second of loading time off of https://wpt.fyi/
Using the raw value meant that 1em would be incorrectly treated as 1px,
for example.
I've updated our canvas-filters test to demonstrate this - without the
code change this would instead have an x-offset of 2px.
This adds formatters for `WebIDL::Exception`, `WebIDL::SimpleException`
and `WebIDL::DOMException`. These are useful for displaying the content
of errors when debugging.
We achieve this by keeping track of the number of HTMLSlotElements
inside each ShadowRoot (do via ad-hoc insertion and removal steps.)
This allows slottables assignment to skip over entire shadow roots when
we know they have no slots anyway.
Massive speedup on https://wpt.fyi/ which no longer takes minutes/hours
to load, but instead a "mere" 19 seconds. :^)
In the very common case that no special constructor options are provided
for the Intl.Collator when calling localeCompare() on a string, we can
cache and reuse a default-constructed Intl.Collator, saving lots of time
and space.
This shaves a fair bit of load time off of https://wpt.fyi/ where they
use Array.prototype.sort() and localeCompare() to sort a big JSON thing.
Time spent in sort():
- Before: 1656 ms
- After: 135 ms
Before this change, it was possible for a second GC to get triggered
in the middle of a first GC, due to allocations happening in the
FinalizationRegistry cleanup host hook. To avoid this causing problems,
we add a "post-GC task" mechanism and use that to invoke the host hook
once all other GC activity is finished, and we've unset the "collecting
garbage" flag.
Note that the test included here only fails reliably when running with
the -g flag (collect garbage after each allocation).
Fixes#3051
Instead of ignoring any paintable immediately when they're invisible to
hit-testing, consider every candidate and while the most specific
candidate is invisible to hit-testing, traverse up to its parent
paintable.
This more closely reflects the behavior expected when wrapping block
elements inside inline elements, where although the block element might
have `pointer-events: none`, it still becomes part of the hit-test body
of the inline parent.
This makes the following link work as expected:
<a href="https://ladybird.org">
<div style="pointer-events: none">Ladybird</div>
</a>
Our layout tree requires that all containers either have inline or
non-inline children. In order to support the layout of non-inline
elements inside inline elements, we need to do a bit of tree
restructuring. It effectively simulates temporarily closing all inline
nodes, appending the block element, and resumes appending to the last
open inline node.
The acid1.txt expectation needed to be updated to reflect the fact that
we now hoist its <p> elements out of the inline <form> they were in.
Visually, the before and after situations for acid1.html are identical.
The existing `::unite_horizontally()` and `::unite_vertically()` tests
did not properly test the edge cases where left/top in the Rect were
updated, so they get re-arranged a bit.
The spec intends to pass through a URL record object as it needs to
be serialized on removal. This has no functional impact on our
implementation other than the double parsing of every URL being
revoked.
It is also missing an error check for an invalid URL being passed
through. This does not impact our implementation currently as we
just end up using an empty URL which is not part of the blob entry
map. This will cause problems once DOMURL::parse is updated to
return an Optional<URL::URL> however.
This is used to put together the list of supported WebGL extensions
based on the available extensions, per-extension required extensions
and WebGL version.
This is done by using the combination of format and type to map to the
appropriate Skia bitmap type. With this, we then read the SkImage of
the TexImageSource into a new SkPixmap with the destination format
information and holding an appropriately sized buffer. Once created,
readPixels is called to convert and write the image into the buffer.
Namely:
- Perform case-insensitive matching
- Return the same extension objects every time
- Only enable the extension if it's supported by the current context
This causes it to enforce the sections "Differences Between WebGL and
OpenGL ES 2.0" from the WebGL 1 specification and "Differences Between
WebGL and OpenGL ES 3.0" from the WebGL 2 specification. It also
disables a bunch of extensions by default, which we must now request
with glRequestExtensionANGLE.
This is required to return original references to the shaders attached
to a program from getAttachedShaders. This is required for Figma (and
likely all other Emscripten compiled applications that use WebGL) to
get it's own generated shader IDs from the shaders returned from
getAttachedShaders.
Use invalidation sets for presentational hint attribute invalidation
instead of falling back to full descendants and siblings invalidation.
The only difference for presentational hint attributes is that we always
have to invalidate the style of element itself.
When we serialize blob URL entries we do not serialize the entire
environment settings object and only use it's origin. To allow
a Blob URL entry to access its relevant storage key, add a getter
that simply takes an origin.
Instead of just putting in members directly, wrap them up in structs
which represent what a URL blob entry is meant to hold per the spec.
This makes more obvious what this is meant to represent, such as the
ByteBuffer being used to represent the bytes behind a Blob.
This also allows us to use a stronger type for a function that needs
to return a Blob URL entry's object.
Specifically, after implementing some recent spec changes to navigables,
we end up calling `get_public_suffix("localhost")` here, which returns
OptionalNone. This would previously crash.
Our get_public_suffix() seems a little incorrect. From the spec:
> If no rules match, the prevailing rule is "*".
> https://github.com/publicsuffix/list/wiki/Format#algorithm
However, ours returns an empty Optional in that case. To avoid breaking
other users of it, this patch modifies Host's uses of it, rather than
the function itself.
Previously most of the calculations for `object-fit` and
`object-position` were based on device pixels, meaning that images would
render differently based on zoom and DPI settings. Instead those
calculations now use css pixels and only the final draw-call is based
on device-pixels.
When an element is displayed as table, an anonymous table wrapper box
needs to be created for it. Among others, the position property of the
table element is then applied to the anonymous table wrapper box
instead. If the table happens to be positioned absolutely, the table
wrapper box may become the containing block for absolutely positioned
elements inside the table.
In the original code however, anonymous layout nodes were excluded from
becoming the containing block for an absolutely positioned element.
Because of this, the containing block was calculated to be the first
suitable parent block of the table wrapper box.
This incorrect containing block would result in a crash later on when
trying to size the absolutely positioned element inside the table. To
prevent this crash, the anonymous table wrapper box is now allowed to
become the containing block for absolutely positioned elements inside
a table.
The definition of containing block for an absolutely positioned element
in the spec does not mention anything about skipping anonymous boxes.
Additionally the rules for absolute positioning of tables
(https://www.w3.org/TR/css-tables-3/#abspos-boxes-in-table-root) imply
that a table wrapper box is indeed able to be the containing block for
absolutely positioned elements.
The "with" statement is its own token (TokenType::With), and thus would
fail to parse as an identifier. We've already asserted that the token
we are parsing is "with" or "assert", so just consume it.
It is currently implemented as a member of CyclicModule. However, as the
spec indicates, this must be invokable with non-CyclicModule modules. In
several of the call sites, we are blindly casting to a CyclicModule;
this will fail for e.g. JSON modules.
Linking a module has assertions about the module's state, namely that
the state is not "new". The state remains "new" if loading the module
has failed. See: https://tc39.es/ecma262/#figure-module-graph-missing
In any case, this exception causes a loading failure, which results
in A's [[Status]] remaining new.
So we must propagate that failure, instead of blindly moving on to the
linking steps.
We currently have a fair amount of code of the form:
if (is<Child>(parent)) {
auto& child = static_cast<Child&>(parent);
if (child.foo()) {
}
}
This new cast allows us to instead write code of the form:
if (auto* child = as_if<Child>(parent); child && child->foo()) {
}
N.B. The name "as_if" was chosen because we are considering renaming
verify_cast to "as".
Co-authored-by: Tim Ledbetter <tim.ledbetter@ladybird.org>
Interestingly, the spec has a note saying:
> window.postMessage() performs StructuredSerializeWithTransfer on
> its arguments, but is careful to do so immediately, inside the
> synchronous portion of its algorithm. Thus it is able to use the
> algorithms without needing to prepare to run script and prepare
> to run a callback.
But there is no note about the deserialization steps. In any case, we do
need callbacks enabled here.
...Which doesn't do anything given start() itself doesn't do anything,
but this is a subtle enough point of the spec that it seems worthwhile
to implement now for whenever this does become meaningful.
For example, running `alert(1)` will pause the event loop, during which
time no JavaScript should execute. This patch extends this disruption to
microtasks. This avoids a crash inside the microtask executor, which
asserts the JS execution context stack is empty.
This makes us behave the same as Firefox in the following page:
<script>
queueMicrotask(() => {
console.log("inside microtask");
});
alert("hi");
</script>
Before the aforementioned assertion was added, we would execute that
microtask before showing the alert. Firefox does not do this, and now
we don't either.
Implements idea described in
https://docs.google.com/document/d/1vEW86DaeVs4uQzNFI5R-_xS9TcS1Cs_EUsHRSgCHGu8
Invalidation sets are used to reduce the number of elements marked for
style recalculation by collecting metadata from style rules about the
dependencies between properties that could affect an element’s style.
Currently, this optimization is only applied to style invalidation
triggered by class list mutations on an element.
We were previously crashing when invoking 'scroll to the fragment' on
such documents as it was unable to find the active document. This is
as a result of our AD-HOC implementation not setting up the document
fully to mark it is active before running the parser.
Fixes a crash on https://tweakers.net.
This is sub-optimal but let's rebuild the whole tree for now, since this
case gets quite complicated and there are more valuable things to chase
after first.
Thanks to Gingeh for the reduced test case!
This updates our local ICU overlay port to use ICU 76.1. This includes
Unicode 16 and CLDR 46.
Upstream vcpkg is not able to supply versions past 74 yet due to various
dependency issues, but we are able to use this version ourselves. The
overlay port now includes a patch to revert ICU's dependence on autoconf
2.72 for now, as this version is not yet available on all systems.
All of the test changes were cross-referenced with Firefox to ensure
correctness.
In ICU 76, the default was changed from "arab" to "latn". See:
c149724509
The whole point of these tests was to use a non-Latin numbering system.
This patch ensures that is the case to make following patches easier to
grok.
This separates the StringBuilder constructor into 2 constructors. This
essentially allows forming code of the following pattern:
StringBuilder foo() { return {}; }
Otherwise, we would get the following compiler error:
chosen constructor is explicit in copy-initialization
Due to the explicitness of the StringBuilder constructor.
This is required for an upcoming update to ICU 76, where we use our
StringBuilder in templated ICU code.
These common cases now cause us to invalidate the layout tree starting
at the relevant parent node instead of invalidating the entire tree.
- DOM node insertion
- DOM node removal
- innerHTML setter
- textContent setter
This makes a lot of dynamic content much faster. For example, demos
on shadertoy.com go from ~18 fps to ~28 fps on my machine.
DOM nodes now have two additional flags:
- Needs layout tree update
- Child needs layout tree update
These work similarly to the needs-style-update flags, but instead signal
the need to rebuild the corresponding part of the layout tree.
When a specific DOM node needs a layout tree update, we try to create
a new subtree starting at that node, and then replace the subtree in the
old layout tree with the newly created subtree.
This required some refactoring in TreeBuilder so that we can skip over
entire subtrees during a tree update.
Note that no partial updates happen yet (as of this commit) since we
always invalidate the full layout tree still. That will change in the
next commit.
Instead just update the existing wrapper with computed values from the
table box, to insure that upside-down "inheritance" works as expected.
This allows table fixup to run on partially updated layout trees without
adding a new layer of unnecessary wrappers every time.
While PendingPullIntos are typically visted by their controller there
were some cases that we were removing those references from the
controller and storing them in a SinglyLinkedList on the stack which
is not safe.
Instead, make PendingPullInto a GC::Cell type, which also allows us
to remove an awkward copy of the struct where the underlying reference
was previously being destroyed.
While we don't yet have a working `using` implementation with our byte
code, we can still keep our DisposableStack implementation up to date.
The changes brought in here are all editorial, and set us up to start
an AsyncDisposableStack implementation.
Previously it only deoptimized the parent scope if the current scope
contains direct eval, which is incorrect because code ran in direct
eval mode has access to the entire scope chain it was executed in.
The fix is to also propagate direct eval's presence if the current
scope is marked as being screwed by direct eval.
This fixes Google's botguard failing to complete on Google sign in, as
it tried to access local variables outside of a direct parent function
with eval, causing it throw "unhandled" exceptions. Unhandled is in
quotes because their bytecode VM _technically_ caught it, but it was
considered an unhandled exception. This was determined by removing get
optimizations and then adding debug output for every get operation.
Using this, I noticed that for these errors, it would access the
'message' and 'stack' properties. This is because their error handler
function noticed this was not a synthesised error, which is never
expected to happen. That was determined by using Chrome Devtools 'pause
on handled exception' feature, and noticing it never threw a '[var] is
not defined' exception, but only synthesized error objects which
contained a sentinel value to let it know it was synthesized.
I added debug output to eval to print out what was being eval'd because
it makes heavy use of eval. This revealed that the exceptions only came
from eval.
I then dumped every generated executable and noticed the variables it
was trying to access were generated as local variables in the top
scope. This led to checking what makes a variable considered local or
not, which then lead to this block of code in ~ScopePusher that
propagates eval presence only to the immediate parent scope. This
variable directly controls whether to create all variables properly
with variable environments and bindings or allow them to be stored as
local registers tied to that function's executable.
Since this now lets botguard run to completion, it no longer considers
us to be an insecure/potential bot browser when signing in, now
allowing us to be able to sign in to Google.
The spec never mentions the possibility for the `hash` member of
`RsaHashedKeyAlgorithm` to be a string, it should be a `KeyAlgorithm`
object containing a `name` string member.
Spec: https://w3c.github.io/webcrypto/#dfn-RsaHashedKeyAlgorithm
If some state has already been tried, skip over it as it would never
lead to a match regardless.
This fixes performance/memory issues in cases like
/(a+)+b/.exec("aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa")
or
/(a|a?)+b/...
Fixes#2622.
Same again, although rotation is more complicated: `rotate`
is "equivalent to" multiple different transform function depending on
its arguments. So we can parse as one of those instead of the full
`rotate3d()`, but then need to handle this when serializing.
The only ways this varies from the `scale()` function is with parsing
and serialization. Parsing stays separate, and serialization is done by
telling `TransformationStyleValue` which property it is, and overriding
its normal `to_string()` code for properties other than `transform`.
https://webaudio.github.io/web-audio-api/#AnalyserNode
Most of the interface is naively implemented. Container types
probably need adjusted (Vector<double> is used for all the processing).
A Fourier Transform is needed, but that's waiting on either a 3rd
party library or a complex number type.
There are lots of simple miscellaneous filters that need to be applied.
It could be reasonable to implement from scratch, supposing that
it can be parallelized. It might be hard to find one library with
everything. Not my call though.
Some additional scaffolding around blocks and render quanta is
probably needed before this is developed much further, which
probably comes in at the level of the AudioNode.
Co-authored-by: Tim Ledbetter <tim.ledbetter@ladybird.org>
The use of this HashMap looks very spooky, but let's at least use
finalize when cleaning them up on destruction to make things slightly
less dangerous looking.
Instead of clamping to the limits allowed by ISOYearMonthWithinLimits,
clamp to the limits allowed by the type we are converting to (i32). This
allows some callers to then reject years outside that range.
In the UTF-8 implementation, this prevents out-of-bounds access of the
underlying text data, as the ICU macro would essentially do something
akin to `text[text.length()]`.
The UTF-16 implementation already checks for out-of-bounds, but would
previously return 0. We now return an empty Optional in both impls. This
doesn't affect LibJS (the user of the UTF-16 impl), as it already does
bounds checking before invoking LibUnicode APIs.
While investigating an issue with Unicode::Segmenter (the result of
which is passed to functions in this file), this missing include was
causing clangd to be unable to find non-virtual implementations.
This commit begins to implement the track processing model. When the
`src` attribute is updated, we now fetch the given source file.
Currently, we always fire an `error` event once fetching is completed,
as we don't support processing the fetched data.
Invalidation for adopted style sheets was broken because we had an
assumption that "active" style sheet is always attached to
StyleSheetList which is not true for adopted style sheets. This change
addresses that by keeping track of all documents/shadow roots that own
a style sheet and notifying them about invalidation instead of going
through the StyleSheetList.
This replaces the old `OAEP` implementation with one backed by OpenSSL.
The changes also include some added modularity to the RSA class by
making the `RSA_EME` and `RSA_EMSE` for encryption/decryption and
signing/verifying respectively.
This commit replaces the old implementation of `EMSA_PKCS1_V1_5` with
one backed by OpenSSL. In doing so, the `sign` and `verify` methods of
RSA have been modified to behave like expected and not just be
encryption and decryption.
I was not able to split this commit because the changes to `verify` and
`sign` break pretty much everything.
It used to be that the caller would supply a buffer to write the output
to. This created an anti-pattern in multiple places where the caller
would allocate a `ByteBuffer` and then use `.bytes()` to provide it to
the `PKSystem` method. Then the callee would resize the output buffer
and reassign it, but because the resize was on `Bytes` and not on
`ByteBuffer`, the caller using the latter would cause a bug.
Additionally, in pretty much all cases the buffer was pre-allocated
shortly before.
Originally we used the `paths-ignore` feature to skip the Lagom jobs if
no source code was changed. Afterwards, we enabled the 'required checks'
feature to prevent merging pull requests that have failing checks.
Unfortunately, marking a check as required means it always needs to be
successfully executed, even if a conditional caused it to be skipped.
By using the `paths-filter` action we were able to add the conditional
to the job instead, theoretically causing the 'required checks' feature
to start working again for documentation-only changes. As it turns out,
not the job, but the steps should get the new conditionals.
As that requires adding an `if` to every individual step and the
conditional execution has caused enough headaches since its
introduction, let's remove it for now and take the unfortunate execution
of CI jobs for granted.
Changed the usage from `add_libweb_test.py test_name.html` to
`add_libweb_test.py test_name.html test_type` with no default,
and supports automatically generating input/output files in the
right directories for test types Screenshot, Text, Ref, and Layout.
Co-authored-by: Sam Atkins <sam@ladybird.org>
`current_property_id()` is insufficient to determine if a quirk is
allowed. For example, unitless lengths are allowed in certain
properties, but NOT if they are inside a calc() or other function. It's
also incorrect when we are parsing a longhand inside a shorthand. So
instead, replace that with a stack of value-parsing contexts. For now,
this is either properties or CSS functions, but in future can be
expanded to include media features and other places.
This lets us disallow quirks inside functions, like we're supposed to.
It also lays the groundwork for being able to more easily determine
what type a percentage inside a calculation should become, as this is
based on the same stack of contexts.
Previously, a crash would occur when attempting to throw an error in
this case because the method used to create the exception tried to get
the current realm from the execution context stack, which is empty. The
realm is now passed explicitly when constructing the error, avoiding
the crash.
Instead of always reporting a colno and lineno of zero try and use the
values from the Error object that may be provided, falling back to the
source location of the invocation if not provided. We can definitely
improve the reporting even more, but this is a start!
Also update this function to latest spec while we're in the area.
This isn't a full fix, as the paint function does not handle this
either. But instead of getting the bitmap from the image source
immediately, follow the spec a bit more closely by creating the
CanvasPatern object with the ImageSource directly.
Fixes a crash for the 5 included WPT tests.
Replace our slow, possibly incorrect RSA key generation with OpenSSL.
This should fix many WPT tests that are timing out because we were too
slow at computing keys.
We added these methods to propagate OOM errors at process startup, but
we longer fret about these tiny OOM failures. Requiring that these init
methods be called prohibits using these strings in processes that have
not set up a MainThreadVM. So let's just remove them and initialize the
strings in a sane manner.
In doing so, this also standardizes how we initialize strings whose C++
variable name differs from their string value. Instead of special-casing
these strings, we just include their string value in the x-macro list.
This makes it more convenient to use the 'relvant agent' concept,
instead of the awkward dynamic casts we needed to do for every call
site.
mutation_observers is also changed to hold a GC::Root instead of raw
GC::Ptr. Somehow this was not causing problems before, but trips up CI
after these changes.
URL::basic_parse has a subtle bug where the resulting URL is not set
to valid when StateOveride is provided and the URL parser early returns
a valid URL.
This has not surfaced as a problem so far, as the only users of the
state override API provide an already valid URL buffer and also ignore
the result of basic parsing with a state override.
However, this bug surfaces implementing the URL pattern spec, which as
part of URL canonicalization:
* Provides a dummy URL record
* Basic URL parses that URL with state override
* Checks the result of the URL parser to validate the URL
While we could set URL validity on every early return of the URL parser
during state override, it has been a long standing FIXME around the code
to try and remove the awkward validity state of the URL class. So this
commit makes the first stage of this change by migrating the basic
parser API to return Optional, which also happens to make this subtle
issue not a problem any more.
This matches the behavior of other browsers, which always set the dirty
checkedness flag when setting checkedness, except when setting the
`checked` content attribute.
Implement the Ed448 curve for signing and verifying using OpenSSL.
The methods could be all made static, but all other curves are not.
I think this is material for further refactoring.
Previously, <a> elements were frequently invalidated because
`set_the_url()` was called by `reinitialize_url()`, which is a
preparation step in every HTMLHyperlinkElementUtils function. As a
result, styles were unnecessarily invalidated each time any of these
functions were invoked without changing the URL.
This change causes explicit role=none and role=presentation attribute
values to be ignored in cases where the elements for which those values
are specified are either focusable, or have global ARIA attributes —
per https://w3c.github.io/aria/#conflict_resolution_presentation_none.
This change implements the role-checking requirement from the ARIA spec
at https://w3c.github.io/aria/#document-handling_author-errors_roles
that the “form” and “region” roles are required to have accessible
names — and that if they don’t have accessible names as required, UAs
must treat them as if they’d not been specified at all.
This change causes explicitly-specified role attributes to be ignored in
the case where the specified role is “orphaned” — that is, when its
element lacks a required ancestor with an appropriate role.
Per https://w3c.github.io/aria/#document-handling_author-errors_roles,
determining whether to ignore certain specified landmark roles requires
first determining whether the element for which the role is specified
has an accessible name.
But if we then try to retrieve a role for such elements, we end up
calling right back into the accessible-name computation code — which
would cause the calls to loop infinitely.
So to avoid that — and to have handling for any other future cases the
spec may introduce of such recursive calls that will loop indefinitely —
this change introduces a parameter that callers can pass to cause
role-attribute lookup to be skipped during accessible-name computation.
This change adds a virtual to_element function to ARIAMixin, and
overrides it in DOM::Element so it can then be used back inside
ARIAMixin to get an element when needed (for example, when computing a
role requires checking the roles of ancestors of an element).
This accurately reflects the spec it's implementing. This algorithm is
used in 5 spots in the spec but the old buggy behavior was never
triggered:
* In both ::extract() and ::clone_the_contents(), invocations to this
method are guarded by a check to see if the start node is the
inclusive ancestor of the end node, or vice versa - effectively
resulting in the inequality checks to be accidentally correct.
* In ::surround_contents(), we forego the usage of this algorithm as
stated in the spec, and instead use a correct and more optimized
version that simply compares the start and end nodes.
A lot of words to say: no functional changes :^)
Functions in AK/GenericShorthands used pass-by-value which results in
copying values when calling those functions. This could be expensive
for complex types. To optimize performance, we switch the functions to
use forwarding references.
Returning numbers instead of booleans for the statuses made Ruffle
(through the wgpu crate) think a shader/program failed to compile/link,
as it does a strict type comparison.
Previously the enforcement was only done on creation. Not enforcing it
on change would cause a crash if the canvas width/height was set to
zero or less.
Required by https://qwasm2.m-h.org.uk, which adds a custom `name`
attribute to objects it generates. It then gets some of these objects
out with getParameter, and expects the `name` attribute to be there.
Rather than partly-converting number, dimension, and ident tokens at the
start of parsing a calculation, and then later finishing it off, we can
just do the whole step in convert_to_calculation_node(). This is a
little less code, but mainly means we are left with only a single use
of the Dimension type in the codebase, so that can be removed soon.
Various places in the spec allow for `<number> | <percentage>`, but this
is either/or, and they are not allowed to be combined like dimensions
and percentages are. (For example, `calc(12 + 50%)` is never valid.)
User code generally doesn't need to care about this distinction, but it
does now need to check if a calculation resolves to a number, or to a
percentage, instead of a single call.
The existing parse_number_percentage[_value]() methods have been kept
for simplicity, but updated to check for number/percentage separately.
An upcoming change requires that we can determine which property we are
parsing before we parse the value. That's the opposite of what this
code previously did, which was to parse a generic dimension or calc()
and then figure out what property would accept it.
When we know what kind of dimension we want, it's awkward to attempt to
parse any dimension type, including quirks that only affect lengths, to
then throw it away unless it's the type we wanted in the first place.
Additionally, move the unitless angle/length behavior for SVG attributes
into these methods, where it belongs.
Instead, only try to parse the type of dimension we want. This is
currently more code, but some could be factored together later.
SVGs are rendered with subpixel precision. As such it can happen that
paths are rendered with less than 1px width or height and that they can
have a bounding box thinner than 1px. Due to an optimization such paths
were ignored when painting because their bounding box was incorrectly
calculated to be empty.
As a result horizontal or vertical lines inside SVGs were missing if:
* The SVG is displayed at viewbox size but the lines are defined with
less than 1px.
* The SVG contians 1px-thin lines, but is displayed at a size smaller
than viewbox size.
To prevent this, the bounding box of the path is now enlarged to contain
all pixels that are partially affected.
Previously, if the NumericCharacterReferenceEnd state was reached when
current_input_character was None, then the
DONT_CONSUME_NEXT_INPUT_CHARACTER macro would restore back before the
EOF, and allow the next state (after the SWITCH_TO_RETURN_STATE) to
proceed with the last digit of the numeric character reference.
For example, with something like `ї`, before this commit the
output would incorrectly be `<code point with the value 1111>1` instead
of just `<code point with the value 1111>`.
Instead of putting the `if (current_input_character.has_value())` check
inside NumericCharacterReferenceEnd directly, it was instead added to
DONT_CONSUME_NEXT_INPUT_CHARACTER, because all usages of the macro
benefit from this check, even if the other existing usage sites don't
exhibit any bugs without it:
- In MarkupDeclarationOpen, if the current_input_character is EOF, then
the previous character is always `!`, so restoring and then checking
forward for strings like `--`, `DOCTYPE`, etc won't match and the
BogusComment state will run one extra time (once for `!` and once
for EOF) with no practical consequences. With the `has_value()` check,
BogusComment will only run once with EOF.
- In AfterDOCTYPEName, ConsumeNextResult::RanOutOfCharacters can only
occur when stopping at the insertion point, and because of how
the code is structured, it is guaranteed that current_input_character
is either `P` or `S`, so the `has_value()` check is irrelevant.
This is mostly a development helper, to move all undefined symbols
in shared libraries to link time rather than load time.
At the same time, set --no-allow-shlib-undefined and -z,defs to
further enforce the rule.
The inspector widget has this functionality, but it's limited to the
site you're currently viewing. This commit adds an option for removing
all cookies globally.
Previously, the workaround was to open a sqlite shell and run:
`DELETE FROM Cookies;` on the database yourself.
When setting the textContent of an element with no children to null or
the empty string, nothing happens. Even so, we were still invalidating
style, layout and collections, causing pointless churn.
Skipping invalidation in this case also revealed that we were missing
invalidation when changing the selected state of HTMLOptionElement.
This was all caught by existing tests already in-tree. :^)
As many other projects, Ladybird has more than a few TODO and FIXME
comments sprinkled throughout the codebase. These can be a great
inspiration or starting point for contributors looking for work, so
we mention them in the docs on contributing to the project.
It is possible to skip inherited style recalculation for children if
parent's recalculation does not cause any changes.
Improves performance on Github where we could avoid dozens of inherited
style calculations that do not produce any visible changes.
Previously, we optimized hover style invalidation to mark for style
updates only those elements that were matched by :hover selectors in the
last style calculation.
This change takes it a step further by invalidating only the elements
where the set of selectors that use :hover changes after hovered element
is modified. The implementation is as follows:
1. Collect all elements whose styles might be affected by a change in
the hovered element.
2. Retrieve a list of all selectors that use :hover.
3. Test each selector against each element and record which selectors
match.
4. Update m_hovered_node to the newly hovered element.
5. Repeat step 3.
6. For each element, compare the previous and current sets of matched
selectors. If they differ, mark the element for style recalculation.
Instead of recalculating styles for all nodes in the common ancestor of
the new and old hovered nodes' subtrees, this change introduces the
following approach:
- While calculating ComputedProperties, a flag is saved if any rule
applied to an element is affected by the hover state during the
execution of SelectorEngine::matches().
- When the hovered element changes, styles are marked for recalculation
only if the flag saved in ComputedProperties indicates that the
element could be affected by the hover state.
Previously, `percentage_of` would be called on the previous value,
potentially changing its numeric type, yet this potential change
was never reflected as the old numeric type was always used. Now,
the numeric type will be re-calculated every time after the
percentage is resolved. As well, VERIFY checks have been placed to
uphold the requirements for the numeric types to match what the
actual values are.
The "strictly split" infra algorithm feels like an inefficient way of
doing basically what our existing split() does, except working with
code points instead of bytes. It didn't seem worth it to implement now.
Unfortunately, there is no explicit and step-by-step spec to perform
the serialization of `color()` declared values, so while being
spec-informed, this is quite ad-hoc.
Fixes 81 subtests in:
- css/css-color/parsing/color-valid-color-function.html
Previously, a method with multiple sequence arguments would cause a
compile error, as the same variable name was redeclared when iterating
the sequence for each argument. This change disambiguates these
variable names.
This brings keyboard shortcuts for the inspector up with common
convention in FF and Chrome: Ctrl+Shift+C now also opens the inspector,
and F12, Ctrl+W, and Ctrl+Shift+I now close the inspector when the
inspector window is focused.
Resolves#972
Some websites (Reddit, for example) create lots of temporary documents
for fragment parsing. Before this change, we had to build a rule cache
for these documents just to determine whether there are :has, :defined,
or attribute selectors, while it should be safe to simply return `false`
right away.
Which will have proper handling of the exection context when performing
a microtask checkpoint. This fixes an assertion to be added in the next
commit where perform_a_microtask_check is invoked on a non-empty
execution context stack.
Instead of storing all storage objects in static memory, we now
follow the the spec by lazily creating a unique Storage object
on each document object.
Each Storage object now holds a 'proxy' to the underlying backing
storage. For now, this proxy is simply a reference to the backing
object. In the future, it will need to be some type of interface
object that stores on a SQLite database or similar.
Session storage is now correctly stored / tracked as part of the
TraversableNavigable object.
Local storage is still stored in a static map, but eventually this
should be factored into something that is stored at the user agent
level.
This is no longer done. One of the comments is also innacurate for a
second reason - the call stack is never empty in that case, and is
verified as such only a few lines above.
From what I understand, the suspension steps are not required now,
or in the future for our implementation, or any other. The intent
is already implemented in the spec pushing on another execution
context to the stack and leaving the running execution context as-is.
The resume steps are a slightly different story as there is some subtle
behavior which the spec is trying to convey where some custom logic may
need to be done when one execution context changes from one to another.
It may be worth implementing those steps at a later point in time so
that this behavior is a bit easier to follow in those cases.
To make the situation more confusing - from what I can gather from the
spec, not all cases that the spec mentions resume actually means
anything normative. Resume is only _actually_ needed in a limited set
of locations.
For now, let's just remove the unneeded FIXMEs that indicate that there
is something to be done for the suspension steps, as there is not, and
leave the resume steps as is.
This is consistent with other functions such as
HTMLElement::offset_width and fixes a crash for the included test.
Returning an offset of zero is not correct for this case, but this is
still an improvement to not crash.
This fixes a bug where, if a non-existent font family is specified in
CSS, whitespaces would be rendered using the emoji font, while letters
would use the default font. This issue occurred because the font was
resolved separately for each code point. Since the emoji font was listed
before the default font, it was chosen for whitespace characters due to
its inclusion of whitespace glyphs (at least in the Apple Color Emoji
font on macOS). This change resolves the issue by placing the default
font before the emoji font in the list.
This change fixes the IDLGenerators.cpp implementation of the “If
reflectedTarget's explicitly set attr-element is a descendant of any of
element's shadow-including ancestors” step from the “If a reflected IDL
attribute has the type T?, where T is either Element or an interface
that inherits from Element” case in the HTML spec at
http://whatwg.org/html/#reflecting-content-attributes-in-idl-attributes
This change updates the BindingsGenerator/IDLGenerators.cpp code to
handle reflected non-HTML::AttributeNames content-attribute names, and
to handle such names as-is — including names with dashes.
This avoids Fedora 41 x86_64 machines using the x64-linux triplet
when building pkg-config. Doing so without our custom linker flags
causes the build to fail.
In particular:
- Don't compute DOM node editability if we don't need it. This was 22%
of CPU time when scrolling on Wikipedia.
- Defer inversion of transformed coordinates until we actually need
them, after we've performed early returns.
This event is fired while both the previous and the current phase are
active.
This prevents this test from timing out:
- css/css-animations/animationevent-types.txt
The associated animations list might be modified on the time change
event. This means that we can't safely iterate over the hashmap during
this period.
This fixes a crash in:
- css/css-animations/CSSAnimation-effect.tentative.html
Previously, the`HTMLInputElement.selectinStart` and
`HTMLInputElement.selectionEnd` IDL setters, and the
`setRangeText()` IDL method were used when updating an input's value
on keyboard input. These methods can't be used for this purpose,
since selection doesn't apply to email type inputs. Therefore, this
change introduces internal-use only methods that don't check whether
selection applies to the given input.
The output of `WPT.sh list-tests` includes test variants, which vary
only by their query string. Since we don't care about this when
importing tests, ignore any query strings and ensure duplicates are
removed from the given test paths.
We now ignore files imported from WPT, if they are in the root `common`
directory, or any directory named `resources`. This matches the
behavior of the WPT test harness.
This change updates the bindings generator for the case defined at
https://html.spec.whatwg.org/#reflecting-content-attributes-in-idl-attributes:element;
that is, the case “If a reflected IDL attribute has the type T?, where T
is either Element or an interface that inherits from Element”.
The change “normalizes” the generator behavior for that case — such that
the generated code expects a getter with a name of the form used in
other cases; e.g., popover_target_element().
Otherwise, without this change, the generator expects a name of the form
get_popover_target_element() for that case.
This change adds computation of ARIA roles for a number of SVG elements
for which, if the element meets the SVG spec criteria for inclusion in
the accessibility tree, the computed ARIA role should be
“graphics-symbol”, and should otherwise be “generic”.
This change also adds similar role computation for the SVG foreignObject
element (the role for which, if the element meets the SVG spec criteria
for inclusion in the accessibility tree, should be “group”, and should
otherwise be “generic”).
This mistakenly implemented the 'piped to' operation on ReadableStream.
No functional difference as the caller was doing the extra work already
of 'piped through' vs 'piped to'.
As far as I can tell there is no change in WPT from this implementation.
But these steps should be more effecient than using the non BYOB steps
as we can make direct use of the provided buffer to the byte stream.
1. Stop using GC::Root in member variables, since that usually creates
a realm leak.
2. Stop putting OrderedHashMap<FlyString, GC::Ptr> on the stack while
setting these up, since that won't protect the objects from GC.
This change ensures that when an accessible name is computed from
multiple labels, the parts computed from each label are separated by
spaces. Otherwise, without this change, the parts are run together in
the accessible name, with no space in between.
For example, https://locals.com/site/discover has a script with an
object of the form:
var f = {
parser: {
sync() {},
async async() {},
}
};
We were previously throwing a syntax error on the async function, as we
specifically did not allow using "async" as a function name here.
Otherwise we will fully read from the cached response and invalidate
it's stream, invalidating it for the next time it is read from. Fixes
a crash when reloading linegoup.lol after two reloads.
I agree that the spec definition of this function isn't super clear
about that, but from "Web Animations 1 - 4.5. Animations"[1]:
An animation is a timing node that binds an animation effect child,
called its associated effect, to a timeline parent so that it runs. Both
of these associations are optional and configurable such that an
animation can have no associated effect or timeline at a given moment.
[1]: https://drafts.csswg.org/web-animations-1/#animations
One day we'll have an eviction strategy, too, but for now let's not
allow these to get collected.
Co-Authored-By: Gingeh <39150378+Gingeh@users.noreply.github.com>
This change fixes selector matching for non-HTML elements that have
mixed-case names — such as the SVG foreignObject element.
Otherwise, without this change, attempting to use a selector to match
such an element — e.g., document.querySelector("foreignObject") — fails.
If there are no :defined pseudo-class selectors anywhere in the
document, we don't have to invalidate style at all when an element's
custom element state changes.
The goal here is to ensure we check for audio backends in a way that
makes sense. On macOS, let's just always use Audio Unit (and thus avoid
any checks for Pulse, to reduce needless/confusing build log noise). We
will also only use the Qt audio backend if no other backend was found,
rather than only checking for Pulse.
Rather than conditionalizing creating an audio plugin, just let the
audio backend factory return an error on its own. If an audio plugin
is not found, we will get a similar error to what is removed here.
Instead of having to duplicate the audio stream backend conditions, just
define PlaybackStream::create in each audio backend implementation file.
We provide a weak definition in PlaybackStream.cpp as the fallback.
This adds a thin wrapper to LibCrypto for generating cryptographically
secure random values and replaces current usages of PRNG within
LibCrypto as well.
Many times, attribute mutation doesn't necessitate a full style
invalidation on the element. However, the conditions are pretty
elaborate, so this first version has a lot of false positives.
We only need to invalidate style when any of these things apply:
1. The change may affect the match state of a selector somewhere.
2. The change may affect presentational hints applied to the element.
For (1) in this first version, we have a fixed list of attribute names
that may affect selectors. We also collect all names referenced by
attribute selectors anywhere in the document.
For (2), we add a new Element::is_presentational_hint() virtual that
tells us whether a given attribute name is a presentational hint.
This drastically reduces style work on many websites. As an example,
https://cnn.com/ is once again browseable.
There's one failing due to the constructor object not having the name
"Global" vs "WebAssembly.Global". This also doesn't include the
tentative test for the type property.
Some WPT tests expect that if you go out of your way to
Object.getOwnPropertyDescriptor on an interface object and extract
the setter out of it, that they throw when called with no arguments.
When the `style` attribute changes, we only need to update style on the
element itself (unless there are [style] attribute selectors somewhere).
Descendants of the element don't need a full style update, a simple
inheritance propagation is enough.
We can now mark an element as needing an "inherited style update" rather
than a full "style update". This effectively means that the next style
update will visit the element and pull all of its inherited properties
from the relevant ancestor element.
This is now used for descendants of elements with animated style.
Becuase we're using dynamic libraries, our configuration is
classified as a "community triplet". To not confuse vcpkg
maintainers when developers create bug reports, name them
properly. This means that the default triplet detection is now
kind of useless, so we have to invent our own for these triplets.
This required multiple changes:
- Make hashes non-copiable because they contain a heap allocated pointer
- Reference classes via `NonnullOwnPtr` only (they are non-copiable)
- Drop all existing hashes implementations
- Use the `OpenSSLHashFunction` base class to implement the same hashes
I was not able to come up with a way to divide this commit into multiple
without increasing the amount of changes.
Nothing breaks with this commit!
The previous VERIFY statement incorrectly asserted that the
interception state was not "committed" or "scrolled". Updated
the condition to ensure the interception state is either
"committed" or "scrolled" as intended.
if (size <= 1)
return;
This means that size is at the very minimum 2,
and pivot_point at the very minimum equals 1 as size / 2;
The condition if (pivot_point) can never be false.
Before this change, StyleComputer would essentially take a DOM element,
find all the CSS rules that apply to it, and resolve the computed value
for each CSS property for that element.
This worked great, but it meant we had to do all the work of selector
matching and cascading every time.
To enable new optimizations, this change introduces a break in the
middle of this process where we've produced a "CascadedProperties".
This object contains the result of the cascade, before we've begun
turning cascaded values into computed values.
The cascaded properties are now stored with each element, which will
later allow us to do partial updates without re-running the full
StyleComputer machine. This will be particularly valuable for
re-implementing CSS inheritance, which is extremely heavy today.
Note that CSS animations and CSS transitions operate entirely on the
computed values, even though the cascade order would have you believe
they happen earlier. I'm not confident we have the right architecture
for this, but that's a separate issue.
This improves the quality of our font rendering, especially when
animations are involved. Relevant changes:
* Skia fonts have their subpixel flag set, which means that individual
glyphs are rendered at subpixel offsets causing glyph runs as a
whole to look better.
* Fragment offsets are no longer rounded to whole device pixels, and
instead the floating point offset is kept. This allows us to pass
through the floating point baseline position all the way to the Skia
calls, which already expected that to be a float position.
The `scrollable-contains-table.html` ref test needed different table
headings since they would slightly inflate the column size in the test
file, but not the reference.
Previously this was proxying the call through javascript, which lead to
unexpected crashes when functions returned things that js-api did not
like.
This commit also adds in the spec comments and fixes a few inaccuracies
that were present in the process.
The spec doesn't say they should exist, so we should not
`VERIFY_NOT_REACHED()` when they don't. Prevents a crash in the WPT
`editing/event.html` tests.
The algorithm referenced to in the Editing spec whenever they talk about
obtaining the "resolved" style or value is actually implemented in
ResolvedCSSStyleDeclaration, so use that instead of going directly to
the computed styles.
We were using the anchor_node() as the boundary point node when
collapsing a selection, but the spec tells us to use the start and end
boundary point nodes.
When we originally implemented calc(), the result of a calculation was
guaranteed to be a single CSS type like a Length or Angle. However, CSS
Values 4 now allows more complex type arithmetic, which is represented
by the CSSNumericType class. Using that directly makes us more correct,
and allows us to remove a large amount of now ad-hoc code.
Unfortunately this is a large commit but the changes it makes are
interconnected enough that doing one at a time causes test
regressions.
In no particular order:
- Update our "determine the type of a calculation" code to match the
newest spec, which sets percent hints in a couple more cases. (One of
these we're skipping for now, I think it fails because of the FIXMEs
in CSSNumericType::matches_foo().)
- Make the generated math-function-parsing code aware of the difference
between arguments being the same type, and being "consistent" types,
for each function. Otherwise those extra percent hints would cause
them to fail validation incorrectly.
- Use the CSSNumericType as the type for the CalculationResult.
- Calculate and assign each math function's type in its constructor,
instead of calculating it repeatedly on-demand.
The `CalculationNode::resolved_type()` method is now entirely unused and
has been removed.
Took the opportunity to pull out a helper function for
entry_with_value_1_while_all_others_are_0(), too.
A couple of these require us to have extra contextual information about
what type percentages should resolve to. Until we have that available,
these are left as FIXMEs with a rough approximation.
This reverts commit 76daba3069.
We're going to need separate types for the JS-exposed style values, so
it doesn't make sense for us to match their names with our internal
types.
By doing that we eliminate the need for the vertical flip flag.
As a side effect it fixes the bug when doing:
`canvasContext2d.drawImage(canvasWithWebGLContext, 0, 0);`
produced a flipped image because we didn't account for different origin
while serializing PaintingSurface into Gfx::Bitmap.
Visual progress on https://ciechanow.ski/curves-and-surfaces/
start_loading_next_url() is a no-op if there's a pending resource load,
but not if that resource load has successfully loaded already. There is
a delay between the font resource loading, and it being processed into
a vector font. Calling font_with_point_size() in that gap would
previously erase the previously-loaded font, if the font had multiple
URLs to choose from.
This fixes the icon font on mods.factorio.com :^)
When applied, corresponding interface object will not be exposed on the
global object, e.g. for the following IDL:
```
Exposed=(Window,Worker), LegacyNoInterfaceObject]
interface ANGLE_instanced_arrays {
};
```
executing `"ANGLE_instanced_arrays" in window` will return `false`.
PlatformObjects with named properties does not qualify as 'has own
property' just by virtue of a named property existing.
This fixes at least one WPT test, which is imported.
This change does replacement of ARIA roles that have newer synonyms.
There are a number of newer ARIA roles that are synonyms for older
roles. https://wpt.fyi/results/wai-aria/role/synonym-roles.html has a
number of subtests which expect that when retrieving the value of an
explicitly- specified role attribute, if the value is one of the older
role values, implementations must replace that with its newer synonym.
This commit enables warnings when using variable length arrays. For
disabling this warning locally use -Wno-vla, as it will work both across
Clang and GCC.
There are some special values for CSS::Selector::PseudoElement::Type
which are after `KnownPseudoElementCount` and therefore not present in
various arrays of pseudo elements, this leads to some errors, if a type
after `KnownPseudoElementCount` is used without checking first. This
adds explicit checks to all usages
Ensure becomes `m1` greater than `m2` even when smaller by more than
one `p`. Since the next operations on `m1` are modulus `p` we can add it
as many times as it's needed.
The official WPT runner supports a `<meta name=timeout content=long>`
tag to let tests opt-in to a longer timeout. Modify our harness to pass
that custom timeout to our runner, so that we don't incorrectly time
out if our default time is shorter than the requested one.
Some tests take longer than others, and so may want to set a custom
timeout so that they pass, without increasing the timeout for all other
tests. For example, this is done in WPT.
Add an `internals.setTestTimeout(milliseconds)` method that overrides
the test runner's default timeout for the currently-run test.
If we have a valid PNG header with geometry info etc, we should still
display it as *something*, even if the image data itself is missing or
corrupted.
This matches the behavior of other browsers, and is something that
Cloudflare Turnstile checks for.
To achieve this, we split the PNG decoder's initialization into two
steps: "everything except reading frame data" and "reading frame data".
If the latter step fails, we yield a transparent bitmap with the
geometry from the PNG's IHDR chunk.
This would fail with EINVAL earlier, due to an attempt to create a
zero-length Core::AnonymousBuffer.
We fix this by transferring the buffer length separately, and only
going down the AnonymousBuffer allocation path if the length is
non-zero.
When message encoding failed for some reason, we'd just swallow the
error without saying a word, and carry on without sending anything.
This led to some very confusing situations.
`WebIDL::invoke_callback()` now takes an `exception_behavior`
parameter, which can be set to `report` to report an exception before
returning. Setting `exception_behavior` to `rethrow` retains the
previous behavior.
This way we could be sure that context object won't be deallocated
before any of the objects that belong to it.
Having a context pointer is also going to be used in upcoming changes
to generate an INVALID_OPERATION error if an object does not belong to
the context it's being used in.
Make sure that `HashAlgorithmIdentifier` is passed through
`normalize_an_algorithm` to verify that the hash is valid and supported.
This is required by the spec, but we are not following it very strictly
in `normalize_an_algorithm` because it is pretty convoluted.
Fixes ~60 tests.
CSS filters work similarly to canvas filters, so it makes sense to have
Gfx::Filter that can be used by both libraries in an analogous way
as Gfx::Color.
Without this, a crashing ref test is able to take down the entire
process because of the `VERIFY(!m_pending_screenshot);` in
`take_screenshot()`. The dialog/prompt fields were not causing crashes
but clearing them feels more hygienic.
Add a formatter output to the flake (`nix fmt`), along with moving +
renaming the devshell so it will work by running `nix-shell` in the root
of the project.
Previously we created a tree of CalculationNodes with dummy
UnparsedCalculationNode children, and then swapped those with the real
children. This matched the spec closely but had the unfortunate
downside that CalculationNodes couldn't be immutable, and couldn't know
their properties at construct-time. UnparsedCalculationNode is also a
footgun, as if it gets left in the tree accidentally we would VERIFY().
So instead, let's parse the calc() tree into an intermediate format, and
then convert each node in that tree, depth-first, into its
corresponding CalculationNode. This means each CalculationNode knows
what its children are when it is constructed, and they never change.
Apart from deleting UnparsedCalculationNode, we can also get rid of the
for_each_child_node() method that was only used by this "replace the
children" code.
Fixes multiple slightly wrong behaviours of the `deriveBits` method
across various algorithms. Some of them might be due to a spec update.
Add tests related to fixes.
Apart from the fact that this workflow is failing every time, we don't
need to check the branch we're on since it's only invoked for the
default branch of the repository. We can also remove the `always()`
since there are no job dependencies nor do we want this to be
uncancelable.
Matching `master` can be a bit simpler and prevent invoking the job
altogether. Additionally, we don't need an `always()` here: we don't
have job dependencies and we want these jobs to be cancelable.
We need to invoke the CI jobs every time, even if nothing relevant has
changed, because we marked them as required status checks for PRs. If
they are not invoked, the associated status checks remain in a 'pending'
state indefinitely not allowing us to (auto-)merge the PR.
This action adds and removes the new "conflicts" label to indicate
whether a pull request is in need of conflict resolution. This both
automatically informs the author of the PR that this is the case, if
they have notifications enabled that is, and makes for an easier
evaluation of the PR queue.
isomorphic encoding a value that has already been encoded will
result in garbage data. `response_headers` is already encoded in
ISO-8859-1/latin1, we cannot use `from_string_pair`, as it triggers
ISO-8859-1/latin1 encoding.
Follow-up of https://github.com/LadybirdBrowser/ladybird/pull/1893
Add support for AES-KW for key wrapping/unwrapping. Very similar
implementation to other AES modes.
Added generic tests for symmetric import and specific AES-KW ones.
Adds ~400 test passes on WPT. Now we do better than Firefox in
`WebCryptoAPI/wrapKey_unwrapKey`!
Auto popovers now correctly establish a close watcher when shown.
This means popovers now correctly close with an escape key press.
Also correctly hide open popovers when removed from the document.
We do not concern ourselves with small OOM handling any longer. But the
main point of this patch is that we are for some reason getting a build
error with clang-19 here about ignoring a nodiscard return type inside
JsonArray::try_for_each.
An AK::String works fine for a USVString as a USVString is just a more
strict version of DOMString. Maybe we will have a different String type
for it in the future, but for now using an AK::String is fine and we do
not need this FIXME.
This change imports the remaining HTML-AAM tests from WPT that haven’t
yet been imported in any previous PRs — giving us complete in-tree
regression-testing coverage for all available WPT tests for the
requirements in the HTML-AAM spec.
This implements the last WebCryptoAPI methods `wrapKey` and `unwrapKey`.
Most of the functionality is already there because they rely on
`encrypt` and `decrypt`. The only test failures are for `AES-GCM` which
is not implemented yet.
The ASN1 structure for PCKS#8 was wrong and missing one wrapping of the
key in a OctetString.
The issue was discovered while implementing `wrapKey` and `unwrapKey` in
the next commits.
The ASN1 structure for PCKS#8 was wrong and missing one wrapping of the
key in a OctetString.
The issue was discovered while implementing `wrapKey` and `unwrapKey` in
the next commits.
The presence of padding in the base64 fields made plenty of WPT tests
fail. Additionally, export was performed with the wrong public key.
The issue was discovered while implementing `wrapKey` and `unwrapKey` in
the next commits.
The presence of padding in the base64 fields made plenty of WPT tests
fail.
The issue was discovered while implementing `wrapKey` and `unwrapKey` in
the next commits.
Previously, if `nullptr` was passed as params for
`wrap_in_private_key_info` or `wrap_in_subject_public_key_info` an ASN1
null was serialized. This was not the intended behaviour for many.
The issue was discovered while implementing `wrapKey` and `unwrapKey` in
the next commits.
The ASN1 structure for PCKS#8 was wrong and missing one wrapping of the
key in a OctetString.
The issue was discovered while implementing `wrapKey` and `unwrapKey` in
the next commits.
The presence of padding in the base64 fields and the typo made plenty of
WPT tests fail.
The issue was discovered while implementing `wrapKey` and `unwrapKey` in
the next commits.
The property was not accessible because it was not exposed to JS.
The issue was discovered while implementing `wrapKey` and `unwrapKey` in
the next commits.
The validation of the key size and specified algorithm was out of spec.
It is now implemented correctly like in `AesCbc`.
The issue was discovered while implementing `wrapKey` and `unwrapKey` in
the next commits.
The textbook RSA decryption method of `c^d % n` is quite slow. If the
necessary parameters are present, the CRT variant will be used.
Performing RSA decryption this way is ~3 times faster.
- Removed the constructor taking a (n, d, e) tuple and moved
it to `RSAPrivateKey`
- Removed default constructor with key generation because it was always
misused and the default key size is quite small
- Added utility constructors to accept a key pair, public key, private
key or both
- Made constructor parameters const
- Updated test to use generated random keys where possible
The previous implementation of `ModularInverse` was flaky and did not
compute the correct value in many occasions, especially with big numbers
like in RSA.
Also added a bunch of tests with big numbers.
The trimmed cache length of the `UnsignedBigInteger` was not reset after
an `add_into_accumulator_without_allocation` operation because the
function manipulates the words directly.
This meant that if the trimmed length was calculated before this
operation it would be wrong after.
The function AnimationEffect::phase() contained duplicated condition
checks for the animation phase determination. This refactor eliminates
the redundant checks by simplifying the logic.
Even though calling delete on a super property will ultimately throw a
ReferenceError, we must return the allocated register for the result of
the delete operation (which would normally be a boolean). If the delete
operation is used in a return statement, the bytecode generator for the
return statement must be able to assume the statement had some output.
In order for public/private key serialization to work correctly we must
store the size of the key because P-521 cannot be stored as full words
inside `UnsignedBigInteger` and therefore is exported as the wrong
length (68 instead of 66).
This makes it also possible to refactor some methods and cleanup
constants scattered around.
Gets almost all import/export tests, expect the JWK ones that calculate
the public key on export. The `SECPxxxr1` implementation currently fails
to do calculations for P-521.
Define SECP521r1 with its constants. Since the parameters cannot be
represented as full bytes, a slight modification has been added to the
byte size.
The current implementation of SECPxxxr1 does not work with this curve.
Tests with different combinations of missing width, height
and viewBox.
All tests confirmed to work on Ladybird:
- exactly the same as Chromium (131.0.6778.85)
- almost the same as Firefox (129.0.2)
- only difference: standalone-w.svg: same size, different alignment
For example, for the following `includes` line in a test262 file:
includes: [sm/non262-TypedArray-shell.js, sm/non262.js]
We currently parse and execute each file in this list as its own script,
in the order they appear in the list.
Tests have recently been imported test262 from SpiderMonkey which fail
with this behavior. In the above example, if the first script references
some function from the second script, we will currently fail to execute
that harness file.
This patch changes our behavior to concatenate all harness files into a
single script, which satisfies the behavior required by these new tests.
This is how test262.fyi and other test262 runners already behave.
Make use of TRY semantics a bit more. And we don't need to store harness
files as a ByteString - we can store the contents as the ByteBuffer that
we receive from reading the file.
In conformance with the requirements of the spec PR at
https://github.com/whatwg/html/pull/9546, this change adds support for
the “switch” attribute for type=checkbox “input” elements — which is
shipping in Safari (since Safari 17.4). This change also implements
support for exposing it to AT users with role=switch.
Due to optimiser shenanigans in the tree alternative form, some
JumpNonEmpty ops might be moved before their Checkpoint instruction.
It is safe to assume the distance between the nonexistent checkpoint and
the current op is zero, so just do that.
Repeat's 'offset' field is a bit odd in that it is treated as a negative
offset, causing a backwards jump when positive; the optimizer didn't
correctly model this behaviour, which caused crashes and misopts when
dealing with Repeats.
This commit fixes that behaviour.
That is what the spec calls it, at least.
In code, this manifests as making the offset very aware
of the element's transform, because the click position comes
relative to the viewport, not to the transformed element.
These were converted to lambdas in
6b921e91d4ce4791c67bf467a2ba519b8e3ca88b
But I merged fcf6cc27f2 without checking
that the code had responded to the change.
What `latest-stable` means exactly also depends on the image version
we're running the workflow on, and unfortunately this can vary wildly
between GitHub runners.
Fixate the version to 16.1 for now. This version will need to be updated
as soon as we want to increase the minimum supported compiler version.
The popoverTargetElement seems to be one of the only cases of a
reflected Element? attribute in the HTML spec, the behaviour of which
is specified in section 2.6.1.
Buttons can't actually toggle popovers yet because showing/hiding
popovers is not implemented yet.
Additionally: For “img” elements with empty “alt” attributes, change the
default role to the newer, preferred “none” synonym for the older
“presentation” role; import https://wpt.fyi/results/html-aam/roles.html
(which provides test/regression coverage for these changes).
Currently, the following JS snippet will hang indefinitely:
new DOMParser().parseFromString("<object>", "text/html");
Because the document into which the object is inserted is not active. So
the task queued to run the representation steps will never run.
This patch implements the spec steps to rerun the representation steps
when the active state changes, and avoid the hang when the object is
created in an inactive document.
We currently (sometimes) copy the observer map to a vector for iteration
to ensure we are not iterating over the map if the callback happens to
remove the observer. But that list was not protected from GC.
This patch ensures we protect that list, and makes all document observer
notifiers protected from removal during iteration.
This change updates the (advanced) build docs to explain how to do a
Debug build with the CXX `-O0` option set — which tells the compiler to
build with no optimizations at all.
Otherwise, Debug builds use the `-Og` option — which, when trying to
check frame variables in a debugger can result in an error of this form:
> error: Couldn't look up symbols: __ZN2AK6Detail10StringBaseD2Ev
> Hint: The expression tried to call a function that is not present in
> the target, perhaps because it was optimized out by
> the compiler.
This change aligns the default roles for “th” and “td” elements with the
requirements in the HTML-AAM spec, and with the corresponding WPT tests
at https://wpt.fyi/results/html-aam/table-roles.html, and with the
behavior in other engines.
Otherwise, without this change, the default role values for “th” and
“td” elements in some cases don’t match the behavior in other engines,
and don’t match the expected results for the corresponding WPT tests.
This change makes Ladybird conform to the current requirements at
https://w3c.github.io/core-aam/#roleMappingComputedRole in the “Core
Accessibility API Mappings” spec for the case of “orphaned” li elements;
that is, any li element which doesn’t have a role=list ancestor.
The core-aam spec requires that in such cases, the li element must not
be assigned the “listitem” role but instead must be treated as if it had
no role at all.
This patch ensure Headers object's associated header list
is ISO-8859-1 encoded when set using `Infra::isomorphic_encode`,
and correctly decoded using `Infra::isomorphic_decode`.
Follow-up of https://github.com/LadybirdBrowser/ladybird/pull/1893
These are non-standard and only needed internally as implementation
details in the implementation of AbstractOperations, so let's keep them
at a file-local level.
Before, libpng would use its own internal logging mechanism to print
non-fatal errors and warnings to stdout/stderr. This made it confusing
when trying to search the Ladybird codebase for those messages as they
didn't exist.
This commit uses `png_set_error_fn` from libpng to redirect those
messages to our own custom logging functions instead.
This eliminates the use of ResourceLoader in HTMLObjectElement. The spec
steps around fetching have been slightly updated since we've last looked
at this, so those are updated here.
Regarding the text test change: we cannot rely on the data: URL being
fetched synchronously. It will occur on a deferred task now. This does
match the behavior of other browsers, as they also will not have run the
fallback representation steps as of DOMContentLoaded.
There are essentially 3 URL parsing AOs defined by the spec:
1. Parse a URL
2. Encoding parse a URL
3. Encoding parse a URL and serialize the result
Further, these are replicated between the Document and the ESO.
This patch defines these methods in accordance with the spec and updates
existing users to invoke the correct method. In places where the correct
method is ambiguous, we use the encoding parser to preserve existing ad-
hoc behavior.
Previously, imported WPT tests didn't display any output if the
internals object was exposed. This change adds the condition that the
browser must also be running headlessly for test output to not be
displayed.
We are currently returning LibJS's invalid code point message, but not
formatting it with the bad value. So we get something like:
Unhandled JavaScript exception: [TypeError] Invalid code point {},
must be an integer no less than 0 and no greater than 0x10FFFF
So not only is the error unformatted, but it's inaccurate; in this case,
the byte cannot be larger than 255.
The WebIDL for the `Headers` constructor specifies that the `init`
parameter is optional and must be of type `HeadersInit`. While the
parameter can be omitted (or explicitly set to `undefined`),
`null` is not a valid value.
This change fixes at least 2 "Create headers with null should throw"
WPT subtests which I have imported in this patch.
It is the responsibility of code that deals with TypedArrays to apply
the byte offset and byte length. Not doing this caused Unity Web to
crash, as they call getRandomValues with views into their full main
memory. Previously, it would fill their entire memory of about 33.5 MB
with random bytes.
This triggers a mouse button press without the up event, allowing us to
e.g. simulate a selection by moving the mouse while keeping the button
depressed.
The DOM spec defines what it means for an element to be an "editing
host", and the Editing spec does the same for the "editable" concept.
Replace our `Node::is_editable()` implementation with these
spec-compliant algorithms.
An editing host is an element that has the properties to make its
contents effectively editable. Editable elements are descendants of an
editing host. Concepts like the inheritable contenteditable attribute
are propagated through the editable algorithm.
Instead of recursively iterating all descendants of the common ancestor
of the new line range that are not contained by that range, skip the
entire node tree as soon as we determine they're not.
Before, on a mouse-move event, if the hovered html element did not have
a tooltip or it was not a link, `page_did_leave_tooltip_area()` and
`page_did_unhover_link()` virtual functions would get called.
Now, the page remembers if it is in a tooltip area or hovering a link
and only informs of leaving or unhovering only if it was.
Before, on *every* mouse-move event, `page_did_request_cursor_change()`
virtual function would get called, requesting to change cursor to the
event's mouse position's cursor.
Now, the page keeps track of the last cursor change that was requested
("page's current cursor") and only requests cursor change again if and
only if the current cursor is not already the one that is required.
Whenever we create a GC function, it should always be so that we can
pass it to a platform event loop spin, HTML event loop spin, or some
queued task on the HTML event loop. For every use case, any local
variables will be out of scope by the time the function executes.
Using a default reference capture for these kinds of tasks is dangerous
and prone to error. Some of the variables should for sure be captured
by value so that we can keep a GC object alive rather than trying to
refer to stack objects.
These variables are all captured in queued events or other event loop
tasks, but are all guarded by event loop spins later in the function.
The IGNORE_USE_IN_ESCAPING_LAMBDA will soon be required for all locals
that are captured by ref in GC::Function as well as AK::Function.
By actually using streams, they get marked as disturbed and the
`.bodyUsed` API starts to work. Fixes at least 94 subtests in the WPT
`fetch/api/request` test suite.
Co-authored-by: Timothy Flynn <trflynn89@pm.me>
The spec for filtered responses states:
Unless stated otherwise a filtered response’s associated concepts
(such as its body) refer to the associated concepts of its internal
response.
This includes setting its associated concepts. In particular, when the
filtered response's body is set upon fetching a request with integrity
metadata, we must set the internal response's body instead.
Further restrictions that apply to filtered response subclasses (such as
opaque filtered responses having a status code of 0) are already
implemented.
In particular, the processBody callback here *can't* move the
processBodyError callback. It is needed a few lines after. Passing by
value is safe and intended here.
Done by forward declaring the required functions and defining the needed
constants. The defines shouldn't collide as they are from memoryapi.h.
This is done to avoid including windows.h.
Add support for shared memory creation in WebAssembly memory API.
This API is needed for WPT tests that use shared array buffers.
Import related WPT tests.
The CenterRight and TopCenter alignment cases were
mistakenly identical due to a copy-paste error,
causing the function to behave unexpectedly.
Rather than attempting to fix it, remove aligned_within entirely.
The stale action has a weird interaction with the cache which requires
the `actions: write` permission. Without this, it is unable to overwrite
any existing cache key.
We want to remind both contributors and maintainers that PRs should be
up-to-date, reviewed and merged. Stale bot does a pretty good job of
notifying the right people :^)
Little refactoring to remove the last bits of ASN1 decoding/encoding
from within the `SECPxxxr1` class. It was a bit confusing for the
`SECPxxxr1` methods to handle ASN1 internally implicitly. Some explicit
methods are available to achieve the same functionality on the data
structures.
This allows to move ASN1 logic from inside the `SECPxxxr1` curve
itself to the data structures. It makes more sense to have dedicated and
explicit methods to handle transformation between formats.
With all the plumbing in place, we can handle this quirk at the
serialization layer.
This allows us to remove the pass where StyleComputer would loop over
all computed values and replace any color values with new values
stripped of their original name strings.
This change imports the https://wpt.live/accname/basic.html — which we
had overlooked earlier when importing all the other WPT accname tests.
(Note that this one of the six upstream WPT tests that directly call
test_driver.get_computed_label(element) — and so that we’d otherwise
need to patch our copy of, if we weren’t patching it in testdriver.js).
This change takes the change we made in 120bc52f23 to patch the imported
WPT aria-utils.js file to use our window.internals.getComputedLabel(el)
function, and moves that patching into the imported WPT testdriver.js
file — in the same way we did in c5966bbdcb.
That centralizes the patching, and avoids the need to patch multiple
other WPT tests we’re likely to import eventually. There are actually
six different WPT test files upstream which we haven’t imported yet that
call window.test_driver_internal.get_computed_label(el) directly — and
that, without this change, we’d otherwise end up needing to patch.
This change implements spec-conformant computation of default ARIA roles
for elements whose expected default role depends on the element’s
context — specifically, either on the element’s ancestry, or on whether
the element has an accessible name, or both. This affects the “aside”,
“footer”, “header”, and “section” elements.
Otherwise, without this change, “aside”, “footer”, “header”, and
“section” elements may unexpectedly end up with the wrong default roles.
This change separates the steps for checking the string value of the
ARIA “role” attribute out from the element.role_or_default() function
into a separate function — in order to expose a way to just check if the
ARIA “role” attribute actually has a value, without also then computing
a default role value if no “role” attribute value was found.
Otherwise, without this change, the only available function for
retrieving ARIA role values is the element.role_or_default() function —
which always does the additional step of computing (and returning) a
default role value if no “role” attribute is found.
This change adds a window.internals.getComputedLabel(element) function,
for use in testing ARIA-related behavior. It also patches the imported
WPT testdriver.js script’s test_driver.get_computed_role(element)
function to call our window.internals.getComputedRole(element) function.
Details' contents matches a new details-content pseudo element.
Further work is required to make this pseudo-element behave per spec.
This pseudo should be element-backed per
https://drafts.csswg.org/css-pseudo/#element-backed
Introduced in bd932858, the code would try to move `result.color_space`
into the ImmutableBitmap created for each single frame of an image. For
animated images, this would result in a use-after-move from the second
frame.
Previously, it ignored 'start', sorting from the array's
beginning. This caused unintended changes and slower
performance. Fix ensures sorting stays within 'start'
and 'end' indices only.
This commit adds an in-tree test for code added in a previous commit:
e89e084798543d567f00337e831f686aa3cb05a7
We want to make sure that the custom reason phrase is making it from
RequestServer to the "statusText" property in the XHR infrastructure.
The `str | None` syntax is only supported in Python 3.10+ and we can
support more Python versions without compromising readability by
importing the `Optional` type.
This commit adds a "echo_server_port" property to `WebContentOptions`.
Additionally, it makes `Application::web_content_options()` return a
mutable reference instead of a const reference so that we can set the
port value from the fixture.
This commit makes the following changes:
- Adds a model "Echo" for the request body
- Changes the main endpoint from /create to /echo (more REST-y)
- Sends "400: Bad Request" for invalid params or a reserved path
- Sends "409: Conflict" when trying to use an already registered path
- Prints the server port to stdout
- Removes unnecessary subcommands/options (start, stop, --background)
This corresponds to https://github.com/whatwg/html/pull/10753
WHile I was at it, I've also moved the checks inside the spin callback,
and reformatted the spec comments to match our style.
The assertions can be hit when Temporal.Duration.prototype.round / total
are provided a PlainDate at the very limits of valid date-times. Tests
were recently added to test262 which trip these assertions, thus we are
now crashing in those tests. Let's throw a RangeError instead, as this
is the behavior expected by the tests.
This is an editorial change in the Temporal proposal. See:
https://github.com/tc39/proposal-temporal/commit/630b043
This also changes the second argument of this AO to not use the alias of
TimeDuration, as that should not be used to refer to epoch nanoseconds.
This was missed in commit 2d9405e5d7.
This commit changes the exception flag to match the description and turn
off exceptions. This matches the behavior of -fno-exceptions. However,
on Windows SEH exceptions are always available, and there's no easy way
to turn them off. This flag should also suppress previous /EHsc flags,
if they get automatically set by cmake. Reference:
https://learn.microsoft.com/en-us/cpp/build/reference/eh-exception-handling-model?view=msvc-170
We only need a Realm to allocate CSSOM objects on the GC heap. Style
values are not such objects, and over time, we'll be changing the
parser to only produce non-CSSOM objects.
These changes make sure that we postpone calls to
ReadableByteStreamControllerCommitPullIntoDescriptor until after all
pull-into descriptors have been filled up by
ReadableByteStreamControllerProcessPullIntoDescriptorsUsingQueue.
Also pulls in the WPT test which was created in relation to this spec
change. The test verifies that a patched then() will see a null
byobRequest.
We are currently constructing the attribute names as FlyStrings every
time we invoke one of the ARIA attributes getters/setters. If there are
not any other instances of these strings in-memory, then we're thrashing
the FlyString cache.
Instead, let's follow suit of all other Web attributes - use an x-macro
to generate the attribute names.
This patch introduces the `Gfx::ColorSpace` class, this is basically a
serializable wrapper for skia's SkColorSpace. Creation of the instances
of this class (and thus ICC profiles parsing) is performed in the
ImageDecoder process. Then the object is serialized and sent through
IPC, to finally be handed to skia for rendering.
However, to make sure that we're not making all LibGfx's users dependent
on Skia as well, we need to ensure the `Gfx::ColorSpace` object has no
dependency on objects from Skia. To that end, the only member of the
`ColorSpace` class is the opaque `ColorSpaceImpl` struct. Though, there
is on issue with that design, the code in `DisplayListPlayer.cpp` needs
access to the underlying `sk_sp<SkColorSpace>`. It is provided by a
template function, that is only specialized for this type.
Doing this work allows us to pass the following WPT tests:
- https://wpt.live/css/css-color/tagged-images-001.html
- https://wpt.live/css/css-color/tagged-images-003.html
- https://wpt.live/css/css-color/tagged-images-004.html
- https://wpt.live/css/css-color/untagged-images-001.html
Other test cases can also be found here:
- https://github.com/svgeesus/PNG-ICC-tests
Note that SkColorSpace support quite a limited amount of color spaces,
so color profiles like the ones in [1] or the v4 profiles in [2] are not
supported yet. In fact, SkColorSpace only accepts skcms_ICCProfile with
a linear conversion to XYZ D50.
[1] https://www.color.org/browsertest.xalter
[2] https://www.color.org/version4html.xalter
If not set, when copying pixels into the SkImage, skia assumes that the
color space is the same as the input, so no transformation is done. We
are currently setting the color space to sRGB, this is fine for now as
it allows us to start making some transformations, but down the road we
will want to set that to the actual output's display color space.
It makes it a little bit easier to distinguish which one of
read_into_bitmap and write_from_bitmap actually modify the Bitmap that
was passed to the method. NFC.
This is a minor change to the Node::name_or_description code to switch
some instances of element.has_attribute("value"_string) over to instead
using element.has_attribute(HTML::AttributeNames::value).
Even in a successful LibWeb test run, on Linux, we currently always have
LSAN output. Causing this step to fail is currently making every run of
CI fail.
If LibWeb tests did fail, they should exit with a non-zero status code
on their own, causing the pipeline to fail.
The requirement to use // for comments wasn't described anywhere, so now
it is! Also updated the rule about TODO comments because we've allowed
those for a while now.
Lines like these were getting a warning to simplify the expanded
boolean expression from `!(a || b)` to `(a && b)`, but since the
`!(...)` is part of the macro, that is never going to happen.
Add a test that verifies selectors inside a shadow
root can only match their host element through :host pseudo-class.
Tests both simple selectors (#id, .class)
and complex selectors (:not, :where) to ensure they are blocked from
matching the host element directly.
Fixes issue #2319
Fixes an issue where selectors inside a shadow root could incorrectly
match their shadow host directly using selectors like #host,
instead of requiring :host pseudo-class selectors.
Fixes issue #2319
Fixes the crash in css/css-color/parsing/color-valid-hwb.html.
The crash was probably introduced in 248e4bb5, as it was the first
commit to VERIFY that the value given to `Color::with_opacity` were in
the correct range. As the values in color-valid-hwb.html were resolved
as NaN, the check never passed.
m_hover_label did not have checks if the mouse is in the same location.
This caused clickable URLs to be hidden.
Also shortened the label text to not be longer than half of the window.
Instead of writing ASAN and UBSAN output into the same stream we use for
test logging, direct them to log files, named asan.log.$PID and
ubsan.log.$PID, and then output them in a separate CI job that runs
afterwards. This should hopefully make it easier to see which tests are
failing.
The downside is that it's now harder to tell which tests the *SAN errors
are related to.
...in HTMLCanvasElement::to_data_url() and HTMLCanvasElement::to_blob().
This fixes the problem when surface is not allocated because context is
not initialized yet, even though canvas size is non-zero.
Fixes broken WebDriver screenshot endpoint.
The length resolution context might be needed even when resolving an
integer value, since it might contain a sign() function with length
values inside. This fixes a WPT subtest.
The WPT tests require the shortest possible serialization that support
an 8 bits roundtrip.
As an example, `128` is serialized to `0.5` while `127` needs more
precision and thus will be serialized to `0.498`.
This commit fixes 33 WPT subtests in css/css-color.
PaintableWithLines created from inline nodes need to examine their
children to fully compute their size and offset as nested children
might have textnodes that are placed before their parent.
Fixes#1286
In the move to a python version of this script, I didn't notice that
running the bootstrap script in shell mode precluded it from actually
accepting the -disableMetrics argument.
Existing vcpkg installs can be un-metrics'd by re-running the bootstrap
script with the disable argument, or by adding a vcpkg.disable-metrics
file in $VCPKG_ROOT
Most computed border-radii contain their initial values, and since the
normalized initial border radii are always zero, there is no need to do
expensive floating point math to normalize them.
In all these cases there should be an ancestor available, but it
definitely cannot hurt to be a bit more defensive about this and prevent
nullptr dereferences.
The refactor in the previous commit was storing a reference to a stack
allocated `Infrastructure::Request::BodyType` which was then immediately
freed. To fix this, we can store the `Infrastructure::Request::BodyType`
in a variable beforehand, so it becomes safe to reference.
C++ will jovially select the implicit conversion operator, even if it's
complete bogus, such as for unknown-size types or non-destructible
types. Therefore, all such conversions (which incur a copy) must
(unfortunately) be explicit so that non-copyable types continue to work.
NOTE: We make an exception for trivially copyable types, since they
are, well, trivially copyable.
Co-authored-by: kleines Filmröllchen <filmroellchen@serenityos.org>
This makes them trivially copyable/movable, silencing
> "parameter is copied for each invocation"
warnings on `Optional<T&>`, which are unnecessairy,
since `Optional<T&>` is just a trivially copyable pointer.
This creates a slight change in behaviour when moving out of an
`Optional<T&>`, since the moved-from optional no longer gets cleared.
Moved-from values should be considered to be in an undefined state,
and if clearing a moved-from `Optional<T&>` is desired, you should be
using `Optional<T&>::release_value()` instead.
OpenGL's origin is at the bottom-left corner, while Skia's origin is at
the top-left corner. This change adds a transformation to compensate for
this difference when rendering PaintingSurface attached to WebGL
context.
For now only macOS is supported.
IOSurface is used as a backing store because it will allow us to read
it from Skia and write to it from OpenGL without any extra copying:
- ANGLE_metal_texture_client_buffer extension is used to create
EGLSurface from IOSurface.
- Then the same IOSurface is wrapped into Metal texture and passed to
Skia allowing to share the same memory between Skia Metal backend and
ANGLE.
Previously, constructing a PaintingSurface from an IOSurface required
wrapping IOSurface into a Metal texture before passing it to the
PaintingSurface constructor. This process was cumbersome, as the caller
needed access to a MetalContext to perform the wrapping.
With this change SkiaBackendContext maintains a reference to the
MetalContext which makes it possible to do:
IOSurface -> MetalTexture -> SkSurface within a PaintingSurface
constructor.
Implement transfer logic for ArrayBuffer and ResizableArrayBuffer.
Change TransferDataHolder data type to Vector<u32> to reuse existing
serialization infrastructure.
Fix 5 WPT tests in `window-postmessage.window.html` that relates to
transport.
Fix `LibWeb/Text/input/Worker/Worker-postMessage-transfer.html`.
The latter is currently ignored due to flakiness, no rebaseline is
needed.
During serialization with transfer, initialize memory with known index
and initialize Serializer at position that dependent on the memory.
This is mandatory to make ArrayBuffer transport to work. It also happens
to fix 4 WPT tests, that are related to curcular references during
serialization.
Previously, it was assumed that nodes must share the same root, prior
to the calculation of their relative boundary point positions. This is
no longer the case, since `Selection.setBaseAndExtent()` now accepts
anchor and focus nodes that may be in different shadow trees.
We have to list the set of allowed values for the DOMTokenList to not
throw when asking if one is supported.
This fixes an issue where YouTube embeds would hang indefinitely trying
to report an endless series of exceptions, seen on https://null.com/
`AsyncBlockStart` was still doing a `DisposeResources` call as
specified in older drafts of the `explicit-resource-management`
proposal, but the latest draft no longer does this, and it is
causing crashes when combined with the `array-from-async` proposal.
The 'reason' was getting initialized to 'empty' state when not
provided through the constructor, which results in a crash when
accessed through throw_dom_exception_if_needed in the generated
IDL getter.
All global objects current need to be event targets so that they
can have events dispatched to them. This allows for removing of
verify_cast for these global objects.
The "copy" command is not in the Miscellaneous commands section. The
"defaultParagraphSeparator" command is, however. Let the accompanying
comment reflect that.
This makes these values the same as `start` and `end`. While this is not
entirely correct, it is better than centering which is what we did
previously.
This fixes misaligned images on https://nos.nl
From this action's documentation, "0 means using default retention."
There's no reason for this to have a different retention duration than
the CI results themselves.
Without this, if two CI runs on Linux both fail and want to upload
screenshots, we get an error like this on the second:
> Error: Failed to CreateArtifact: Received non-retryable error: Failed
> request: (409) Conflict: an artifact with this name already exists on
> the workflow run
Using all the inputs as part of the name should make this kind of
conflict impossible.
This allows us to disable test output, which performs expensive assert
tracking. This was making our imported tests run significantly slower
than tests run via `WPT.sh`.
Formatting the output ourselves also allows us to remove unnecessary
information from the test output.
This commit also rebaselines all existing imported WPT tests to follow
the new format.
```
VERIFICATION FAILED: !_temporary_result.is_error()
```
is not really a helpful error message.
When we are including `AK/Format.h`, which is most of the time,
we can easily print a much more useful error message:
```
UNEXPECTED ERROR: Cannot allocate memory (errno=12)
```
The source code position cache was moved from a line based approach
to a "chunk"-based approach to improve performance on large, minified
JavaScript files with few lines, but this has had an adverse effect
on _multi-line_ source files.
Reintroduce some of the old behaviour by caching lines again, with
some added sanity limits to avoid caching empty/overly small lines.
Source code positions in files with few lines will still be cached
less often, since minified JavaScript files can be assumed to be
unusually large, and since stack traces for minified JavaScript
are less useful as well.
On WPT tests with large JavaScript dependencies like
`css/css-masking/animations/clip-interpolation.html` this reduces the
amount of time spent in `SourceCode::range_from_offsets` by as much as
99.98%, for the small small price of 80KB extra memory usage.
The rules for parsing integers don't specify an upper bound on the
value that can be returned, so the `parse_integer_digits` method can be
used to check whether the given arbitrarily-large StringView is valid
according to these rules. The `parse_integer` and
`parse_non_negative_integer` methods would fail for values larger than
2147483647 when they shouldn't have.
restore() corresponding to ApplyFilters should be called after stacking
context content is painted, not before.
Fixes regression introduced in c94b4316e7
Many dependencies aren't currently included in the devShell. As ladybird
is already packaged downstream, we can pull in those buildInputs along
with the extra dev dependencies already defined.
ApplyOpacity internally calls canvas.saveLayer() which requires a
matching canvas.restore() to be called.
Fixes missing header on https://supabase.com/
Setting the `width` or `height` properties of `HTMLCanvasElement` to a
value greater than 2147483647 will now cause the property to be set to
its default value.
ApplyFilter internally calls canvas.saveLayer() which requires a
matching canvas.restore() to be called.
Fixes painting on https://supabase.com/ regressed by
8562b0e33b
This is part of a normative change to the HTML space for WebAssembly JS
module integration and the source phase import proposal, see:
https://github.com/whatwg/html/commit/10ed38ee7
Further changes are required, but this is a start :^)
Previously, we leaked the `curl_slist`s on every request. This also
validates the pointer we get from `curl_slist_append` before setting the
option.
Also, use the `set_option` helper for CURLOPT_RESOLVE as it will print
when there is an error.
Previously, we would stop the repeat timer even if we got a null result.
This caused the pending lookup to:
- Never resolve, and
- Never get purged for too many retries
I believe the underlying issue is something on the socket level, but we
should handle this case regardless.
When the "Consume a component value from input, and do nothing."
step in `Parser::consume_the_remnants_of_a_bad_declaration` was
executed, it would allocate a `ComponentValue` that was then
immediately discarded.
Add explicitly `{}_and_do_nothing` functions for this case that never
allocate a `ComponentValue` in the first place.
Also remove a `(Token)` cast, which was unnecessarily copying a `Token`
as well.
Lazily coercing might have made sense in the past, but since hashing
and comparing requires the `PropertyKey` to be coerced, and since a
`PropertyKey` will be used to index into a hashmap 99% of the time,
which will hash the `PropertyKey` and use it in comparisons, the
extra complexity and branching produced by lazily coercing has
become more trouble than it is worth.
Remove the lazy coercions, which then also neatly allows us to
switch to a `Variant`-based implementation.
Our GCC pipeline is regularly timing out. When it doesn't time out, it
finishes just under the default 25 minute limit. Let's bump the timeout
to 30 minutes to give it a bit more wiggle room.
We currently have some tests that hang. In order to find which tests
these are, let's enable verbose logging to get a log of each running
test and its individual duration.
This adds a verbosity option to log the start and end of each test, with
the duration taken for the test. To be able to use this option with our
exisiting verbosity flag, without cluttering stdout with other data, we
add verbosity levels to headless-browser. The level is increased by
providing the -v flag on the command line multiple times.
Anchor the minimum functionality for this. WPT has an extensive suite
to test editing functionalities, but they all take a long time to
execute - so let's have a simple regression test in-tree for now.
To facilitate the implementation of "delete" and all associated
algorithms, split off this piece of `Document` into a separate
directory.
This sets up the infrastructure for arbitrary commands to be supported.
Corresponds to https://github.com/whatwg/html/pull/10683
As part of this, I noticed we incorrectly were setting the "is popup"
flag on the Navigable instead of the BrowsingContext. I've fixed that
and removed the erroneous flag from Navigable.
This lets us move a few Host-related functions (like serialization and
checks for what the Host is) into Host instead of having them dotted
around the codebase.
For now, the interface is still very Variant-like, to avoid having to
change quite so much in one go.
A couple of reasons:
- Origin's Host (when in the tuple state) can't be null
- There's an "empty host" concept in the spec which is NOT the same as a
null Host, and that was confusing me.
Recently reported against the shadow realm proposal after running into
issues with WPT tests.
In a nested shadow realm, the associated realm is a shadow realm, not
the principal realm. One such issue this fixes is a crash when a nested
shadow realm performs an operation which requires the principal settings
object.
There was a bug in the HTML proposal where a synthetic realm settings
object's principal realm was a shadow realm if there were nested shadow
realms, which this assertion catches more directly (rather than later
down the track, where it is used).
We were meant to also assert for this case, but we were previously
returning early.
This is to resolve naming conflicts between the ServiceWorker JS exposed
object and the internal representation of a ServiceWorker which is going
to be stored cross process.
Replicate what we are doing with RSA and parse both the private and
public key when parsing the ASN1.
The only thing that changed in the tests is the error message.
The decoding inside `RSA::parse_rsa_key` is quite complex because it
tries to understand if it's decoding PKCS#8 or PKCS#1. Simplify the code
by moving the burden to the PEM decoder.
I have divided ANS1 constants by length so that they don't have
trailing zeros that need to be removed.
Also moved OIDs lists to the only place they are used for clarity.
Fixed a couple of WPT tests by adding SECP521r1 to the list of known
curves.
Regardless of what the shorthand property is, if all its longhands are
the same CSS-wide keyword such as "initial" or "inherit", then it's the
same as the shorthand being that value.
This gets us 2 WPT subtest passes.
This wins us 65 new WPT subtest passes! It also shows up that we're
doing the wrong thing in ShorthandStyleValue in places, notably with
the grid properties. However, having one place to fix instead of two
will make it easier to correct them. :^)
In order to be fully correct, we should use the algorithm here:
https://drafts.csswg.org/cssom/#serialize-a-css-value
However, it's quite hand-wavy. What we do have in the meantime is
`ShorthandStyleValue::to_string()`, where we special-case the
serialization rules for shorthands with a generic fallback that's
equivalent to what the previous `get_property_value()` code was doing.
libpxbackend was not being installed from vcpkg which lead to use of
libpxbackend on the system causing library mismatch in rolling
distros(ie arch linux)
a
When attempting to set `HTMLProgressElement.max` to a value not greater
than 0, we were previously setting the value to 1. We now retain the
previous value.
This change ensures that the correct default value of 0 is used and
that values greater than 2147483647 will fall back to the default value.
It also splits the display size concept into a separate method, as
this isn't supposed to be used when getting the IDL property.
This change makes Ladybird give the value of the aria-label attribute
the correct precedence for accessible-name computation required by the
“Accessible Name and Description Computation” and HTML-AAM specs and by
the corresponding WPT tests.
Otherwise, without this change, Ladybird fails some of the WPT subtests
of the test at https://wpt.fyi/results/accname/name/comp_label.html.
This change implements full support for the “A. Hidden Not Referenced”
step at https://w3c.github.io/accname/#step2A in the “Accessible Name
and Description Computation” spec — including handling all hidden nodes
that must be ignored, as well as handling hidden nodes that, for the
purposes of accessible-name computation, must not be ignored (due to
having aria-labelledby/aria-describedby references from other nodes).
Otherwise, without this change, not all cases of hidden nodes get
ignored as expected, while cases of nodes that are hidden but that have
aria-labelledby/aria-describedby references from other nodes get
unexpectedly ignored.
The gist is that we need to construct an ICU date-time formatter for
each possible Temporal type. This is of course going to be expensive.
So instead, we construct the configurations needed for the ICU objects
in the Intl.DateTimeFormat constructor, and defer creating the actual
ICU objects until they are needed.
Each formatting prototype can also now accept either a number (as they
already did), or any of the supported Temporal objects. These types may
not be mixed, and their properties (namely, their calendar) must align
with the Intl.DateTimeFormat object.
Attempting to set `HTMLInputElement.size` to 0 via IDL now throws an
IndexSizeError DOMException. Attempting to set it to a value larger
than 2147483647 results in it being set to the default value.
This reverts commit aeaa284be3.
This didn't actually help. It appears we actually just get stuck, so
increasing the timeout to an hour just makes CI take longer.
We seem to be brushing up against the 1500s limit (25 minutes) on both
macOS and Linux+GCC workflows. As we keep importing more WPT tests, we
will exasperate the issue.
I dug through the code and the WebCryptoAPI spec to figure out the
reason for `... mixed case parameters` WPT tests and figured out that
our implementation was slightly wrong.
By being closer to the spec we can now pass those tests and also remove
a bunch of duplicated code.
Context: https://github.com/LadybirdBrowser/ladybird/pull/2598#discussion_r1859263798
To check whether a NavigationParams is null, we have to check whether
it's `Empty` or `NullWithError`. Instead, we can merge both of these
possible variants into an optional error. If `NullOrError` has no
value it's null, otherwise it contains an error message.
While the code that did this referred to the HTML spec, other browsers
appear not to have this behavior when parsing XML, and it breaks a WPT
subtest.
This change does not appear to break any tests, and fixes 1 WPT subtest.
Size negotiation should not occur for other viewports, such as iframe,
since that would allow content inside the iframe to affect the size of
the iframe in the containing document.
300 new subtest passes on WPT. :^)
If `HTMLMarqueeElemnt.scrollAmount` or `HTMLMarqueeElemnt.scrollDelay`
is set to a value larger than 2147483647, then it should be set to its
default value.
Previously, `ECDH::generate_key` was implemented by storing a
`ByteBuffer` in the `InternalKeyData`. This improves the implementation
by using internal structures of already-parsed data.
Parse and store the `ECPrivateKey` extracted from the
`privateKeyAlgorithm` field of the ASN.1 `PrivateKeyInfo` sequence when
the algorithm identifier is `ec_public_key_encryption`.
The parsing function returns `ErrorOr` instead of an "empty" key, like
`parse_rsa_key` does. To me, this seemed better in terms of reliability.
As mentioned in the previous commit, there is room for improvement.
Added basic EC private and public key definitions as well as ASN.1
encoding and decoding.
A lot of refactoring can be made around the ASN.1 processing (here and
in other parts of the codebase) by utilizing what is available
in `LibCrypto::Certificate` as macros, but I think it's outside the
scope of implementing ECDH support for WebCryptoAPI.
It looks like the `SECPxxxr1` was made mainly to work with the TLS
implementation which requires everything to be bytes. This is not always
the case and a loss of generality.
I have added some methods that take and return `UnsignedBigInteger`s
for better interoperability with ASN.1 stuff. I would like to remove
the old methods relying on bytes, but I haven't made my mind around how
to generalize it for all curves.
Add support for encoding parameters in `wrap_in_private_key_info` and
`wrap_in_subject_public_key_info` as well as turn `Span<int>` into
`Span<int const>`.
If we were able to parse an ISO8601 Date string, but the parse results
in an invalid date (e.g. out of the min/max range), we should abort
parsing immediately.
And make ErrorType definitions use a better hanging-indent style, to
make it easier to maintain going forward.
i.e. instead of:
M(VeryLongErrorNameHere, "very long error "
"message across multiple "
"lines"
We now have:
M(VeryLongErrorNameHere,
"very long error message across multiple "
"lines")
We were passing types like ISODate by reference so that they could be
used as forward-declarations. But after commit 021a5f4ded, we now have
their full definitions anywhere they're needed. So let's pass ISODate by
value everywhere consistently - it is only 8 bytes.
It was a bit of a semantic mistake too use this alias too eagerly.
Namely, it should not be used to refer to epoch nanoseconds. We now only
use the TimeDuration alias where the spec refers to a value as a time
duration.
When experimenting with different inheritance structures, I ended up
seeing compilation failures whenever a parent class defined a `call`
method. This seems more in line with the rest of the code.
According to the HTML specification, the `size` attribute of an input
element must be a valid non-negative integer greater than zero. If the
value is invalid or set to `0`, the default size of `20` should be used.
This small change fixes one issue identified in
https://wpt.live/html/rendering/widgets/input-text-size.html
The WPT test suite was also automatically imported.
Use discrete animation when the number of components or the types
of corresponding components do not match. This commit does not cover
all cases, but adds FIXME comments in the appropriate places.
This fixes a bug in `pause()` that canceled the pause task instead
of the play task. This issue prevented the animation from being paused
while a play task is scheduled.
Previously any existing ElementInlineCSSStyleDeclaration would get
overwritten by e.setAttribute("style", ...), while it should be updated
instead.
This fixes 2 WPT subtests.
This test could fail when `/tmp/testfile` already exists.
There also shouldn't be a need to wait so long for this test,
except for on MacOS, where the `FileWatcher` implementation is
apparently less reliable.
By moving `Certificate` to `LibCrypto` it is possible to reuse a bunch
of code from in `LibCrypto` itself. It also moves some constants
and pieces of code to a more appropriate place than `LibTLS`.
This also makes future work on WebCryptoAPI easier.
The implementation of `Certificate::is_valid` and
`Certificate::is_self_signed` were in `TLSv12.cpp` and they have been
moved to `Certificate.cpp`.
This is in preparation of the next commits to split the changes.
The declaration of `DefaultRootCACertificates` was in `Certificate.h`
and its implementation in `TLSv12.cpp`. It has been moved over
to `TLSv12.h` for consistency.
This is in preparation of the next commits to split the changes.
This library isn't used by anything but the Android build which
currently doesn't work. We most likely won't be using a homegrown
implementation for archive formats in the future, regardless.
Two differences:
1. An extra step inserted to record timing info, which we don't yet
implement.
2. The last step in the loop breaks instead of returning the value
directly. (But this is functionally the same, as the following step
does return that value.)
(Also removed the duplicated part of the comment in step 11 née 10.)
So, there's no actual change in behavior.
This change adds handling for the “Determine Child Nodes” substep at
https://w3c.github.io/accname/#comp_name_from_content_find_child in the
“Accessible Name and Description Computation” spec. Specifically, it
adds handling for the “If the current node has an attached shadow root”
and “if the current node is a slot with assigned nodes” conditions.
Otherwise, without this change, AT users don’t hear the expected
accessible names in cases where the content for which an accessible name
being computed is in a shadow root or slot element.
When serializing an sRGB color value that originated from a named color,
it should return the color name converted to ASCII lowercase. This
requires storing the color name (if it has one).
This change also requires explicitly removing the color names when
computing style, because computed color values do not retain their name.
It also requires removing a caching optimization in create_from_color(),
because adding the name means that the cached value might be wrong.
This fixes some WPT subtests, and also required updating some of our own
tests.
...and make sure it will eventually complete (or fail) by adding a
timeout retry sequence.
Fixes an issue where RequestServer would stick around after exit,
waiting for piled up DNS requests for a long time.
This change makes Ladybird correctly handle all “encapsulation” tests in
the https://wpt.fyi/results/accname/name/comp_host_language_label.html
set of tests in WPT.
Those all test the requirement that when computing the accessible name
for a <label>-ed form control, then any content (text content or
attribute values) from the control itself that would otherwise be
included in the accessible-name computation for it ancestor <label> must
instead be skipped and not included.
The HTML-AAM spec seems to try to achieve that result by expressing
specific steps for each particular type of form control. But what all
that reduces/optimizes/simplifies down to is just, “skip over self”.
Otherwise, without this change, Ladybird includes that “self” content
from those “encapsulated” elements when doing accessible-name
computation for the elements — which results in AT users hearing
unexpected extra content in the accessible names for those elements.
This change makes Ladybird conform to the requirements in the HTML-AAM
spec at https://w3c.github.io/html-aam/#accname-computation for the
cases of HTML input@type=button, input@type=submit, and input@type=reset
elements. Otherwise, without this change, Ladybird fails to expose the
expected accessible names for those cases.
This change makes Ladybird conform to the requirements in the HTML-AAM
spec at https://w3c.github.io/html-aam/#accname-computation for the
cases of HTML table, fieldset, and input@type=image elements. Otherwise,
without this change, Ladybird fails to expose the expected accessible
names for those cases.
There was an existing check to ensure that `U` was an lvalue reference,
but when this check fails, overload resolution will just move right on
to the copy asignment operator, which will cause the temporary to be
assigned anyway.
Disallowing `Optional<T&>`s to be created from temporaries entirely
would be undesired, since existing code has valid reasons for creating
`Optional<T&>`s from temporaries, such as for function call arguments.
This fix explicitly deletes the `Optional::operator=(U&&)` operator,
so overload resolution stops.
The Qt style hints' color scheme can return ::Unknown, in which case we
still need to fall back to our 'old' method of using the luma component
of the background role color.
Before this change, skipping a word to the left in a non empty text
input element would crash when near the end of the text as the
offset + length of the substring would exceed the length of the string.
This was removed from the ShadowRealm HTML spec integration PR after my
suggestion as it is not used anywhere, and I don't believe it would ever
need to be used in the future or by other specs.
An environment settings object will return a copy to the URL. From a
quick glance, we could probably make an environment settings object
return a reference to one, but let's just change this code to make
a copy since its not safe to rely on that.
Prior to this commit LibTLS closed the connection but did not consider
it terminated after receiving and acknowledging a CloseNotify from the
server, which led to hangs in DoT (and possibly other users).
Instead of collecting all documents in a big vector and then filtering
the vector (twice!) with remove_all_matching(), we now pass a filter
callback to documents_in_this_event_loop_matching() and avoid all the
extra shuffling work.
Saw this stuff hogging ~20% of CPU time when profiling a WebContent
process in the middle of a WPT run.
Rather than accumulating margins into a vector, and then looping through
them when resolving the margin, it's much simpler to just update two
fields, and sum them when resolving.
When inserting a new utf-16 surrogate next to an existing surrogate
with replaceData, the surrogates would not get merged correctly into a
single code point. This is because internally the text data is stored
as utf-8, and the two surrogates would be converted seperately. This
has now been fixed by first recreating the whole string in utf-16 and
then converting it back to utf-8.
It's not the most efficient solution, but this fixes at least 6 WPT
subtests.
This isn't directly in the spec, but since replaceChild is implemented
in terms of remove + insert, the removal step may cause arbitrary code
to execute, and so we have to verify that the replaceChild inputs still
make sense afterwards, before doing the insertion.
This roughly matches what WebKit does, and makes a bunch of HTML parsing
tests in WPT stop asserting.
In particular, input character lookahead now knows how to stop at the
insertion point marker if needed.
This makes it possible to do amazing things like having document.write()
insert doctypes one character at a time.
This is necessary to avoid a circular reference when including
Serializable.h in DOMException.h.
This moves the definition of SerializationRecord, SerializationMemory,
and DeserializationMemory into LibWeb/Forward.h so that Serializable.h
only needs to include LibWeb/Forward.h.
Take record of the named capture group prior to parsing the group's
body. This requires removal of the recorded minimum length of the named
capture group directly, and now needs to be looked up via the group
minimu lengths table.
Elements with transforms were tested on their pre-transformed
positions, causing incorrect hits.
Copy the position transformation done in `StackingContext::hit_test`
to ensure that hit tests are done on the _actual_ position.
For Qt >= 6.5, the system theme can be determined reliably, so no
guesswork is needed. A fallback remains for Qt < 6.5, but it is
hacky and less reliable.
If we reach the insertion point at the same time as we switch to another
tokenizer state, we have to bail immediately without proceeding with the
next code point. Otherwise we'd fetch the next token, get an EOF marker,
and then proceed as if we're at the end of the input stream, even though
more data may be coming (with more calls to document.write()..)
The repository being in static storage is a bit of a hodgepodge, but in
line with how our current storage partitioning is done. We should
eventually move this, along with other across browsing context APIs to a
proper location at a later stage. But for now, this makes progress on
the meat of the BroadcastChannel API.
In the case where we had a preferred aspect ratio and a natural height
but no natural width, we'd get into ping-ponging infinite recursion by
trying to find the width to resolve the height to resolve the width to
resolve the height...
These fields are modified by an invocation to CalendarResolveFields. All
callers currently don't care about these modifications, so they move the
CalendarFields struct into the AO. But it turns out that at least one AO
will care (InterpretTemporalDateTimeFields), thus we will need to take
by reference here.
Currently, AbstractOperations.h needs to include Time.h for the Time
structure. Soon, Time.h will need to include AbstractOperations.h for
structures defined there. To avoid a circular include, let's put the
ISO types into their own file, so that AO.h does not need to include
any JS type header.
Our BigInt implementation can store the negativity of the division
result in either the quotient or the remainder. This is exposed by an
upcoming prototype, which will reach this division with:
-432000000000000 / 86400000000000000000000
Which has the BigInt division result:
quotient = 0
remainder = -432000000000000
(Our old Temporal implementation also had this bug).
For example, consider the following operation:
const instance = new Temporal.PlainDate(2000, 5, 2);
instance.until("-271821-04-19");
The spec would have us enter a loop to compute the difference between
these dates by incrementing an intermediate date one day at a time.
Instead, we can do some math to skip ahead much closer to the desired
date.
This one is particularly weird as there's a priority order, and we even
have to look at attributes from the container element if we're inside a
subframe.
This is an ad-hoc hack papering over the fact that we can apparently
end up in these places without an active window, and proceeding without
one leads to assertions on WPT.
This change adds support for computing accessible names for SVG
elements, per the https://w3c.github.io/svg-aam/#mapping_additional_nd
spec requirements. Otherwise, without this change, accessible names for
SVG elements don’t get exposed as expected.
Also removing a FIXME about not covering all of the event names as it is
not exactly clear when such a FIXME would be addressed, especially as
these come from multiple specifications.
- Hue now wraps properly when negative or larger than 360
- The hsl to rgb conversion now closely mirrors the code example from
the spec.
This fixes a number of WPT tests in
/css/css-color/parsing/color-computed-hsl.html
We now have the Temporal facilities to implement the Date AOs which
parse UTC offset strings using the ISO8601 parser. This patch updates
those AOs and their callers in accordance with the Temporal spec.
This started with implementing TemporalMonthDayString. It turns out that
the facilities needed to parse that production includes nearly all the
helpers to parse each of:
TemporalDateTimeString
TemporalInstantString
TemporalMonthDayString
TemporalTimeString
TemporalYearMonthString
TemporalZonedDateTimeString
As most of these invoke the same helpers with different options. So,
all 6 of those productions are implemented here.
Now, along with the mouse events we also dispatch pointerup, pointerdown
and pointermove.
With this change shape painting works on https://excalidraw.com/
Previously, the list was copied when constructing the FormData object,
then the original list was passed to the event, meaning any changes to
the list that happened within the event would not be reflected outside
of it.
This commit introduces proper handling of three intrinsic size keywords
when used for CSS heights:
- min-content
- max-content
- fit-content
This necessitated a few plumbing changes, since we can't resolve these
values without having access to containing block widths.
This fixes some visual glitches on https://www.supabase.com/ as well
as a number of WPT tests. It also improves the appearance of dialogs.
An inopportune garbage collection may cause collected `ResizeObserver`s
to unregister themselves from `m_resize_observers` while we are
iterating over it, resulting in a use-after-free.
For example, in the following HTML:
```html
<label>
<input type="radio" name="fruit" value="apple" id="radio1">
<span class="box"></span>
</label>
```
When any descendant of a <label> element is clicked, a "click" event
must be dispatched on the <input> element nested within the <label>, in
addition to the "click" event dispatched on the clicked descendant.
Previously, this behavior was implemented only for text node descendants
by "overriding" the mouse event target using `mouse_event_target()` in
the TextPaintable. This approach was incorrect because it was limited to
text nodes, whereas the behavior should apply to any box. Moreover, the
"click" event for the input control must be dispatched *in addition* to
the event on the clicked element, rather than redirecting it.
The clientX and clientY values are, as per the spec, the offset from
the viewport.
This makes them actually be that and also fixes up the calculations
for offsetX, offsetY, pageX and pageY.
I assume all of these got messed up in some sort of refactor in the
past.
The spec comment from the now-removed
compute_mouse_event_client_offset() function sadly has no convenient
place to be anymore so, for now, it is just gone as well.
Personally, I think it'd make sense to refactor a lot of this file so
that not every mouse event repeats a large chunk of (almost) identical
code. That way there'd be a nice place to put the comment without
repeating it all over the file.
But that is out of the scope of this PR.
Also: I know, offsetX and Y are not fully fixed yet, they still
don't ignore the element's CSS transforms but I am working on that
in a new PR.
This change fixes unhoverable toolbar on https://excalidraw.com/
The problem was that React.js uses setProperty() to add style properties
specified in the "style" attribute in the virtual DOM, and we were
failing to add the CSS variable used to set the "pointer-events" value
to "all".
See previous the commit description for more details about the floating
points operations.
The hwb test cases in `css-color-functions` are now rendered identically
to what firefox does (I haven't checked the others tests, but they
aren't affected by this commit).
Without this change the math in `CSSHWB::to_color()` is lacking some
precision to generate the correct value to hand to `Color::from_hsv()`.
More precisely, when converting `hwb(120 20 30)`, the HSV's value would
be calculated as `1 - .3`. However, it turns out that `1 - .3f != .7f`
and `1 - .3f` gives bad results down the road in `Color::from_hsv()`.
This example actually only requires `resolve_with_reference_value()` to
return a double. I changed the two others for symmetry.
In line with the ShadowRealm proposal changes in the WebIDL spec:
webidl#1437 and supporting changes in HTML spec.
This is required for ShadowRealms as they have no relevant settings
object on the shadow realm, so fixes a crash in the QueueingStrategy
test in this commit.
These interfaces are exposed on *, meaning it should work for workers
and our newly added shadow realm global object by being stored on the
universal global scope mixin.
This also includes a stubbed Temporal.Duration.prototype.
Until we have re-implemented Temporal.PlainDate/ZonedDateTime, some of
Temporal.Duration.compare (and its invoked AOs) are left unimplemented.
Our Temporal implementation is woefully out of date. The spec has been
so vastly rewritten that it is unfortunately not practical to update our
implementation in-place. Even just removing Temporal objects that were
removed from the spec, or updating any of the simpler remaining objects,
has proven to be a mess in previous attempts.
So, this removes our Temporal implementation. AOs used by other specs
are left intact.
This is required by mini Cloudflare invisible challenges, as it will
only run if the readyState is not "loading". If it is "loading", then
it waits for readystatechange to check that it's not "loading" anymore.
Initial about:blank iframes do not go through the full navigation and
thus don't go through HTMLParser::the_end, which sets the ready state
to something other than "loading". Therefore, the challenge would never
run, as readyState would never change.
Seen on https://discord.com/login
`StyleComputer::font_matching_algorithm` was creating a copy of a
`FlyString` every time a `MatchingFontCandidate` was constructed or
copied, causing millions of unnecessairy reference updates when a
lot of fonts are loaded.
While a more permanent solution would be to not load so many unused
fonts, let's do the right thing and remove the unnecessairy copies of
`FlyString`.
size >> 31 >> 1 is used instead of size >> 32 to support 32-bit Windows
(size_t is 32 bit there, and you cannot shift 32-bit value by 32 bits
on x86)
This is equivalent to sizeof(size) == 4 ? 0 : size >> 32
Windows doesn't have a concept of zombie children, hence:
* `disown` not needed
* we need a process handle because otherwise if the process have ended
by the time `wait_for_termination` is called
its pid may be reassigned to other process
This includes a protocol for creating LibGC Heap allocated Swift
objects. Pay no attention to the Unmanaged shenanigans, they are
all behind the curtain.
This will allow us to use the GC to manage the lifetime of objects
that are not C++ objects, such as Swift objects. In the future we
could expand this cursed FFI to other languages as well.
While we don't want arbitrary callers deferring GC, we do want
deferral to be available to the Swift. In order for Swift to
understand the RAII nature of DeferGC, we need to mark it as
non-copyable.
CDATASection inherits from Text, and so it was incorrect for them to
claim not to be Text nodes.
This fixes at least two WPT subtests. :^)
It also exposed a bug in the DOM Parsing and Serialization spec,
where we're not told how to serialize CDATASection nodes.
Spec bug: https://github.com/w3c/DOM-Parsing/issues/38
Instead of checking the header in ZlibDecompressor::create(), we now
check it in read_some() when it is called for the first time. This
resolves a FIXME in the new DecompressionStream implementation.
We don't need nanosecond precision here anyways, as we only display
millisecond resolution.
This uses our simple duration formatter from AK, which is updated to
accept a Duration here. This method did not have any users after the
move from Serenity.
The Duration record no longer exists in Temporal. Implement it according
to the DurationFormat spec to prepare for its removal from our Temporal
implementation.
We also implement the DurationSign AO here as well, as the Temporal
implementation will now require a Temporal.Duration JS object.
Implemented by reusing AddMask display list item that was initially
added for `background-clip` property.
Progress on flashlight effect on https://null.com/games/athena-crisis
This was causing issues for my Ubuntu 24.04 build when building
the Distribution preset, so just stash this constant config in
the CMake cache to not worry about it anymore.
For a while we used the wider Paintable type for stacking context,
because it was allowed to be created by InlinePaintable and
PaintableBox. Now, when InlinePaintable type is gone, we can use more
specific PaintableBox type for a stacking context.
I believe this is an error in the UI Events spec, and it should be
updated to match the HTML spec (which uses WindowProxy everywhere).
This fixes a bunch of issues already covered by existing WPT tests.
Spec bug: https://github.com/w3c/uievents/issues/388
Note that WebKit has been using WindowProxy instead of Window in
UI Events IDL since 2018:
816158b4aa
These compressors will be used by w3c's CompressionStream, which can run
arbitrary JS, and thus never reach their "finish" steps. Let's not crash
the WebContent process if that happens.
GzipCompressor is currently written assuming that it's write_some method
is only called once. When we use this class for LibWeb, we may very well
receive data to compress in small chunks. So this patch makes us write
the gzip header and footer only once, which now resembles the zlib and
deflate compressors.
Some LibCompress API changes for LibWeb will make these utilities a bit
difficult to keep up to date. Given that these are unused anways, let's
just not bother.
TL;DR: There are two available sets of coefficients for the conversion
matrices from XYZ to sRGB. We switched from one set to the other, which
is what the WPT tests are expecting.
All RGB color spaces, like display-p3 or rec2020, are defined by their
three color chromacities and a white point. This is also the case for
the video color space Rec. 709, from which the sRGB color space is
derived. The sRGB specification is however a bit different.
In 1996, when formalizing the sRGB spec the authors published a draft
that is still available here [1]. In this document, they also provide
the matrix to convert from the XYZ color space to sRGB. This matrix can
be verified quite easily by using the usual math equations. But hold on,
here come the plot twist: at the time of publication, the spec contained
a different matrix than the one in the draft (the spec is obviously
behind a pay wall, but the numbers are also reported in this official
document [2]). This official matrix, is at a first glance simply a
wrongly rounded version of the one in the draft publication. It however
has some interesting properties: it can be inverted twice (so a
roundtrip) in 8 bits and not suffer from any errors from the
calculations.
So, we are here with two versions of the XYZ -> sRGB matrix, the one
from the spec, which is:
- better for computations in 8 bits,
- and official. This is the one that, by authority, we should use.
And a second version, that can be found in the draft, which:
- makes sense, as directly derived from the chromacities,
- is publicly available,
- and (thus?) used in most places.
The old coefficients were the one from the spec, this commit change them
for the one derived from the mathematical formulae. The Python script to
compute these values is available at the end of the commit description.
More details about this subject can be found here [3].
[1] https://www.w3.org/Graphics/Color/sRGB.html
[2] https://color.org/chardata/rgb/sRGB.pdf
[3] https://photosauce.net/blog/post/making-a-minimal-srgb-icc-profile-part-3-choose-your-colors-carefully
The Python script:
```python
# http://www.brucelindbloom.com/index.html?Eqn_RGB_XYZ_Matrix.html
from numpy.typing import NDArray
import numpy as np
### sRGB
# https://www.w3.org/TR/css-color-4/#predefined-sRGB
srgb_r_chromacity = np.array([0.640, 0.330])
srgb_g_chromacity = np.array([0.300, 0.600])
srgb_b_chromacity = np.array([0.150, 0.060])
##
## White points
white_point_d50 = np.array([0.345700, 0.358500])
white_point_d65 = np.array([0.312700, 0.329000])
#
r_chromacity = srgb_r_chromacity
g_chromacity = srgb_g_chromacity
b_chromacity = srgb_b_chromacity
white_point = white_point_d65
def tristmimulus_vector(chromacity: NDArray) -> NDArray:
return np.array([
chromacity[0] /chromacity[1],
1,
(1 - chromacity[0] - chromacity[1]) / chromacity[1]
])
tristmimulus_matrix = np.hstack((
tristmimulus_vector(r_chromacity).reshape(3, 1),
tristmimulus_vector(g_chromacity).reshape(3, 1),
tristmimulus_vector(b_chromacity).reshape(3, 1),
))
scaling_factors = (np.linalg.inv(tristmimulus_matrix) @
tristmimulus_vector(white_point))
M = tristmimulus_matrix * scaling_factors
np.set_printoptions(formatter={'float_kind':'{:.6f}'.format})
xyz_65_to_srgb = np.linalg.inv(M)
# http://www.brucelindbloom.com/index.html?Eqn_ChromAdapt.html
# Let's convert from D50 to D65 using the Bradford method.
m_a = np.array([
[0.8951000, 0.2664000, -0.1614000],
[-0.7502000, 1.7135000, 0.0367000],
[0.0389000, -0.0685000, 1.0296000]
])
cone_response_source = m_a @ tristmimulus_vector(white_point_d50)
cone_response_destination = m_a @ tristmimulus_vector(white_point_d65)
cone_response_ratio = cone_response_destination / cone_response_source
m = np.linalg.inv(m_a) @ np.diagflat(cone_response_ratio) @ m_a
D50_to_D65 = m
xyz_50_to_srgb = xyz_65_to_srgb @ D50_to_D65
print(xyz_50_to_srgb)
print(xyz_65_to_srgb)
```
fe46b2c141 added the reset-temp-inverse flag, but set it up so all
tempinverse ops were negated at the start of the next op; this commit
makes it so these flags actually persist for one op and not zero.
Fixes#2296.
We were previously throwing an exception if the generated code was
throwing an exception before it hit the implementation of the interface.
Instead, we are meant to catch any exception, and wrap that in a
rejected promise.
For example, this was impacting the fixed test in this commit as an
exception was being thrown when invoking WebIDL::convert_to_int<T>
as the given number was out of range, and the [EnforceRange]
extended attribute decorates that attribute.
This same type of case is seen for a few tests in WPT.
This change ensures that:
- if an element for which an accessible name otherwise wouldn’t be
computed is referenced in an aria-labelledby value, the accessible
name for the element will be computed as expected.
- if an element has both an aria-label value and also an
aria-labelledby value, the text from the aria-label value gets
included in the computation of the element’s accessible name.
Otherwise, without this change, some elements with aria-labelledby
values will unexpectedly end up without accessible names, and some
elements with aria-label values will unexpectedly not have that
aria-label value included in the element’s accessible name.
This color space is often used as a reference in WPT tests, having
support for it makes us pass 15 new tests:
- css/css-color/rec2020-001.html
- css/css-color/rec2020-002.html
- css/css-color/rec2020-003.html
- css/css-color/rec2020-004.html
- css/css-color/rec2020-005.html
- css/css-color/predefined-011.html
- css/css-color/predefined-012.html
That makes us pass the following WPT tests:
- css/css-color/prophoto-rgb-001.html
- css/css-color/prophoto-rgb-002.html
- css/css-color/prophoto-rgb-003.html
- css/css-color/prophoto-rgb-004.html
- css/css-color/prophoto-rgb-005.html
- css/css-color/predefined-009.html
- css/css-color/predefined-010.html
This color space is often used as a reference in WPT tests, having
support for it makes us pass 15 new tests:
- css/css-color/display-p3-001.html
- css/css-color/display-p3-002.html
- css/css-color/display-p3-003.html
- css/css-color/display-p3-004.html
- css/css-color/display-p3-005.html
- css/css-color/display-p3-006.html
- css/css-color/lab-008.html
- css/css-color/lch-008.html
- css/css-color/oklab-008.html
- css/css-color/oklch-008.html
- css/css-color/predefined-005.html
- css/css-color/predefined-006.html
- css/css-color/xyz-005.html
- css/css-color/xyz-d50-005.html
- css/css-color/xyz-d65-005.html
This makes us pass the following WPT tests:
- css/css-color/a98rgb-001.html
- css/css-color/a98rgb-002.html
- css/css-color/a98rgb-003.html
- css/css-color/a98rgb-004.html
- css/css-color/predefined-007.html
- css/css-color/predefined-008.html
In commit 1b82cb43c2 I accidentally
removed the paint transformation altogether. The result was that
zoomed-in SVGs, or SVG elements with a transformation applied could have
their gradient coordinates misplaced significantly.
This was also exposed in the `svg-text-effects` test by way of a slight
visual difference. Add a new test that very clearly exposes the fixed
issue by rotating the gradient coordinates by 45 degrees.
In #1537, determine_the_origin() changed to take
`Optional<URL::URL> const&` as first parameter, but it's passed
`Web::Fetch::Infrastructure::Response::url()`, which returns
`Optional<URL::URL const&>`. Ladybird does not have
SerenityOS/serenity#22870 (yet?), so this mismatch silently creates
a copy.
Change determine_the_origin() to take `Optional<URL::URL const&>`
instead. No behavior change, saves a copy, and is probably what
was originally intended.
Before this change, a StringView with a character-data pointer would
never compare as equal to one with a null pointer, even if they were
both length 0. This could happen for example if one is
default-initialized, and the other is created as a substring.
The WebSocket spec tells us to queue tasks instead of firing events
synchronously at WebSockets, so this commit does exactly that.
The way we've implemented web sockets means that the work is spread
across multiple libraries and even processes, which is why it doesn't
look like the spec verbatim.
That makes us pass the following WPT tests:
- css/css-color/srgb-linear-001.html
- css/css-color/srgb-linear-002.html
- css/css-color/srgb-linear-003.html
Previously it was only pushing the module context for the call to
capture the module execution context. This is incorrect, as the capture
occurs upon function construction. This resulted in it capturing the
execution context that execute_module was called from, instead of the
newly created module_context.
f87041bf3a/Libraries/LibJS/Runtime/ECMAScriptFunctionObject.cpp (L92)
This can be demonstrated with the following setup:
index.html:
```html
<script>
var foo = 1;
</script>
<script type="module">
import {test} from "./scriptA.mjs";
</script>
```
scriptA.mjs:
```js
function foo() {
return {a: "b"};
}
export let test = await foo();
```
Before this fix, this would throw:
```
[TypeError] 1 is not a function (evaluated from 'foo')
at module code with top-level await
at module code with top-level await
at <unknown>
at <unknown>
```
Fixes#2245.
Fixes a bug when https://wpt.live/css/CSS2/positioning/abspos-001.xht
saved as file fails because we incorrectly recognized its MIME type
as HTML, leading to incorrect self-closing tag handling and thus
incorrect rendering.
The MessagePort one in particular is required by Cloudflare Turnstile,
as the method it takes to run JS in a worker is to `eval` the contents
of `MessageEvent.data`. However, it will only do this if
`MessageEvent.isTrusted` is true, `MessageEvent.origin` is the empty
string and `MessageEvent.source` is `null`.
The Window version is a quick fix whilst in the vicinity, as its
MessageEvent should also be trusted.
Resulting in a massive rename across almost everywhere! Alongside the
namespace change, we now have the following names:
* JS::NonnullGCPtr -> GC::Ref
* JS::GCPtr -> GC::Ptr
* JS::HeapFunction -> GC::Function
* JS::CellImpl -> GC::Cell
* JS::Handle -> GC::Root
The spec just says to follow "most backwards-compatible, then shortest"
when serializing these (and it does so in a very hand-wavy fashion).
By omitting some keywords when they are implied, we end up matching
other engines and pass a bunch of WPT tests.
This change removes the `--headless` option, which is now the default
behavior and adds the `--show-window` option to force tests to run in a
visible browser window.
This shouldn't just be a simple reflection of the label attribute.
It also needs fallback to the HTMLOptionElement.text property if the
label attribute is absent.
None of the algorithms actually set the `extractable` internal slot in
their implementations, and looking at `SubtleCrypto::import_key()` it
seems likely that a step is missing here.
Fix the function signatures of Canvas.toDataURL() and Canvas.toBlob()
and make both functions accept non-numbers as the quality parameter, in
which case it will just use the default quality instead of raising an
exception.
This makes toDataURL.arguments.1.html, toDataURL.arguments.2.html and
toDataURL.jpeg.quality.notnumber.html in
wpt/html/semantics/embedded-content/the-canvas-element pass :^)
This means that an `<input type=password>` will show the correct number
of *s in it when non-ASCII characters are entered.
We also don't need to perform text-transform on these as that doesn't
affect the output length, so I've moved it earlier.
After we absolutize the contents of :has(), we check that those child
selectors don't contain anything that :has() rejects.
This is a separate path than the checks inside the parser, which is
unfortunate.
Fixes a WPT ref test. :^)
It's possible for absolutizing a selector to return an invalid selector
(eg, it could cause `:has()` inside `:has()`) so we need to be able to
express that.
The CSSOM spec tells us to potentially add up to three different IDL
attributes to CSSStyleDeclaration for every CSS property we support:
- A camelCased attribute, where a dash indicates the next character
should be uppercase
- A camelCased attribute for every -webkit- prefixed property, with the
first letter always being lowercase
- A dashed-attribute for every property with a dash in it.
Additionally, every attribute must have the CEReactions and
LegacyNullToEmptyString extended attributes specified on it.
Since we specify every property we support with Properties.json, we can
use that file to generate the IDL file and it's implementation.
We import it from the Build directory with the help of multiple import
base paths. Then, we add it to CSSStyleDeclaration via the mixin
functionality and inheriting the generated class in
CSSStyleDeclaration.
This allows us to specify multiple base paths to look for imported IDL
files in. This will allow us to import IDL files from sources and from
the Build directory (i.e. for generated IDL files).
We currently have 2 virtual methods to inform DOM::Element subclasses
when an attribute has changed, one of which is spec-compliant. This
patch removes the non-compliant variant.
Instead, smuggle it in as a `void*` private data and let Javascript
aware code cast out that pointer to a VM&.
In order to make this split, rename JS::Cell to JS::CellImpl. Once we
have a LibGC, this will become GC::Cell. CellImpl then has no specific
knowledge of the VM& and Realm&. That knowledge is instead put into
JS::Cell, which inherits from CellImpl. JS::Cell is responsible for
JavaScript's realm initialization, as well as converting of the void*
private data to what it knows should be the VM&.
NanBoxedValue is intended to be a GC-allocatable type which is not
specific to javascript, towards the effort of factoring out the GC
implementation from LibJS.
Attempt 2! Reverts 2a5dbedad4
This time, set up a different combinator when producing a relative
invalid selector rather than a standalone one. This fixes the crash.
Original description below for simplicity because it still applies.
---
Selectors like `:is(.valid, &!?!?!invalid)` need to keep the invalid
part around, even though it will never match, for a couple of reasons:
- Serialization needs to include them
- For nesting, we care if a `&` appeared anywhere in the selector, even
in an invalid part.
So this patch introduces an `Invalid` simple selector type, which simply
holds its original ComponentValues. We search through these looking for
`&`, and we dump them out directly when asked to serialize.
This gets rid of a couple FIXMEs and allows reusing the logic of
validating this field between different algorithms. While we're here,
expand its logic to match the constraints as outlined in RFC 7517.
LibWebView now knows how to launch RequestServer and ImageDecoderServer
without help from the UI, so let's move ownership of these services over
to LibWebView for de-duplication.
This was a silly mistake on my end and percentages values are not
covered by device-independent color space, so I had to add support for
srgb to run a WPT test that made me realize the mistake.
This makes the following test pass:
- css/css-color/predefined-002.html
It makes the following WPT tests pass:
- css/css-color/predefined-001.html
- css/css-color/xyz-003.html
- css/css-color/xyz-d50-003.html
- css/css-color/xyz-d50-004.html
- css/css-color/xyz-d65-003.html
Also we now render the reference of color-mix-currentcolor-nested-for-
color-property.html properly. Which means that it's now different from
the actual test, that is still rendered incorrectly. In other word, the
false positive for this test is now turned into a true negative.
Now that the heap has no knowledge about a JavaScript realm and is
purely for managing the memory of the heap, it does not make sense
to name this function to say that it is a non-realm variant.
The main motivation behind this is to remove JS specifics of the Realm
from the implementation of the Heap.
As a side effect of this change, this is a bit nicer to read than the
previous approach, and in my opinion, also makes it a little more clear
that this method is specific to a JavaScript Realm.
Selectors like `:is(.valid, &!?!?!invalid)` need to keep the invalid
part around, even though it will never match, for a couple of reasons:
- Serialization needs to include them
- For nesting, we care if a `&` appeared anywhere in the selector, even
in an invalid part.
So this patch introduces an `Invalid` simple selector type, which simply
holds its original ComponentValues. We search through these looking for
`&`, and we dump them out directly when asked to serialize.
We will want to re-inform WebContent of the system visibility state when
we create a new process after a crash. This changes the IPC to just send
the enum value directly, instead of a boolean, so that we can just store
that enum value directly on the ViewImplementation class.
These operations should still apply even if they are off screen, because
they affect painting of things outside of their bounding rectangles.
This commit makes us always apply these, regardless of if they are in
the visible region. However, if they are outside that region, we
replace them with simple clip-rect commands, which have the same
effect (not painting anything) but are cheaper than computing a full
mask bitmap.
The insertion steps for iframes were following an old version of the
spec, where it was checking if the iframe was "in a document tree",
which doesn't cross shadow root boundaries. The spec has since been
updated to check the shadow including root instead.
This is now needed for Cloudflare Turnstile iframe widgets to appear,
as they are now inserted into a shadow root.
Previously, the inclusive descendant, which is the node that
for_each_shadow_including_inclusive_descendant was called on, would not
have it's shadow root traversed if it had one.
This is because the shadow root traversal was in the `for` loop, which
begins with the node's first child. The fix here is to move the shadow
root traversal outside of the loop, and check if the current node is an
element instead.
While this does mean that we keep one copy of the stack info in the VM,
and another in the Heap; keeping a separate instance removes one more
instance of coupling between the heap and LibJS specific details.
There is definitely a possibility I am misunderstanding the reason
behind it - but this does not appear neccessary. The VM owns both the
string cache and Heap. On destruction, the VM should clear out both
the heap and its string cache.
While this is used in the implementation of Runtime objects itself, Heap
seems like a more appropriate home. This will also help in factoring out
the GC implementation into it's own library as the heap explicitly has
knowledge of WeakContainer.
When the cached value was not an accessor, it was simply ignored.
This is the value we really want, so we can just return it.
Shows up to 5x improvements on some benchmarks,
and 1.4x in general js-benchmarks.
Async functions whose promise is never resolved were leaking, since they
had a strong root JS::Handle on themselves.
This doesn't appear to actually be necessary, since the wrapper will be
kept alive as long as it's reachable (and if it's not reachable, nobody
is going to resolve/reject the promise either).
This fixes the vast majority of leaks on Speedometer, bringing memory
usage at the end of a full run from ~12 GiB to ~3 GiB.
This fixes an issue where a badly-timed garbage collection could swallow
a static field initializer.
Caught by running test262 in GC-on-every-allocation mode.
We were miscalculating the length of the buffer in pointer-sized chunks,
which is what the conservative root scan cares about.
This could cause some values to be prematurely garbage-collected.
This was preventing https://ubereats.com/ from fully loading, because
they are attempting to overwrite setItem. They seem to be trying to add
error logging to setItem if it throws, as all they do is add a
try/catch block that emits an error log to their monitoring service if
it throws.
However, because Storage is a legacy platform object with a named
property setter (setItem), it will call setItem with the stringified
version of the function. This is actually expected as per the spec,
Firefox (Gecko) and Epiphany (WebKit) does this too, but Chromium does
not as it actually overwrites the function with the new function and
does not store the stringified function.
The problem is that we had the LegacyOverrideBuiltIns flag accidentally
set, so it would return the stored string instead of the built-in
function (hence the name), then it would try and call it and throw a
"not a function" error. This prevented their JS from going any further.
This fix allows their UI to fully load and be fully interactive, though
it is quite slow at the moment!
This change removes the append_without_space, append_with_space,
prepend_without_space, and prepend_with_space functions from DOM::Node.
All those methods were added with the initial “Implement Accessible Name
and Description Calculation” commit in da5c918 and were only used in the
code related to accessible-name computation. But subsequent changes to
that code have removed all the calls to those functions — so now they’re
all completely unused.
This change ensures that when the aria-labelledby attribute is used, the
expected text from the element referenced in the aria-labelledby value
appears in the computed accessible name. Otherwise, without this change,
the expected text doesn’t appear in the computed accessible name.
This change fixes handling for substep ii of the “F. Name From Content”
step at https://w3c.github.io/accname/#step2F in the “Accessible Name
and Description Computation” spec — to correctly include any ::before
and ::after pseudo-element content in the computation of accessible
names. Otherwise, without this change, accessible names unexpectedly
don’t include that pseudo-element content.
This change implements the https://w3c.github.io/accname/#comp_append
step in the “Accessible Name and Description Computation” spec — so that
when an accessible name is computed from multiple sources in a document
subtree, the parts of the computed text are joined together with spaces.
Otherwise without this change, in accessible names computed from
multiple sources in a document subtree, the parts of the computed text
are unexpectedly run together, with no spaces between the parts.
This is really just a type alias for NonnullGCPtr<T>, but it provides
a way to have non-owning non-visited NonnullGCPtr<T> without getting
yelled at by the Clang plugin for catching GC errors.
We were previously dumping the address of the cell pointer instead of
the address of the cell itself. This was causing mysterious orphans
in GC dumps, and it took me way too long to figure this out.
compute_inset() was incorrectly retrieving the containing block size
because containing_block() is unaware of grid areas that form a
containing block for grid items but do not exist in the layout tree.
With this change, we explicitly pass the containing block into
compute_inset(), allowing it to correctly provide the containing block
sizes for grid items.
Explicitly pass containing block width in
resolve_vertical_box_model_metrics() instead of doing containing block
box lookup.
This is a part of refactoring towards removing containing_block() usage
that will allow us introduce partial layout.
The following syntax is valid:
```js
e?.example / 1.2
```
Previously, the `/` would be treated as a unterminated regex literal,
because it was calling the regular `consume` instead of
`consume_and_allow_division`.
This is what is done when parsing IdentifierNames in
parse_secondary_expression when a period is encountered.
Allows us to parse clients-main-[hash].js on https://ubereats.com/
We now skip so many tests that the list of skipped tests exceeds the
height of my terminal. Let's skip logging these by default, as it is
too noisy to find actually relevant information.
We currently compile the Qt event loop files multiple times, for every
target which wants to use them. This patch moves these to LibWebView as
a central location to avoid this.
If available space is definite it should always match the size of the
containing block. Therefore, there is no need to do containing block
node lookup.
The inline capacity on ThreadEventQueue::Private::queued_events caused
us to reserve (and importantly, not initialize!) 2 KiB of stack memory
when entering ThreadEventQueue::process().
This was causing any leftover pointers to GC-allocated objects within
that memory range to keep those objects alive, even when all other
references were gone.
The example shows how to write a test that depends on custom HTTP
headers in the response. This will be useful for testing browser JS
that depends on how Ladybird processes response headers, eg CORS
headers like Access-Control-Allow-Origin and others.
This allows us to simulate HTTP responses from Browser JS tests.
Instead of using hacks like data URLs to "load" external data,
we can now generate an actual HTTP response that contains
arbitrary headers, body, and has a defined response delay.
...by constructing ImmutableBitmap directly from SkImage.
This is a huge optimization for the case when content of canvas is
painted onto another canvas, as it allows pixels to remain in GPU memory
throughout the process.
Fixes performance regression on https://playbiolab.com/ introduced by
switching to GPU-backend for canvas.
By caching the SkImage that is reused across repaints, we allow Skia t
optimize GPU texture caching.
ImmutableBitmap is chosen to own the SkImage because it guarantees that
the underlying pixels cannot be modified. This is not the case for
Gfx::Bitmap, where invalidating the SkImage would be challenging since
it exposes pointers to underlying data through methods like scanline().
Supported:
* Normal absolute and relative paths: C:\Windows\Fonts, AK\LexicalPath.h
* Forward slashes and multiple separators: C:/Windows///\\\notepad.exe
Not supported:
* Paths starting with two backslashes: \\?\C:\Windows, \\server\share
* Unusual relative paths like C:, C:a\b, \, \a\b
More on Windows path formats: https://learn.microsoft.com/en-us/dotnet/standard/io/file-path-formats
Gap values are now represented by Variant<LengthPercentage, NormalGap>.
NormalGap is just an empty struct to represent the `normal` keyword.
This fixes a long-standing issue where we were incorrectly storing gaps
as CSS::Size, which led to us allowing a bunch of invalid gap values.
This method was being used to check for invalid `PropertyKey`s.
Since invalid `PropertyKey`s are no longer created, and since the
associated method has also been removed in the latest edition of
ECMA-262, the method can now be removed here as well.
While we are removing all its calls, lets also update any surrounding
spec comments to the current latest edition, where possible.
This object property kind had completely different behaviour.
By adding an instruction for it, we can remove a bunch of special
casing, and avoid creating dummy `PropertyKey` values.
This constructor was creating an "invalid" `PropertyKey`, but every
code path constructing such a `PropertyKey` was either never used or
immediately `VERIFY`ing that the `PropertyKey` was valid.
The `VERIFY` call has been moved to the `PropertyKey::from_value(...)`
call, and the array/object literal spreading code could be refactored
to not take a `PropertyKey` but creates dummy values for now.
The default constructor for `Reference` has similairly be deleted,
because it was never used.
Lines like these were getting a warning to simplify the expanded
boolean expression from `!(a || b)` to `(a && b)`, but since the
`!(...)` is part of the macro, that is never going to happen.
We will be moving the variants of these files from Ladybird to the
Userland/Services directory. To make the diffs in those commits actually
make sense, let's remove these unsused variants ahead of time.
This change allows you to give http[s]://wpt.live/ URLs to the WPT.sh
script for both the “WPT.sh run” and “WPT.sh import” commands.
That facilitates the use case where you’ve navigated to a wpt.live URL
in a browser, and you want to just directly copy-paste the URL in order
to either run the test in Ladybird, or import the test into the repo.
Otherwise, without this change, when using WPT.sh, you’re limited to
needing to specify either a WPT path fragment or filesystem pathname —
which doesn’t allow for easy copy-paste directly from wpt.fyi.
This also adds support for `xyz` as it defaults to `xyz-d65`. We now
pass the following WPT tests:
- css/css-color/xyz-001.html
- css/css-color/xyz-002.html
- css/css-color/xyz-004.html
- css/css-color/xyz-d65-001.html
- css/css-color/xyz-d65-002.html
- css/css-color/xyz-d65-004.html
This will be usefully on its own later on when supporting this color
space directly, but it also allows us to do some factorization in the
current codebase.
If a & simple selector is on a style rule with no parent style rule,
then it behaves like :scope - but notably, :scope provides 1
specificity in the class category, but & is supposed to provide 0.
To solve this, we stop replacing it directly, and just handle the & like
any other simple selector. We know that if the selector engine ever
sees one, it's equivalent to :scope, because the nested ones will have
been replaced with :is() before that point.
This gets us one more subtest pass. :^)
When we first parse a nested CSSStyleRule's selectors, we treat them as
relative selectors and then patch them up with an `&` as needed.
However, we weren't doing this when assigning the `cssText` attribute.
So, let's do that!
This gives us a couple of subtest passes. :^)
We were delaying sending an IPC message until the HTML PortMessage
task was run, but we were not queuing the task in on the receiver
side. The result of this was that message port posting was
needlessly creating an HTML task just to send an IPC message.
On the flip side, we were directly calling dispatch_event from the
socket notifier of the target port's message queue. This is a huge
problem, because it means that we were effectively running
javascript-aware code from an 'in parallel' context.
By switching around which side of the IPC interface is responsible
for queuing a task, we can avoid problems where a document is
destroyed from a port message-attached callback and crashes.
The Gfx::Bitmap encoder and decoder have been replaced
by BitmapSequence in #1435, making them redundant and safe to remove.
Additionally, the base IPC::encode and IPC::decode implementations
will trigger a compiler error if these methods are called.
.ipc files already exclude references to Gfx::Bitmap.
This change fixes accessible-name computation for:
- an element that has empty text content but that also has a title
attribute (“tooltip”) with a non-empty value
- an img element whose alt attribute is the empty string but that also
has a “title” attribute with a non-empty value
Otherwise, without this change, the accessible name unexpectedly isn’t
computed correctly for those cases.
This lets us redraw the WebView while live resize events are ongoing. By
doing so, we can also update the window rect from within the WebView,
rather than requiring a separate method invocation (which the Inspector
and Task Manager windows were missing).
When the user resizes the browser window, AppKit will spin the current
run loop in the common mode. This patch allows socket events from IPC to
be processed while the resize is ongoing. This includes the WebView hook
to repaint.
Co-Authored-By: Andreas Kling <andreas@ladybird.org>
This adds the vcpkg triplets and CMake preset to perform release
builds for distribution. These builds are fully static, and currently
intended to be used for the `js` ESVU release.
In the future, linking everything statically into the final binary is
probably not what we will do for released Ladybird builds. Instead, we
may have a "libladybird.so", which is then linked into the binary. But
this should be fine for `js` for now.
As shown in the test added by this patch, it was possible to re-assign
the `this` value of a member function call while it was executing.
Let's copy the original this value like we already do with the callee.
Fixes#2226.
The dynamic shared lib build of skia doesn't seem to actually express
any dependencies in its DT_NEEDED section, so we need to force-load
fontconfig into the dependencies of the skia target to avoid runtime
linker errors.
For some reason, Microsoft have decided to remove Xcode 16 from macOS 14
images. We require Xcode 16 for Swift 6.
See: https://github.com/actions/runner-images/issues/10703
Because macOS 15 images are still in preview, their availability is much
lower than macOS 14 images. To hopefully alleviate the amount of time we
are waiting in the runner queue, for now this only upgrades the workflow
which uses Swift.
Paths in `headless-browser` tests were being compared by string value.
With this patch, they will be compared by their real path instead,
ensuring relative paths and symlinks are handled correctly.
A few are skipped for now:
- A few ref tests fail
- Crash tests are not supported by our runner and time out
- top-level-is-scope.html crashes and needs further investigation
This will no longer be true once we implement at-rules nested inside
style rules, for example:
```css
.foo {
color: yellow;
@media (min-width: 800px) {
color: orange;
}
}
```
If a rule gets its caches cleared because it's moved in the OM, then its
child rules' caches are likely invalid and need clearing too.
Assuming that caches only point "upwards", this will correctly clear
them all. For the time being that will be true.
This implements the `is_valid_in_the_current_context()` methods by
maintaining a stack of contexts, such as whether we're inside a style
rule, or an `@media` block, or the condition of a `@supports` rule.
This is implemented by using a GPU-accelerated surface for <canvas> when
a GPU context is available.
A side effect of this change is that all canvas modifications have to be
performed through Gfx::Painter, and whenever its content has to be
accessed, we need to take a snapshot of the corresponding GPU surface.
A new DrawPaintingSurface display list command is introduced to allow
cheap blitting of canvas content without having to read GPU surface
content into RAM.
Since the Vulkan context is currently only used by LibGfx, it could be
moved there to avoid having Vulkan as a dependency for everything that
uses LibCore.
Since the Metal context is currently only used by LibGfx, it could be
moved there to avoid having Metal as a dependency for everything that
uses LibCore.
This is required to share GPU context creation code between display list
player, which resides in LibWeb, and PainterSkia, which handles <canvas>
painting.
We want the "raw" value here. The spec sort of assumes we are sending JS
values over the wire, whereas we are actually sending AK::JsonValue. So
using the JSON clone AO here is our best bet.
Chromium has a larger list of boolean attributes, and WPT depends on it.
We must also compare attribute names case-insensitively.
Note this method is only used by WebDriver.
We can't invalidate after the removal has taken effect, since that means
invalidation won't be able to find potentially affected siblings and
ancestors by traversing from the invalidation target.
This causes 36 new subtests to pass locally. :^)
Unfortunately at least one of these is flaky when it's able to load the
font file, apparently because we don't wait for the font and its
stylesheet to actually load before the tests run.
For example, a few ref tests have a match like this:
```
<link rel="match"
href="/css/reference/ref-filled-green-100px-square-only.html">
```
Previously we'd interpret this as `http://css/reference/...` which made
the test runner sad.
Because of this we no longer have to handle ahem.css in a special way.
This should find:
- <link rel=stylesheet>
- CSS `@import`s
- Any resources linked from a stylesheet with `url()`
There's a good chance there are other resources we'll want to copy too,
but CSS was a big hole.
Eventually we want to stop skipping these, so it's helpful to know why
they were skipped in the first place. :^)
I've grouped them together by reason, so the order has changed a little.
For some of these the reason isn't clear.
This works around an issue in upstream skia where a debug build with
dynamic libraries includes an extra file in the skparagraph module
that is not compilable on macOS.
By using static linkage, we are opening ourselves up to ODR violations
that result in runtime crashes (rather than build-time link errors). For
example, we've seen this happen while trying to use Skia context in both
LibGfx and LibWeb.
In addition to changing the build-type dependent build directories, we
can take this opportunity to move the vcpkg cache directory to the Build
folder itself. This probably isn't 100% needed, but it ensures that no
leftover artifacts are used from non-dynamic vcpkg builds, and it's also
generally nice to have all build artifacts under Build.
The reason for this change is that CMake/vcpkg are unable to detect a
change to VCPKG_LIBRARY_LINKAGE. So when we switch to dynamic builds,
the switch would be non-functional, and every developer would have to
remove their Build and vcpkg cache directories manually. By changing
these directories, vcpkg is able to detect it must rebuild.
Using install(IMPORTED_RUNTIME_ARTIFACTS), we can re-export the
shared library and frameworks we imported from outside the build
tree into our install treer. This is required for simdutf as it
is a dependency of the AK library, and we need liblagom-ak.so to
be loadable when doing fuzzer and cross-compile builds for the
Lagom tools.
While traversing the DOM tree, looking for nodes that need a style
update, we were recomputing style for every node visited along the way,
even nodes that didn't themselves need a style update (but one of their
descendants did).
This avoids a bunch of completely unnecessary style recomputation on
basically every website.
This class was being copied all over the place, however, most of these
cases can be easily prevented with `auto const&` or `NonnullRawPtr<>`.
It also didn't have a move constructor, causing `Vector` to copy on
every resize as well.
Removing all these copies results in an almost 15% increase in
performance for CSS parsing, as measured with callgrind.
It is a non-null `T*` with reference semantics.
Or a `T&` whose address can be copied and re-assigned.
Or a `NonnullRefPtr` whose memory is not managed.
It can be useful when you want to store a reference in a
data structure that needs to be copyable or assignable.
Range API uses UTF-16 code units to represent offsets, so replace_data()
needs to use it instead of bytes count while calculating new offsets.
Fixes incorrectly thrown exception when non-latin string is passed into
replace_data().
The spec is a bit out-of-date here, so this works around an issue with
closing top-level traversables while a dialog is open in another window
within the same agent.
For example, if a dialog is open and the session is configured to ignore
dialogs (instead of automatically closing them), then the Close Window
endpoint will have failed. We can't remove the client-side window handle
in that case.
Capabilities are configured on a per-session basis, but we were only
applying these options to the first WebContent process created. We will
need to pass these options to new windows created from that process.
This patch re-organizes capability processing so that our session can
remember them for new windows.
This is enough for a basic shadow realm to work :^)
There is more that we still need to implement here such as module
loading and fixing up the global object, but this is enough to get some
basic usage working.
This object represents the global object for a shadow realm. The IDL
generator will need to be adjusted to the '[Global]' extended attribute
and no '[Exposed]' field (the change in the test is not correct, as I
understand it), but this should be enough to get us started on
shadow realms.
With the introduction of shadow realms, there will be two different
possible host defined objects. For clarity, rename the existing host
defined object to PrincipalHostDefined.
We were neglecting to check the namespace when looking for a specific
type of element on the stack of open elements in many cases.
This caused us to confuse HTML and SVG elements.
Element::tag_name() returns an uppercased string for HTML elements,
which is usually not what's expected by the parser algorithms that look
at tag names.
Letter spacing is applied during text shaping and `shape_text` is used
in places other `InlineLevelIterator` so way may have more work to do,
however this is a good start :^).
This fixes 3/4 of the remaining test failures in wpt/png/apng. Also,
about half of the APNG files on
https://commons.wikimedia.org/wiki/Category:Animated_PNG_files
used to be broken, and they all seem to work properly now :^)
I also refactored the code to be (at least to me) simpler and more
similar to what the spec describes. As a nice side effect, there are now
fewer lines of code than before.
Lastly, I replaced DeprecatedPainter with Painter.
The current shadow realm constructor implementation was based off a
merge request to the shadow realm proposal for integration into the
web platform. However, this merge request had a bug which we had
applied a workaround for by popping off the execution stack.
Since then, the spec has had an update of:
https://github.com/tc39/proposal-shadowrealm/commit/28b0cc
Which closer aligned the mainline spec to the proposed merge request.
Now, the original shadow realm proposal merge request has been closed
and a new one has been reopened with a much more minimal set of changes
that merely adds extra arguments to HostInitializeShadowRealm, which
this commit aligns with.
Noone needs to use this any more :^)
This is somewhat AD-HOC in the constructor as it is based on an open
shadow realm merge request, but applies the intent of the change without
any change in behaviour.
The use of extract_parameter_arguments_and_body() here is to make things
a little less awkward. If we were to exactly follow spec there would be
an awkward handling of the case that no arguments were provided and we
needed to provide an empty string.
To do this, we would need to either:
- Provide an Optional<Value> for bodyString to CreateDynamicFunction
- Create a new empty PrimitiveString wrapped in a JS Value.
Either case is somewhat awkward. Instead, just refactor this logic
outside of CreateDynamicFunction and make the caller do it.
Otherwise, this commit prepares for the new definition of
HostEnsureCanCompileStrings.
All callers have been made async, i.e. they now all provide a completion
callback. Let's make the callback required to discourage any future sync
usage.
If the attribute value is the empty string `(lang="")`, the language
is set to unknown. `lang` attribute higher up in the document tree
will no longer be applied to the content of that element.
This updates the CSS parser to support the keyword 'none' in the CSS4
color functions. The underlying CSSColorValue already supports this
keyword, meaning the parser can instantiate the color directly.
We track this node ID to navigate to a default element when opening the
Inspector. So for all intents and purposes, <frameset> and <body> should
be treated the same.
We guarded one step against a null navigable, but the very next step
also needs to be protected. Let's just abort early instead. This was
caught by the following imported WPT test:
html/dom/elements/the-innertext-and-outertext-properties/innertext-setter.html
This test adds a <frame> element and immediately removes it, but the
task to process the src attribute is already queued. Note that <iframe>
would have the same issue, but this test does not include them.
NavigableContainer is our home grown concept which already contains the
AOs needed for frame and iframe elements. This patch simply aligns our
HTMLFrameElement implementation with this class.
A couple of notes:
1. The <script> in the <head> element is intentional. The <frameset>
element effectively takes the place of the <body> element, and we
cannot add a <script> to a <frameset> element.
2. We don't render <frameset> or <frame> at all. Rendering is defined
in the following spec:
https://html.spec.whatwg.org/multipage/rendering.html#frames-and-framesets
3. If you load the test page in your browser, you won't see anything,
regardless of (2). Our test infra adds a <pre> element to the "body"
element (which is the <frameset> element here). Such children will
never be rendered. In the future, we could come up with something
better for our test infra to do, but this isn't important anyways
for this test - we can still grab the <pre> element's innerText.
This change allows the test list given to “WPT.sh run” to include full
filesystem relative or absolute pathnames. That facilitates using tab
completion in the shell to browse for pathnames, and also facilitates
copy-paste of full filesystem pathnames. For example:
./Meta/WPT.sh run Tests/LibWeb/WPT/wpt/dom/historical.html
./Meta/WPT.sh run /opt/ladybird/Tests/LibWeb/WPT/wpt/dom/historical.html
Otherwise, without this change, the test list can’t include full
filesystem pathnames, but is instead limited to only path fragments that
specify WPT subdirectory pathnames — which doesn’t allow for tab
completion on pathnames in the shell, nor copy-paste of full pathnames.
Unfortunately, there isn't an exact spec method to get the rendered text
of an element, including its shadow DOM. The WebDriver spec requires
just doing exactly what Selenium does.
This patch does not implement this, but is a step in the right direction
as we will now handle text transforms.
This fixes structured serialization of DataView. It was expected
to be uniform with TypedArray, which returns u32 for byte_offset().
This was covered by a number of WPT infrastructure tests, which this
commit also imports.
Previously, callers were passing the size in bytes, but the method
expected bits. This caused a crash in LibCrypto when verifying the key
size later on.
Also make the naming of local variables and parameters a little more
clear between the different AES algorithms :^)
It's actually possible for there to be no adjusted current node, when
the stack of open elements is empty. This was covered by one of the WPT
parsing tests.
We currently only use these methods with AK JSON objects, but they're
actually spec'd for JS objects, as the WebDriver spec sort of assumes
we are sending encoded JS objects over the wire.
When we fully implement JSON deserialization for executing scripts, we
will need to invoke these AOs with JS objects.
This is a bit unfortunate, but if a function provided to this struct is
overloaded, the C++ compiler cannot distinguish which overload should be
assigned to the Function object. This is explained in detail here:
https://stackoverflow.com/a/30394755
C-style function pointers do work, however, and are fine here because
we only ever assign global free functions to these members.
Instead of maintaining a list of script execution result types, which we
then map to WebDriver error types, just return the WebDriver error that
is specified by the spec. Then perform the JSON clone algorithm from the
caller in WebDriverConnection, again as specified by the spec. To do so,
this moves the JSON clone algorithm to its own file. This will also be
the future home of the JSON deserialize algorithm, which will need some
of the internal AOs implemented there.
DOMTokenList and FileList do not have the 'length' own property - their
prototypes have this property instead. So we must go through [[Get]] to
retrieve this property, which will consider the prototype.
This is an ad-hoc change to account for the fact that we may run
arbitrary code while waiting for the tasks in this function to complete.
Same exact idea as ac48222ed7, once again.
If we reach the end of the token stream when "ignoring and moving on
to the next token" in the "in body" state, we should just not move
on to the next token, since there isn't one.
Covered by various WPT HTML parsing tests that will be imported in
a subsequent commit.
Use the `writing-mode` property to determine what values should be used
for computing each element's rect on the screen. If it is a vertical
mode, swap the inline and block, lengths and offsets.
This only lays out whole inline formatting contexts vertically, and does
not currently support mixing the two orientations in a single context.
This patch separates the notion of x, y, width, and height, from
inline_offset, block_offset, inline_length, and block_length.
These can then be used to compute the final screen positions,
in respect of the desired layout direction. This is the terminology
used in https://drafts.csswg.org/css-writing-modes/#text-flow
This makes it possible to use this layout algorithm to flow
text in any direction. For example, vertically.
This save() call did not have matching restore(). For mask application
it's display list builder responsibility to emit save() and restore()
so mask is applied only to relevant portion.
Progress on https://www.jetbrains.com/
The `[[GetOwnProperty]]` internal method invocation in
`OrdinarySetWithOwnDescriptor` was being invocated again with the same
parameters in the `[[DefineOwnProperty]]` internal method that is also
later called in `OrdinarySetWithOwnDescriptor`.
The `PlatformObject.[[DefineOwnProperty]]` has similair logic.
This change adds an optional parameter to the `[[DefineOwnProperty]]`
internal method so the results of the previous `[[GetOwnProperty]]`
internal method invocation can be re-used.
We have to handle user prompts during the exection of most WebDriver
endpoints. Of the ~50 endpoints which call this AO, ~20 are currently
currently async and handled here.
There are approximately 1000 WPT subtests that rely on the handling of
user prompts being completely asynchronous. It will take a bit of elbow
grease to make all of our WebDriver endpoints comply with this. So for
now, we will deprecate the currently synchronous implementation, and a
future patch will implement an asynchronous version that already-async
endpoints can use.
Some WebDriver hooks will need to inform the client that execution has
completed, but without any knowledge of what endpoint was running. Since
there can only ever be a single WebDriver endpoint executing at once, we
can replace the completion callbacks with a single callback.
If a dialog is opened while a script is executing, we must give control
back to the WebDriver client. The script must also continue executing
though, so once it completes, we ignore its result.
This allows the script to end with a comment, which is tested by WPT.
Otherwise, an ending comment would create a function of the form:
function() { return 1; // comment }
And the script would fail to parse.
This allows you to get the module map for any realm, whether it is a
principal or synthetic realm. We don't yet have the concept of a
synethetic realm, but this puts the groundwork in place for it.
When mime essence matching, the spec only asks for a string comparison
ignoring ascii case. The whitespace trimming and parsing of the mime
produces unexpected and wrong results.
Fixes tests on WPT html/semantics/scripting-1/the-script-element :^)
This change implements the “is a descendant of a native host language
text alternative element” condition in the “F: Name From Content” step
at https://w3c.github.io/accname/#step2F in the “Accessible Name and
Description Computation” spec — to ensure that all descendant nodes get
included as expected in computations for accessible names for elements.
Otherwise, without this change, Ladybird unexpectedly skips descendant
element nodes when computing accessible names — which can result in the
wrong accessible name being returned.
This is a bit of a chonkier commit as it results in both:
clean_up_after_running_callback and prepare_to_run_callback being
changed to accept a realm instead of an environment settings object,
which has a bunch of fallout, particuarly for IDL abstract operations.
Instead of a settings object. This matches updates to the HTML spec as
part of the shadow realm proposal, and begins the refactor of running
scripts on a realm instead of a settings environment object.
Some of the spec steps are slightly messy here (such as in
MainThreadVM.cpp) as this partially implements the ShadowRealm changes
but not other pieces which we have not implemented yet, such as
preparing to run a script also being based on a realm instead of an
environment. But this will be addressed in further commits.
In terms of the 'current principal realm' definition.
No functional impact, as we still need to implement current principal
realm once the surrounding infrastructure is in place. But it is one
less place which needs to be updated when that is all in place :^)
Aligning the name with the the PR implementing the javascript
shadow realm proposal into the web platform. This commit
simply performs the rename before implementing the behaviour
change.
The actual change to the behaviour of the AO is not implemented in this
commit to support 'synthetic' shadow realms as the surrounding
infrastructure is not in place yet.
Not all specs have a MR open to align with this proposed change to the
HTML standard. But in this case we can just apply the same mechanical
change everywhere.
At some point we must have broken the ability of running XHTML test
cases, so the whitespace expectation changes have not been rebaselined
for this test case.
This is an attempt to bring the size of Parser.cpp down. No code
changes, just moves and some explicit template instantiations now that
we're using them from a different file.
With a8077f79cc Selection object is no
longer aware of selection state inside text controls (<textarea> and
<input>), so this change makes them responsible for dispatching
`selectionchange` if their selection state was changed.
echo"deb [signed-by=/usr/share/keyrings/llvm-snapshot.gpg.key] http://apt.llvm.org/${DISTRIB_CODENAME}/ llvm-toolchain-${DISTRIB_CODENAME} main"| tee -a /etc/apt/sources.list.d/llvm.list
if[ ! "${LLVM_VERSION}"="trunk"];then
echo"deb [signed-by=/usr/share/keyrings/llvm-snapshot.gpg.key] http://apt.llvm.org/${DISTRIB_CODENAME}/ llvm-toolchain-${DISTRIB_CODENAME}-${LLVM_VERSION} main"| tee -a /etc/apt/sources.list.d/llvm.list
echo"deb [signed-by=/usr/share/keyrings/llvm-snapshot.gpg.key] http://apt.llvm.org/${DISTRIB_CODENAME}/ llvm-toolchain-${DISTRIB_CODENAME} main"| tee -a /etc/apt/sources.list.d/llvm.list
if[ ! "${LLVM_VERSION}"="trunk"];then
echo"deb [signed-by=/usr/share/keyrings/llvm-snapshot.gpg.key] http://apt.llvm.org/${DISTRIB_CODENAME}/ llvm-toolchain-${DISTRIB_CODENAME}-${LLVM_VERSION} main"| tee -a /etc/apt/sources.list.d/llvm.list
run:sqlite3 $HOME/Library/Application\ Support/com.apple.TCC/TCC.db "INSERT OR IGNORE INTO access VALUES ('kTCCServiceMicrophone','/usr/local/opt/runner/provisioner/provisioner',1,2,4,1,NULL,NULL,0,'UNUSED',NULL,0,1687786159,NULL,NULL,'UNUSED',1687786159);"
// Bail if left is somehow bigger than right and return default constructed result
// FIXME: This can also occur when the collection is empty maybe propagate this error somehow?
// returning 0 would be a really bad thing since this returns and index and that might lead to memory errors
// returning in ErrorOr<size_t> here might be a good option but this is a very specific error that in nearly all circumstances should be considered a bug on the callers site
VERIFY(left<=right);
// If there's only one element, return that element
// This header should be included only in cpp files.
// It should be included after all other files and should be separated from them by a non-#include line to prevent clang-format from changing header order.
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.
