There are a bunch of situations where we need to treat cross axis
max-size properties as "none", notably percentage values when the
reference containing block size is an intrinsic sizing constraint.
This fixes an issue where flex items with definite width would get
shrunk to 0px by "max-width: 100%" in case the item itself is an
SVG with no natural width or height.
For consistency, we now use the should_treat_max_width/height_as_none
helpers throughout FFC.
This makes the search/account/cart icons show up in the top right
on https://twinings.co.uk :^)
Mostly because I audited all places that assigned to `m_text_matrix`
after #22760.
This one is very difficult to trigger in practice.
`show_text()` marks the text rendering matrix dirty already,
so this only has an effect if the `TJ` array starts with a
number, and the matrix isn't marked dirty going in.
`Tm` caches the text rendering matrix, so I changed text.pdf
to contain:
```
1 0 0 1 45 130 Tm
[ 200 (Hello) -2000 (World) ] TJ T*
```
This first sets an x offset of 5 (on top of the normal 40), and
then undoes it (`200` is multiplied by font size (25) / -1000,
and `200 * 25 / -1000` is -5). Before this change, the topmost
"Hello World" ended up slightly indented.
Likely no behavior change in practice, but makes the code easier
to understand, and maybe it helps in the wild somewhere.
0000342.pdf page 5 contains this snippet:
```
/T1_1 10.976 Tf
0 -31.643 TD
(This)Tj
1 0 0 1 54 745.563 Tm
22.181 -31.643 Td
[(vehicle)-270.926(uses)...
```
The `Tm` marked the text rendering matrix as dirty at the start,
but it then calls calculate_text_rendering_matrix() almost in the
next line, which recalculates the text rendering matrix and caches
the new matrix. The `Td` used to not mark it as dirty, and we'd
draw "vehicle" with an incorrect matrix.
This commit effectively just does a bulk update of this function to the
spec. Since there have been so many spec changes, no specific change was
made in mind, and many FIXMEs have been left for where we are still out
of date.
These changes also appear to include a normative change to the temporal
spec which was previously resulting in timeouts for some tests, and is
now resulting in a timeout.
Furthermore, this also resolves some crashes by protecting against
division by zero, instead throwing a RangeError. This can only happen
when a custom calender is provided which returns funky values. See:
https://github.com/tc39/proposal-temporal/commit/ed85e9
Diff Tests:
+8 ✅ -4 💀 -4 💥️
The spec has moved along quite a bit since this was originally
implemented. Catch up on at least some of these changes, and leave
FIXMEs for what is missing.
No change to test262.
This really just takes the [App].Name config value and removes the
ampersands `&` from the name. These ampersands are hotkeys for the
system menu. Instead of typing them twice - error prone - use the
fact that name is just menu_name without the ampersand. Remove the
ampersand to use as the name and use it as is as the menu_name.
All the data is passed using the `Metadata` object, which has a
`main_tags` method. This method should be used when displaying only a
few main tags, for example to fill the property window of a file
manager. Another method returning the entire list of tags will be
implemented later on.
Previously, any TableView column could be made visible through a
context menu shown by right clicking on the table header. This change
allows columns to be marked as non-selectable, so their visibility
cannot be toggled in this way.
Because of the way sockets are implemented, a local socket object inside
pflocal has little to do with the filesystem node that has an ifsock
translator sitting on it; and the latter doesn't even exist until you
bind() the socket.
So it's not possible to set the socket's mode using Unix-level APIs
(other than by temporarily changing umask). It's still possible to do
this with Mach-level APIs, essentially doing the same thing as glibc's
bind() implementation does, but supplying the desired mode instead of
0666, but let's not go there.
Previously, when calling `BigFraction::from_string()`, the fractional
part of the number was always treated as positive. This led to an
incorrect result if the input string was negative.
By default, a SpinBox's value should be unlimited, (or as close as we
can get to that,) and then the GML or code can impose a limit if
needed. This saves the developer from entering an arbitrary "big" max
value when they want the value to have no maximum.
I've audited the use of SpinBox and added `min: 0`, and removed a `max`,
where appropriate. All existing SpinBoxes constructed in code have a
range set explicitly as far as I can tell.
This is the same fix as 07928129dd
also applied to the SVG script element case. Fixes a crash for the
following HTML:
```html
<svg>
<script>
location.reload();
</script>
</svg>
```
As with the linked commit:
> With this change WebContent does not crash when `location.reload()` is
> invoked but `Navigable::reload()` still not working because of spec
> issue (whatwg/html#9869) so we can't add a
> test yet.
HTML fragments are parsed with a temporary HTML document that never has
its flag set to say that it is ready to have scripts executed. For these
fragments, in the HTMLParser, these scripts are prepared, but
execute_script is never called on them.
This results in the HTMLParser waiting forever on the document to be
ready to have scripts executed.
To fix this, only wait for the document to be ready if we are definitely
going to execute a script.
This fixes a hang processing the HTML in the attached test, as seen on:
https://github.com/SerenityOS/serenityFixes: #22735
Previously, constructing a `UnsignedBigInteger::from_base()` could
produce an incorrect result if the input string contained a valid
Base36 digit that was out of range of the given base. The same method
would also crash if the input string contained an invalid Base36 digit.
An error is now returned in both these cases.
Constructing a BigFraction from string is now also fallible, so that we
can handle the case where we are given an input string with invalid
digits.
This will be useful in the upcoming listdir utility (in the next commit)
to get the file type which is obtained in the get_dir_entries syscall,
so it's not changed later by the fstatat syscall.
This will ensure that we get the raw file type value as it's represented
by directory entries from the get_dir_entries syscall.
The paintable tree structure more closely matches the painting order
when fragments are owned by corresponding inline paintables. This
change does not affect the layout tree, as it is more convenient for
layout purposes to have all fragments owned by a block container in
one place.
Additionally, this improves performance significantly on pages with
many fragments, as we no longer have to walk the ancestor chain up
to the closest block container to determine if a fragment belongs
to an inline paintable.
