Add PDF import support for books (Issue #93)

[dgc08]

Implement issue #93

I took the PR for the epub support as template and chnaged it accordingly.

I used PyPDF2 as library, the license should be fine: https://pypi.org/project/PyPDF2/

[jzohrab]

Hi @dgc08 -- this failed one of the acceptance tests, can you take a look?

[jzohrab]

Failed acceptance test.

[dgc08]

The acceptance test should work now. The sample PDF file contained an extra page number, which also got imported. Didn't run the test beforehand, sorry.

[jzohrab]

No problem that’s what the tests are there for. Thanks for adding the test it’s a big help for stability. El El lun, ene. 8, 2024 a la(s) 8:13 p. m., Sinthoras39 < ***@***.***> escribió:

…

The acceptance test should work now. The sample PDF file contained an extra page number, which also got imported. Didn't run the test beforehand, sorry. — Reply to this email directly, view it on GitHub <#119 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAMPWDO4BF26GIDGZPS3MHDYNPWIZAVCNFSM6AAAAABBQL5TSKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQOBQHE4DIMZVHA> . You are receiving this because you commented.Message ID: ***@***.***>

[dgc08]

kinda screwed up my fork, wait a moment before merging

[dgc08]

ok its fine now

[jzohrab]

One last change requested: what needs to be put into the pyproject.toml file (in project root)?

(I don't have CI checking this, couldn't sort out how to do it well.)

[dgc08]

I hope that's it. Thank you for your patience with me, Lute is the first Open Source project / larger project in general that i contribute to, pytest and that stuff is still new to me

[jzohrab]

It’s a super contribution so thank you. I’ll try it out soon with a bigger pdf to see how it works, and will review the code again. For a first PR it’s great! 👍 El El lun, ene. 8, 2024 a la(s) 11:03 p. m., Sinthoras39 < ***@***.***> escribió:

…

I hope that's it. Thank you for your patience with me, Lute is the first Open Source project / larger project in general that i contribute to, pytest and that stuff is still new to me — Reply to this email directly, view it on GitHub <#119 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAMPWDP7C55KRLXBTIIPHHLYNQKGXAVCNFSM6AAAAABBQL5TSKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQOBRGM3DEOJTG4> . You are receiving this because you commented.Message ID: ***@***.***>

[jzohrab]

Hi @dgc08 -- I did an import of a large file and found a few places where the import adds some spaces, eg

vs the original pdf:

This happens fairly frequently. It's almost certain that it's due to the PDF parser library ... should check.

[jzohrab]

Yes, it's the library, there's a good write up by the author(s) in these links:

• https://stackoverflow.com/questions/76546531/why-does-pypdf-stuff-text-with-extra-spaces-when-extracting-text
• Random whitespaces are inserted when using page.extract_text() py-pdf/pypdf#1507

Per the authors:

Getting whitespaces right is notoriously hard. @pubpub-zz is the expert in that topic; I'll leave it to him to decide if we should leave this issue open. The issue is that PDF does not (necessarily) represent the words as words internally. In the worst case, it just gives the absolute position of each character in the document.

So, I think this should be fine as it is, but we should mention somewhere that PDF imports are extremely tricky -- I'll draft that before merging this PR, as users should be aware of the limitations.

[jzohrab]

Added a flash: