Allow bulk conversion to safe PDFs

[micahflee]

A few different people have requested this feature: the ability to convert documents to safe PDFs in bulk.

This shouldn't be too difficult to implement, but I think the biggest issue is how the user interface should work. One idea is to allow the user to browse for a folder, and then try to automatically convert all documents it can in that folder to save versions (appending -safe.pdf to the end). It probably makes sense to convert them one at a time, and also display a report of all unsupported files in the folder that couldn't get converted.

What happens if there's a file called test.pdf and another called test-safe.pdf in the same folder though? What would the safe version of test.pdf get named? It could maybe detect this edge case, and tell you that you must not have any files that end in -safe.pdf in the folder to convert.

[haplo]

It probably makes sense to convert them one at a time, and also display a report of all unsupported files in the folder that couldn't get converted.

Running multiple conversions in parallel would improve running times and make use of multiple CPU cores. Just run a process pool bound to the number of CPUs (ideally would be a configurable setting), queue all PDF to be converted (using multiprocessing.Queue and have the worker processes pull from the queue and do the conversion. I can provide an implementation for this.

What happens if there's a file called test.pdf and another called test-safe.pdf in the same folder though? What would the safe version of test.pdf get named? It could maybe detect this edge case, and tell you that you must not have any files that end in -safe.pdf in the folder to convert.

I think a good UX would be to identify all PDF files that already have a -safe version and skip them, telling the user that they seem to be already converted. User can then manually rename them if they want Dangerzone to work on them.

[pettitjr]

I would also like to see this feature!

[JesseKrembsNYT]

What James said..

[tzmnyt]

Bulk conversion of PDFs is very much needed.

[RLburrito]

Bulk conversion is desperately needed here as well.

[DirtyNoob]

I agree to the consensus here, this is much needed!

[anarcat]

as part of #110, i implemented a webdav processor to do batch processing. the idea is that you dump your files in a webdav folder (e.g. on nextcloud), share that folder with the dangerzone bot, which pulls the files, processes them in docker, and pushes the sanitized files back.

see https://gitlab.torproject.org/tpo/tpa/dangerzone-webdav-processor/ for details.

[ninavizz]

Hey hey! I've got some hours to spare to get a nice UX together for this, and to potentially improve the single-file experience. Can do & share next week.

[deeplow]

An extra thing to consider is that a user may want to add extra files while some are already processing.

[eloquence]

Tentatively adding to 0.4.0 milestone; initially, this may only include backend changes to support bulk conversion + CLI support.