The spec states to only try to close the session *if* it exists. This
situation can occur when closing a session after a Close Window command,
as the session will be closed automatically if it was the last window.
These steps now have more than one caller; specifically, they may be
called from the Delete Session and Close Window endpoints. The session
was only removed from the active session map for the former endpoint.
Instead, let's more accurately handle removing the session where the
spec tells us to, so that all callers properly perform this step.
When some WebDriver spec steps are implemented a bit more literally, we
will end up in a situation where we remove a session from its client's
active session map, but still have more steps to perform. Currently,
when we remove the session, it is immediately destroyed because it is
stored in an OwnPtr. Instead, we can store it as a RefPtr, which will
let the caller to such steps keep the session alive until the subsequent
steps are complete.
While here, this also changes the storage of active sessions to a
HashMap, as all lookups into it are currently a linear search.
WebDriver::Session::close_window may invoke Session::stop, which needs
the WebContent connection to still exist. Do not remove the window's
handle (thus destroying the connection) until it is no longer needed.
With current architecture every window has its own WebContent process
and there is one WebDriver process that is responsible for talking to
all opened windows. It thus make sense to manage open windows from
WebDriver process instead of WebContent process that is not supposed
to know about all other opened WebContent processes.
This mostly reverts 826d5f8f9a but also
adds `web_content_connection` to window structure and window id
generation (currently out of spec).
With these changes `get_window_handles`, `switch_to_window` and
`close_window` start to actually switch, close and returned handles
of currently opened windows.
This class had slightly confusing semantics and the added weirdness
doesn't seem worth it just so we can say "." instead of "->" when
iterating over a vector of NNRPs.
This patch replaces NonnullRefPtrVector<T> with Vector<NNRP<T>>.
This changes the parameters parsed from a WebDriver HTTP request to
String for transferring over IPC. Conveniently, most locations these
were ultimately passed to only need a StringView.
This also removes DirIterator::error_string(), since the same strerror()
string will be included when you print the Error itself. Except in `ls`
which is still using fprintf() for now.
This is not guaranteed to always work correctly as ArgsParser deals in
StringViews and might have a non-properly-null-terminated string as a
value. As a bonus, using StringView (and DeprecatedString where
necessary) leads to nicer looking code too :^)
With the GC heap conversion, the functionality of legacy platform
objects was broken. This is because the generated implementation of one
of them was used for all of them, removing functionality such as
deletion.
This re-adds all functionality, where questions such as "does the
object support indexed properties?" is instead answered by virtual
functions instead of by the IDL generator checking the presence of
certain keywords/attributes.
The name "initial containing block" was wrong for this, as it doesn't
correspond to the HTML element, and that's specifically what it's
supposed to do! :^)
Nobody made use of the ErrorOr return value and it just added more
chance of confusion, since it was not clear if failing to sniff an
image should return an error or false. The answer was false, if you
returned Error you'd crash the ImageDecoder.
We are currently converting parsed expiry times to local time, whereas
the RFC dictates we parse them as UTC. When expiring cookies, we must
also use the current UTC time to compare against the cookies' expiry
times.
Instead of just calling JS::Value::to_string_without_side_effects() when
printing values to the console, have all the console clients use
the same JS::Print that the REPL does to print values.
This method leaves some things to be desired as far as OOM hardening
goes, however. We should be able to create a String in a way that
doesn't OOM on failure so hard.
This subdirectory is meant to hold all constant data related to the
kernel. This means that this data is never meant to updated and is
relevant from system boot to system shutdown.
Move the inodes of "load_base", "cmdline" and "system_mode" to that
directory. All nodes under this new subdirectory are generated during
boot, and therefore don't require calling kmalloc each time we need to
read them. Locking is also not necessary, because these nodes and their
data are completely static once being generated.
This creates a cached bitmap for each unique screen resolution, which
allows us to share it between displays with the same resolution. If
the resolution is the same as the wallpaper, we can just use the
wallpaper as-is.
The patch also contains modifications on several classes, functions or
files that are related to the `JPGLoader`.
Renaming include:
- JPGLoader{.h, .cpp}
- JPGImageDecoderPlugin
- JPGLoadingContext
- JPG_DEBUG
- decode_jpg
- FuzzJPGLoader.cpp
- Few string literals or texts
Because of interdependencies between DOM::Event and UIEvents::MouseEvent
to template function fire_an_event() in WebDriverConnection.cpp, the
commit: 'LibWeb: Make factory methods of UIEvents::MouseEvent fallible'
have been squashed into this commit.
This reverts the SystemServer exec() logic to how it was before
81bd91c, but now with some extra TRY()s. This allows the HOME var
to always be propagated from LoginServer which prevents needing
to unveil() /etc/passwd everywhere.
This filesystem is based on the code of the long-lived TmpFS. It differs
from that filesystem in one keypoint - its root inode doesn't have a
sticky bit on it.
Therefore, we mount it on /dev, to ensure only root can modify files on
that directory. In addition to that, /tmp is mounted directly in the
SystemServer main (start) code, so it's no longer specified in the fstab
file. We ensure that /tmp has a sticky bit and has the value 0777 for
root directory permissions, which is certainly a special case when using
RAM-backed (and in general other) filesystems.
Because of these 2 changes, it's no longer needed to maintain the TmpFS
filesystem, hence it's removed (renamed to RAMFS), because the RAMFS
represents the purpose of this filesystem in a much better way - it
relies on being backed by RAM "storage", and therefore it's easy to
conclude it's temporary and volatile, so its content is gone on either
system shutdown or unmounting of the filesystem.
To do this properly, we also create Strings with formatting of device
nodes' names, taking into consideration errors when doing that.
Also, we use LibCore System mknod method instead of raw LibC functions
to be able to propagate errors from these calls too.
It's currently possible for the callback of a file request to request
more file objects. This could cause the hash map storing these requests
to be rehashed while one of its callbacks is being invoked. AK::Function
explicitly forbids this with an assertion.
Instead, remove the callback from the hash map before invoking the
callback function.
The shadowRoot property getter that will be added in subsequent commits
has an additional check that checks whether the shadow root is opened.
I didn't update the function logic to match with the IDL interface,
because it's very likely we don't want that check in the existing code,
so that for example closed shadow root elements can still be updated.
This adds support for WebSocket subprotocols to WebSocket DOM
objects, with some necessary plumbing to LibWebSocket and its
clients.
See the associated pull request for how this was tested.
When a file cannot be accessed, we currently send errno as the error
code. However, there are system calls which occur (by way of dbgln)
between the failed file access and accessing errno. This prevents the
client-side detection of ENOENT from working.
Instead, send over the error we already have stored in the ErrorOr
object.
This patch also includes some changes in the way that the environment
and arguments are passed to `exec`. It was needed to fit the signature
of `Core::System::exec`. That's beneficial though, as we are now doing
`String` manipulation in a fallible environment, so we can propagate
more errors.
There is currently a memory leak with these file request objects due to
the callback on_file_request_finish referencing itself in its capture
list. This object does not need to be reference counted or allocated on
the heap. It is only ever stored in a HashMap until a response is
received from the browser, and it is not shared.
`Stream` will be qualified as `AK::Stream` until we remove the
`Core::Stream` namespace. `IODevice` now reuses the `SeekMode` that is
defined by `SeekableStream`, since defining its own would require us to
qualify it with `AK::SeekMode` everywhere.
ARIA has its own spec and is not part of the DOM spec, which is what the
Web::DOM namespace is for (https://www.w3.org/TR/wai-aria-1.2/).
This allows us to stay closer to the spec with function names and don't
have to add the word "ARIA" to identifiers constantly - the namespace
now provides that clarity.
Note that as of this commit, there aren't any such throwers, and the
call site in Heap::allocate will drop exceptions on the floor. This
commit only serves to change the declaration of the overrides, make sure
they return an empty value, and to propagate OOM errors frm their base
initialize invocations.
This replaces the FlyStrings for ARIA roles that were constructed in
a [[gnu::constructor]] with a single enum. I came across this as the
DOM inspector was crashing due to a null FlyString for an ARIA role.
After fixing that, I was confused as to why these roles were not an
enum. Looking at the spec there's a fixed list of roles and switching
from references to static strings to an enum was pretty much an
exercise in find and replace :).
No functional changes (outside of fixing the mentioned crash).
Before 649f78d0a4, the is_animated and
loop_count objects were set directly when making a return object.
That commit moved the decode logic to a separate function but forgot to
assign `is_animated` and `loop_count`. The compiler didn't throw an
error about unused variables because we were also VERIFY()ing that these
variables were zero-initialized at the beginning of the function.
Having an alias function that only wraps another one is silly, and
keeping the more obvious name should flush out more uses of deprecated
strings.
No behavior change.
The main change here is to implement and use the "container for element"
algorithm. But also, adjust the errors we return. Errors thrown by
`scroll_element_into_view()` are not related to the scrolling itself,
so should not claim to be. `UnsupportedOperation` is more accurate than
`InvalidArgument` when we're expressing that the operation isn't fully
implemented.
This used to be optional and was disabled in two cases:
- On a mouse move event during dragging; because double clicks are
only possible on mouse up events, this had no effect.
- On a mouse event for automatic cursor tracking; this has now gained
support for double click events.
Since it's always enabled now, we can remove the `bool` argument.
This header has always been fundamentally a Kernel API file. Move it
where it belongs. Include it directly in Kernel files, and make
Userland applications include it via sys/ioctl.h rather than directly.
When trying to figure out the correct implementation, we now have a very
strong distinction on plugins that are well suited for sniffing, and
plugins that need a MIME type to be chosen.
Instead of having multiple calls to non-static virtual sniff methods for
each Image decoding plugin, we have 2 static methods for each
implementation:
1. The sniff method, which in contrast to the old method, gets a
ReadonlyBytes parameter and ensures we can figure out the result
with zero heap allocations for most implementations.
2. The create method, which just creates a new instance so we don't
expose the constructor to everyone anymore.
In addition to that, we have a new virtual method called initialize,
which has a per-implementation initialization pattern to actually ensure
each implementation can construct a decoder object, and then have a
correct context being applied to it for the actual decoding.
Originally I simply thought that passing file paths is quite OK, but as
Linus pointed to, it turned out that passing file paths to ensure some
files are able to be decoded is awkward because it does not work with
images being served over HTTP.
Therefore, ideally we should just use the MIME type as an optional
argument to ensure that we can always fallback to use that in case
sniffing for the correct image type has failed so we can still detect
files like with the TGA format, which has no magic bytes.
Because TGA images don't have magic bytes as a signature to be detected,
instead assume a sequence of ReadonlyBytes is a possible TGA image only
if we are given a path so we could check the extension of the file and
see if it's a TGA image.
When we know the path of the file being loaded, we will try to first
check its extension, and only if there's no match to a known decoder,
based on simple extension lookup, then we would probe for other formats
as usual with the normal sniffing method.
InodeWatcherFlags is an enumeration from the Kernel. To avoid using it
outside of Serenity, add a FileWatcherFlags for FileWatcher, much like
we already have FileWatcherEvent::Type.
This is currently being implicitly including by InodeWatcherEvent.h by
way of FileWatcher.h. The former will soon be removed from the latter,
which would otherwise cause a compile error in these files.
