Handle Zipped Document Bundles Properly — Funded

[freelawbot]

Issue by thinkcomp
Wednesday May 08, 2013 at 07:55 GMT
Originally opened as https://github.com/freelawproject/recap-firefox/issues/10

---

I just downloaded a bunch of documents for a case I filed (CAND). The PDFs got uploaded through RECAP without a problem, but the .zip bundles didn't make it. Specifically, for...

http://ia601700.us.archive.org/20/items/gov.uscourts.cand.265908

...I downloaded documents (2) containing 2-main.pdf and 2-1.pdf, (6) containing 6-main.pdf and 6-1.pdf, and (8) containing 8-main.pdf and 8-1.pdf.

[freelawbot]

Comment by mlissner
Tuesday Apr 07, 2015 at 21:41 GMT

---

I can say with pretty high certainty that we aren't going to support zips. There's a lot of complexity to them and I just don't expect to be able to sort it out. What we could do is something that discourages people from using zips. That might be an option, so I'll leave this open. If anybody has ideas for ways of doing something like that, I'm all ears.

[mlissner]

A simple solution here would be to inject some kind of warning about the zip files either before the user clicks the link or after they do (as a modal, say). Another user just got bit by this.

[mlissner]

Depending on the format of the zip files, the new RECAP server could probably handle this. I suggest handling this by:

1. Tweaking the extension to upload zips.

2. Track how many of these we get vs regular uploads.

3. Make a decision as to whether we want to bother with them or throw an error.

[johnhawkinson]

Track how many of these we get vs regular uploads.

This is kind of chasing our tail. "Everybody" knows that ZIP files don't work, so people who care about RECAP don't use them. There are cases where I would download ZIPs, but I don't. (Or cases where I am willing to forgo RECAP for reasons of urgency.)

The better question would be to try to bound how often a ZIP download would be useful to the user. For me, that would be the time at which I am downloading all attachments to a single document, with increasing utility with a greater number of attachments. Zip isn't very important when there are 1 or 2 attachments, but it's super-helpful when there are 30.

For instance, http://ia600805.us.archive.org/19/items/gov.uscourts.mad.191926/gov.uscourts.mad.191926.docket.html has 66 attachments to doc 1. (I can't give you the CL docket now because the CL search seems to be down.)

[mlissner]

The good news is that the zip file is blissfully simple:

Really, we could probably just start collecting these. The processing, unless I'm missing something, is a piece of cake:

• Open the zip
• Match up file names to document and attachment numbers
• The rest of the merging and saving routines we already have.

Really it's just parsing file names (easy) and opening a zip (easy, but I forget how). This is about as easy as it comes. Forget tracking.

[mlissner]

Just got another email from a user bit by this.

[mlissner]

Happy to share this feature got sponsored by a user that was annoyed by it. Looks like we'll be building it out! First steps:

• Tweak the extension to send the zips & release new version
• Tweak the RECAP API to receive the zips and store them somewhere

After that:

• Get the parser going for the zips.