Current code was only taking into account cancelation, so we could end
with weird plug-in return:
> HEIF/HEIC plug-in returned SUCCESS but did not return an image
Instead properly set an error status, and echo back the error from the
plug-in or libheif.
Use gimp_image_{freeze,thaw}_layers(), added in the previous
commit, to suppress updates to the Layers dialog while constructing
the decomposed layers, which significantly speeds up the operation
if the dialog is mapped.
Thanks a lot to Dirk for contributing this, added him to AUTHORS.
Import the code from https://github.com/strukturag/heif-gimp-plugin.git
as of today. Merged the files into a single-file plug-in. Changed
the code a lot to match our coding style, but only formatting,
no logic changes.
Still uses deprecated GimpDrawable API and no GIO, but I wanted to do
actual code changes separately from the initial import. Also disabled
metadata support because updating that to GimpMetadata was too much
for the initial import.
Some locales (like German) use ',' as the decimal separator, so using
printf like functions to output a float value fails when the result
should always use '.'. Since there are only two distinct possibilities
we just write the string representation and be done with it.
Exchanging with OpenJPEG developers and searching more on the topic, it
seems that YUV is more often refered to as YCbCr. Wikipedia says:
> typically the terms YCbCr and YUV are used interchangeably, leading to
> some confusion. The main difference is that YUV is analog and YCbCr is
> digital
As for eYCC, I am told this is extended YCC. It seems this is refered as
xvYCC (I really can't find much under "eYCC"). So let's rename it too.
Hopefully I made no mistakes!
JPEG 2000 codestream doesn't have a header and guessing the color space
in particular is not foolproof (especially when 3 or 4 components, which
can be many spaces). Therefore the need of a parameter on the API.
Note that JP2 images should always have the color space information. In
interactive mode, I try to be a bit flexible to salvage broken JP2 with
no color space information in the header, but I am not adding a
parameter in file-jp2-load() (on purpose, since we are not going to add
in the API a parameter for a case not supposed to happen with properly
encoded files).
JPEG 2000 codestream (.j2k/.j2c) are only compressed code stream data,
without header. In particular we don't have color information, such as
the color space. So we need to open a dialog asking to set the color
space in interactive mode.
Note: according to OpenJPEG developers, a JP2 image (not codestream)
should always have a color space defined in its header. But just to be
flexible, the same dialog may get raised as well if we try to load a JP2
with no valid color space defined in header and no ICC profile embedded.
Maybe if such a thing happened, it means the image is corrupt, yet we
may as well try and salvage it anyway.
Note 2: I also removed a weird test which was setting some images as
being YUV color space by mistake. This actually fixes bug 794413 as a
side effect.
Even though I haven't seen working samples with this extension,
according to some references, this is a common extension for compressed
JPEG 2000 code stream. Also our old plug-in was listing this extension,
so let's do so now as well.
To this day, the only 2 extensions we used to list in the JasPer-based
plug-in and not in the OpenJPEG one are .jpf and .jpx (JPEG 2000 Part-2)
since OpenJPEG does not have support yet. But actually I think the old
plug-in may have simply been "lying" since JasPer website says the
library is meant to implement JPEG-2000 Part-1 standard.
So I believe we are now on par (and even better on many aspects) with
the former plug-in implementation based on libjasper.
After moving up the profile extraction, I was running
gimp_image_set_color_profile() with a non-existing image id, which was
obviously wrong. Reorder a bit the operations.
Also try to guess the color space from the profile not only with
OPJ_CLRSPC_UNSPECIFIED but also OPJ_CLRSPC_UNKNOWN images. Indeed I
encountered a case of .jp2 image with no color space in the header, but
with an embedded profile. And unlike the .j2c files I encountered
earlier, the color space was now *_UNKNOWN.
See https://github.com/uclouvain/openjpeg/issues/1103
Current OpenJPEG code only supported the base JP2 container. It now
supports also the JPEG 2000 codestream (which is usually contained
inside other formats, like the JP2 container format, but can also
sometimes be on its own).
The current magics and extension strings were also mixing all kind of
formats. This is now cleaned up a bit.
As explained in the previous commit, the color space is not always
properly declared, in particular with J2K files. If a profile is present
in such a case, try to deduct the color space from this information.
It seems that the color space is not necessarily declared for a JPEG2000
image. From tests, it looks like it especially happens with JPEG2000
codestream (.j2c or .j2k). This variant is apparently mostly designed to
be embedded (from what I read), which may explain why the color space is
not always set (I assume the embedding format would have the color space
information). Mostly a guess.
Rather than just assuming all non-gray images are RGB, do a bit more
robust check and reject unknown formats. Indeed even though I see we
took care of YUV, e-YCC and CMYK images above (and normally either
converted them to RGB or already exited with an error), I can see that
the OpenJPEG library could still return OPJ_CLRSPC_UNKNOWN or
OPJ_CLRSPC_UNSPECIFIED. Let's be thorough and not assume we got a SRGB
here.
Also add the alpha-variant tests inside their parent image type
respective test. This should not change anything by any logics, but
let's not leave anything for chance to strike us.
Finally minor coding style fixes:
- Add a space before "if|for" and parenthese.
- Remove some spaces after parentheses.
- Get rid of 2 trailing whitespaces.
- Align function call parameters, declarations, assignments…
Made plug-in support the RGB and grayscale with alpha.
Comment by Jehan: this makes the original branch work finally usable on
some JPEG 2000 images. Support of the format is not complete yet though
but at least the port to OpenJPEG is now in usable test.
Use the new gimp_get_pdb_status() to forward the error returned by
gimp_file_load(). Previous code was always returning
GIMP_PDB_EXECUTION_ERROR when the file load was failing, but this was
not granular enough. In particular when the file load is actually
interactively cancelled through Esc or the "Cancel" button, we don't
want to display an error message on screen. Therefore we forward the
actual error raised by the underlining plug-in.
While we are at it, let's just add RGBE exporting. It's just as easy.
Also rename s/file-gegl-load-rgbe/file-load-rgbe/. All formats just use
the "file-FORMAT-(load|save)" naming for their procedure, even when
implemented directly through "gegl:load|save".
It seems that various software use something different after the "#?",
and even Blender code just ended up only use these 2 characters as magic
number. See also bug 792453.
Various plug-ins exporting metadata should now follow preferences, which
would override any default. Of course these preferences can still be
overriden by saved settings (global parasite), previous run settings,
and finally through the GUI when interactive.
This is a privacy concern. Whereas importing metadata is usually a good
idea, exporting it should be a conscious action. A lot of private data
can be leaked through metadata and many people don't realize it (which
also usually means they don't need it). On the other hand, the people
who realize it are the ones who would explicitly edit the metadata and
check what they want to be exported or not.
This is only a first step. Some people may want to always export the
metadata and for these people, there should be abilities to change the
default.
After discussing with Mitch and understanding better the X bitmap/pixmap
history, I make the warning more specific to X bitmap cursors only (not
pixmap).
Also I can see our code always exports RGBA data, so I am not quite sure
if this warning even makes sense since X bitmaps are bicolor maps. On
the other hand, Mitch tells me that "these days gdk turns pixbufs into
bitmaps if the x server doesn't support rgba cursors", so maybe that can
still be of use to warn cursor designers for max compatibility.
Still that's pretty old compatibility stuff, so let's replace "may" by
"might".
Leading 0s have no special value, we use base 10 anyway. Removing
leading 0s allows to not trigger the 8-digit test, hence modify a valid
cursor size unecessarily.
Ell