I want to report it to vips, can you suggest any method I can easily use to determine if a given file is a full jp2 or raw jpeg2000, to make my report as accurate as possible?
I don't think…
It turns out, that this is a file without the jp2 metadata around it but just a raw jpeg2000 bitstream. I added rudimentary support for those in 09064e8e70a0549147db172eb5b122e50479fd36. Tell me…
Hrm... I had to touch the code around the "no width in jp2 header" error message when I implemented support for your images because not only did your image have transparency, it also was 16 bits…
You are not giving me trouble -- quite the opposite! Thanks to you, img2pdf now supports JPEG2000 with transparency. 😄
If your organization is fine with that, maybe I can add it to a new…
@jrochkind you asked for an estimation for a release including this fix. It turns out that the unit test I wrote for JPEG2000 with alpha channel fails on Fedora. So I cannot release until I figure…
Okay, e05580a49a1d2cc2822dcd0d13b8386da6293e2b fixes the baseType problem. I can also confirm that installing icc-profiles-openicc makes the tests succeed that failed before.
Unfortunately,…
e05580a49a
src/img2pdf_test.py: IM7 dropped 'baseType' in json output, so use 'type' instead which works for both IM6 and IM7
Thank you, I'm able to reproduce the problem with rawhide inside podman:
podman run -it --rm quay.io/fedora/fedora:rawhide
[...]
# dnf install ImageMagick
# curl -O https://upload.wikime…Fedora has /usr/share/color/icc/OpenICC/sRGB.icc and /usr/share/color/icc/colord/sRGB.icc via icc-profiles-openicc and colord. However, I'll have to explicitly declare one of them as build…
Then you are in luck: I'm also maintaining the img2pdf package in Debian so once I release a new version here I'll also upload that version to Debian unstable. :)
Indeed I forgot about that syntax. It looks funny but it's the same mechanism as when writing
def myfun(foo, bar, *args):
[...]
And *args will catch any additional…
@mara0004 things are a bit more complicated concerning the enum comparison. There are good reasons to use the is operator instead of ==. But above you also suggested to use:
imgformat…I just pushed a commit with the patch I attached in an earlier message.
This does not yet fix the baseType bug (I need to know what you get when you run convert truecolor.jpg json: to fix…