Ignore encryption

[unixnut]

Type of pull request

• New feature (involves code and configuration changes)
• Documentation update

About

In some cases PDF files may be internally marked as encrypted even though the content is not encrypted and can be read.

This MR provides a config option to inform the PDF parser to ignore the encryption and attempt to read the PDF anyway.

This therefore provides a work around for the following issues:

• Encrypted but not really _encrypted_ PDF isn't parsed and detected as "secured pdf" #320
• Getting false "Secured pdf file are currently not supported." #488

Thanks to @DivineOmega for making the original pull request.

(PHP-CS-Fixer has not been run.)

[unixnut]

PHP-CS-Fixer has been run successfully.

All tests passing including new ones in tests/PHPUnit/Integration/EncryptionTest.php

[unixnut]

@k00ni Looks like there were pre-existing PHP-CS-Fixer violations in the code; I ignored those because I assumed they were OK. (And just fixed ones in files I added/changed.) Might have come from PR 632 commits.

Any chance you could triage these, please?

[k00ni]

@unixnut thank you for this pull request. I am very busy these days, but I will try to get back to you until next week. Only skimmed your code, but it looks great so far.

[unixnut]

More re. PHP-CS-Fixer:

• src/Smalot/PdfParser/PDFObject.php is showing up in the list from GitHub actions, but not locally for me
• src/Smalot/PdfParser/Config.php was my bad, fixed
• src/Smalot/PdfParser/Parser.php is showing up locally for me (blank_line_before_statement, line 80 in parseFile()) but not in the list from GitHub actions
• src/Smalot/PdfParser/RawData/RawDataParser.php (statement_indentation, line 405) appears to be spurious, i.e. the correction is bad
• src/Smalot/PdfParser/Document.php (statement_indentation, line 258): comment line should be moved into else clause

Except the one I fixed, I believe all of these are out of scope for the PR.

[k00ni]

CS related stuff doesn't really matter, because I will clean/correct it myself locally, if needed. The CI stuff is important and there are no problems in your code. Thank you for investing extra time on that 👍

[unixnut]

Hi, @k00ni Sorry to bug you, but it looks like the PR is ready to approve. Any ideas when you will have merged it and released a new version? This is so I can give an update to my client.

[k00ni]

For some reason my Git was not able to push my local changes back into your branch, even though you allowed changes by maintainers. I had to make the changes by hand, therefore the messy history.

Summary:

• Adapted a few texts, only minor changes.
• Moved the tests into ParserTest, because a separate EncryptionTest class makes no sense at this point. Our tests need better organisation, but for now that the most practical solution I can think of.

Your pull request is not based on our latest master branch. I tried it locally and everything worked fine. Please check that next time.

I will merge it later this day and can prepare a new version. But it will be a release candidate, because we are in the middle of a big transition and watching how things are going.

[k00ni]

@unixnut: https://github.com/smalot/pdfparser/releases/tag/v2.8.0-RC2

[UnnitMetaliya]

@k00ni any idea why this was deprecated later on? Do we have a permanent fix? If yes, what is that?

[k00ni]

It was marked deprecated because our current implementation is incomplete/not working. In case of a false positive one can override the check setting the appropriate config value (we had a few issues with those already). In the future there might be someone who implements this part and then this config value can be removed.

FYI: https://github.com/smalot/pdfparser/blob/master/doc/Usage.md#pdf-encryption