The verified pixel output in this test just reflects currently observed
behavior, I have not verified that all cases output the correct data wrt
what the spec expects.
The `Bitmap` type was referring to to its internal pixel format by a
name that represents the order of the color components as they are layed
out in memory. Contrary, the `Color` type was using a naming that where
the name represents the order of the components from most to least
significant byte when viewed as a unsigned 32bit integer. This is
confusing as you have to keep remembering which mental model to use
depending on which code you work with.
To unify the two, the naming of RGBA-like colors in the `Color` type has
been adjusted to match the one from the Bitmap type. This seems to be
generally in line with how web APIs think about these types:
* `ImageData.pixelFormat` can be `rgba-8unorm` backed by a
`Uint8ClamedArray`, but there is no pixel format backed by a 32bit
unsigned type.
* WebGL can use format `RGBA` with type `UNSIGNED_BYTE`, but there is no
such format with type `UNSIGNED_INT`.
Additionally, it appears that other browsers and browser-adjacent
libraries also think similarly about these types:
* Firefox:
https://github.com/mozilla-firefox/firefox/blob/main/gfx/2d/Types.h
* WebKit:
https://github.com/WebKit/WebKit/blob/main/Source/WebCore/platform/graphics/PixelFormat.h
* Skia:
https://chromium.googlesource.com/skia/+/refs/heads/main/include/core/SkColorType.h
This has the not so nice side effect that APIs that interact with these
types through 32bit unsigned integers now have the component order
inverted due to little-endian byte order. E.g. specifying a color as hex
constant needs to be done as `0xAABBGGRR` if it is to be treated as
RGBA8888.
We could alleviate this by providing endian-independent APIs to callers.
But I suspect long-term we might want to think differently about bitmap
data anyway, e.g. to better support HDR in the future. However, such
changes would be more involved than just unifying the naming as done
here. So I considered that out of scope for now.
We should invert CMYK data only if color space is JCS_CMYK and either
there is no Adobe marker, or the Adobe transform is 0. Transform 2
indicates YCCK data, which we should not invert.
The decoder was requiring GIF files to be at least 32 bytes, but the
actual minimum for a valid GIF is only 26 bytes:
- 6 bytes for the header
- 7 bytes for the Logical Screen Descriptor
- 10 bytes for the Image Descriptor
- 2 bytes for the LZW minimum code size and block terminator
- 1 byte for the GIF trailer
This change allows us to load minimal 1x1 GIFs with empty LZW data.
They are commonly used on the web as transparent placeholders with
minimal file size.
These actually were always working since we first enabling LibGfx
on Windows. I was just running them outside of the ctest context
and therefore had the wrong working directory so the test-inputs folder
could not be found
All the file-based tests left out build, but they all fail at run time
with the error "No such file or directory" or a Core::File-based
assertion failure for the Benchmark test
Gfx::Bitmap only supports a bit depth of 8, therefore we refused to
load AVIF images which didn't have this bit depth.
However, we can tell the libavif decoder to reduce the output depth by
setting avifRGBImage.depth to 8. This allows us to support any input
depth.
Makes images load on https://www.ikea.com/ which uses Cloudflare Images
to re-encode their images to 16-bit AVIF.
In the previous fix, we were still drawing IDAT data to the reference
frame even when no fcTL was present. This would cause rendering issues
when subsequent frames use APNG_BLEND_OP_OVER blending mode, as they
would composite over the incorrect reference frame. This commit adds a
simple check to properly skip any frame without an fcTL chunk.
We now properly handle OS/2 format BMPs that use 3 bytes per color
entry instead of 4. While OS/2 2.x officially specified 4 bytes per
color, some tools still produce files with 3-byte entries. We can
identify such files by checking the available color table space.
This extends ICO loader to support Windows cursor files. There is no
point in creating a separate loader for this, as the ICO format is
very similar to the CUR format. The only differences are bytes used to
identify the file and a presence of a hotspot in the CUR header.
Color masks should only be used when the compression type is either
BITFIELDS or ALPHABITFIELDS. They were always read before and produced
corrupted images when there was random data in the mask fields.
Other browsers don't think that BMP files with more than 1024 colors are
invalid. They clamp the palette instead, and now we do the same. This
allows us to load more BMPs.
Typeface::try_load_from_externally_owned_memory() relies on that
external owner keeping the memory around. However, neither WOFF nor
WOFF2 do so - they both create separate ByteBuffers to hold the TTF
data. So, rename them to make it clearer that they don't have any
requirements on the byte owner.
Before this change, IDAT data was mistakenly always included in the
animation. Now we only include frames with explicit fcTL chunks.
As per the PNG spec (third edition):
"The static image may be included as the first frame of the animation
by the presence of a single fcTL chunk before IDAT. Otherwise, the
static image is not part of the animation."
We also fall back to the IDAT data when APNG has acTL but no fcTL
chunks. Test image is 062.png from fDAT-inherits-cICP.html from WPT.
Bitmap::get_pixel() was only handling two out of the four possible pixel
formats, asserting when called with the other two. The asserting code
path was triggered when loading JPEG XL images, causing crashes on pages
like https://jpegxl.info/resources/jpeg-xl-test-page or
https://html5test.co/.
GIF loader was completely failing when encountering errors with
frame descriptors or individual frames, even when some frames were
successfully loaded. Now we attempt to decode at least some frames
and fail only when no frames can be decoded at all.
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.
- 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
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)
```
Fixes wpt/png/exif-chunk.html.
At some point there should probably be some mechanism to handle this
outside of the individual decoder plugins. The TIFF decoder seems to
have its own version of this, and as far as I can tell, the JPEG decoder
doesn't handle this at all, even though that's probably the most common
use case for Exif orientations. :^)
These files seem to have been marked as executable by error.
Found by running the command:
find \( -name WPT -or -name Toolchain -or -name Build \) \
-prune -or -executable \! -type d -print \
| grep -Pv '\.(sh|py)$'
This will be the first step is making better use of system libraries
like fontconfig and CoreText to load system fonts for use by the UI
process and the CSS style computer.
This change removes wrappers inherited from Gfx::Typeface for WOFF and
WOFF2 fonts. The only purpose they served is owning of ttf ByteBuffer
produced by decoding a WOFF/WOFF2 font. Now new FontData class is
responsible for holding ByteBuffer when a font is constructed from
non-externally owned memory.
It currently doesn't support animated image.
Note that Gfx::Bitmap has no support for get_pixel when the format is
RGBA8888. This is why it has been removed from the tests.
This matches libwebp (see ZeroFillCanvas() call in
libwebp/src/demux/anim_decode.c:355 and ZeroFillFrameRect() call
in line 435, but in WebPAnimDecoderGetNext()) and makes files
written e.g. by asesprite look correct -- even though the old
behavior is also spec-compliant and arguably makes more sense.
Now nothing looks at the background color stored in the file.
See PR for an example image where it makes a visible difference.
Cherry-picked from serenityos master
276a904d20ffe260b5544a9ace9841d083e0243
