Introduce pdf support (#7318)

[yabramuvdi]

First implementation of the Pdf feature to support pdfs (#7318) . Using pdfplumber as the default library to work with pdfs.
@lhoestq and @AndreaFrancis

[AndreaFrancis]

Great contribution, @yabramuvdi! 🎉
I added a couple of minor comments. It would also be great to add pdfplumber to the dependencies in the setup.py file and perhaps include a few tests for the new features.
Thanks in advance! This will help us unblock some other features in the datasets-viewer project as well.

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

[lhoestq]

Awesome ! the implementation looks all good to me :)

could you add a test before we merge ? something as simple as test_video_feature_encode_example and test_dataset_with_video_feature from test_video.py maybe

[yabramuvdi]

Hi @AndreaFrancis and @lhoestq ! Thanks for looking at the code and for all the changes and suggestions. I have worked on all your suggestions. I need to work a bit more on the tests, but I created a first version and uploaded a simple pdf to use for testing. The pdf has a couple of pages some of which have images, figures, and tables (this is exactly the pdfs for which I think keeping the pdf format is very interesting because they have multiple types of content). Will try to finish the tests as soon as possible.