BUG: Cascaded filters in image objects #1913
Conversation
Codecov ReportPatch coverage:
Additional details and impacted files@@ Coverage Diff @@
## main #1913 +/- ##
==========================================
- Coverage 92.85% 92.83% -0.03%
==========================================
Files 34 34
Lines 7056 7058 +2
Branches 1389 1389
==========================================
Hits 6552 6552
- Misses 358 359 +1
- Partials 146 147 +1
☔ View full report in Codecov by Sentry. |
|
Do not have test data for the moment to improve coverage |
pypdf/filters.py
Outdated
| except KeyError: # pragma: no cover | ||
| filters = x_object_obj.get(SA.FILTER, [None]) | ||
| lfilters = filters[-1] if isinstance(filters, list) else filters | ||
| if lfilters == FT.FLATE_DECODE: |
There was a problem hiding this comment.
I think the code might be more readable if we had some extra functions. Something like this:
if lfilters == FT.FLATE_DECODE:
img, image_format, extension = _handle_flate(size, data, mode, color_space)
elif lfilters in (FT.LZW_DECODE, FT.ASCII_85_DECODE, FT.CCITT_FAX_DECODE):
# Code block: Maybe leave that as-is
elif lfilters == FT.JPX_DECODE:
img, image_format, extension = _handle_flate(size, data, mode, color_space)
elif lfilters == FT.CCITT_FAX_DECODE:
# Code block: Maybe leave that also as is
elif lfilters is None:
# Code block: Maybe leave that also as is
MartinThoma
added
the
soon
|
I will have another look at it tomorrow (I want to try the extraction + inspect the files manually). But I'm very certain that I'm taking this into pypdf tomorrow + realease it :-) |
|
I can confirm that the extraction to jp2 worked. I needed to follow https://askubuntu.com/a/1151781 to convert it to something I can view though xD
imagemagick-ASCII85Decode / imagemagick-images.pdf / imagemagick-lzw.pdfExtracting
However, I cannot find a version of pypdf where it worked |
MartinThoma
changed the title
closes #1912
the issue is dealing with some images being encoded with 2 filters being cascaded