BUG: Avoid extracting inline images twice and dropping other operators #3002
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## main #3002 +/- ##
=======================================
Coverage 96.36% 96.36%
=======================================
Files 52 52
Lines 8751 8752 +1
Branches 1593 1593
=======================================
+ Hits 8433 8434 +1
Misses 190 190
Partials 128 128 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
pubpub-zz
approved these changes
Dec 18, 2024
stefan6419846
added a commit
that referenced
this pull request
Jan 26, 2025
## What's new ### Deprecations (DEP) - Deprecate with replacement CCITParameters (#3019) by @j-t-1 - Correct deprecation of interiour_color (#2947) by @j-t-1 ### New Features (ENH) - Support alternative (U)F names for embedded file retrieval (#3072) by @stefan6419846 - Adding support for reading .metadata.keywords (#2939) by @Lucas-C ### Bug Fixes (BUG) - Handle further Tf operators in text extraction layout mode (#3073) by @blushingpenguin - Ensure `add_metadata` can deal with `_info = None` (#3040) by @xmo-odoo - Handle IndirectObject in CCITTFaxDecode filter (#2965) by @stefan6419846 - Handle chained colorspace for inline images when no filter is set (#3008) by @stefan6419846 - Avoid extracting inline images twice and dropping other operators (#3002) by @stefan6419846 - Fixed reference of value with `str.__new__` in TextStringObject (#2952) by @thomas-forte - Handle indirect objects in font width calculations (#2967) by @nsw42 - Title sometimes is bytes and not str (#2930) by @reformy - Fix undefined variable for text extraction (regression) (#2934) by @stefan6419846 - Don't close stream passed to PdfWriter.write() (#2909) by @alexaryn ### Robustness (ROB) - Handle zero height fonts when extracting text (#3075) by @blushingpenguin - Deal with content streams not containing streams (#3005) by @stefan6419846 - Gracefully handle some text operators when the operands are missing (#3006) by @stefan6419846 - Fall back to non-Adobe Ascii85 format for missing end markers (#3007) by @stefan6419846 - Ignore odd-length strings when processing cmap lines (#3009) by @stefan6419846 - Skip annotation destination being NullObject in PdfWriter (#2964) by @stefan6419846 - Skip destination page being None in PdfWriter (#2963) by @dxsooo - Fix infinite loop case when reading null objects within an Array by @jakep-allenai - Fixing infinite loop in ArrayObject read_from_stream (#2928) by @jakep-allenai ### Documentation (DOC) - Add note about default line colors (#3014) by @stefan6419846 ### Developer Experience (DEV) - Remove ignoring Ruff rule PGH004 (#3071) by @j-t-1 - Tidy ignore array in tool.ruff.lint (#3069) by @j-t-1 - Move Windows CI to Python 3.13 (#3003) by @stefan6419846 - Move to Ubuntu 22.04 (#3004) by @stefan6419846 ### Maintenance (MAINT) - Fix formatting of warning message and include exception message (#3076) by @stefan6419846 - Narrow return type for `ContentStream.operations` (#2941) by @kmurphy4 ### Testing (TST) - Fix image similarity for upcoming Ubuntu 24.04 (#3039) by @stefan6419846 - Replace broken Apache Tika Corpora urls (#3041) by @stefan6419846 ### Code Style (STY) - Add form feed to WHITESPACES (#3054) by @j-t-1 - Lots of small internal changes by @j-t-1 [Full Changelog](5.1.0...5.2.0) Signed-off-by: stefan6419846 <96178532+stefan6419846@users.noreply.github.com>
stefan6419846
added a commit
to stefan6419846/pypdf
that referenced
this pull request
Feb 12, 2025
When fixing the inline image extraction in py-pdf#3002, the fallback behavior would break accidentally by leaving the stream before the EI marker. We never saw this due to the tests somehow being not strict enough, thus leading to the error > b'EI' operator met whereas not expected, please share usecase with pypdf dev team
stefan6419846
added a commit
that referenced
this pull request
Feb 12, 2025
When fixing the inline image extraction in #3002, the fallback behavior would break accidentally by leaving the stream before the EI marker. We never saw this due to the tests somehow being not strict enough, thus leading to the error > b'EI' operator met whereas not expected, please share usecase with pypdf dev team
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR fixes a regression introduced in version 4.3.0 which would always try to extract inline images twice due to setting the wrong input stream position. The corresponding behavior is not directly obvious and only becomes visible with some PDF files as the one used in the test, where the faulty change would silently drop a
Qoperator and thus basically destroy the remaining layout of the PDF file.During analysis, I have added some basic comments and fixed a faulty error message I stumbled upon as well.
Fixes #2927.