Content-Disposition filename ignored somehow

[schmitch]

Hello, currently I render my Browser documents with PDFJS however I have patched the viewer.js to support my Authorization header i just added httpHeaders to paramters before PDFJS.getdocument so nothing special, however PDFJS ignore's my Content-Disposition Header which has a filename=
it always gives me document.pdf

I tested against a CORS version and non cors currently this is my content-disposition:

 Content-Disposition:inline; filename="Angebot_412356.pdf"

Is there a setting that it should use this filename or do I miss something?

Currently it isn't related to #2407 since that will work on save + s.
However on my side either Firefox/Chrome nobody gets the correct content-disposition i tried:

 Content-Disposition: inline; filename=name.pdf
 Content-Disposition: attachment; filename=name.pdf
 Content-Disposition: inline; filename="name.pdf"
 Content-Disposition: attachment; filename="name.pdf"

Also that's the Code I added in viewer.js starting after Line 6373 and the PDFJS line is already there:

 var token = window.localStorage.getItem('token');
 var typ = window.localStorage.getItem('type');
 parameters.httpHeaders = {'Authorization': typ + ' ' + token};
 PDFJS.getDocument(parameters, pdfDataRangeTransport, passwordNeeded,

btw. I could also set the filename by myself, I think of the possibility that I know the filename already.

[Rob--W]

The Content-Disposition header is currently not parsed.

Patches are welcome! If you're interested:

• Get the header (search for onHeadersReceived in the src/core directory to find an entry point), and parse it. (note that this parsing is not trivial, e.g. account for special encodings, see other open-source libraries for inspiration).
• Add an API or use an existing API to communicate the Content-Disposition header from the worker to the main thread.
• Search for uses of getPDFFileNameFromURL in files in web/, e.g. in web/viewer.js and update it to take the filename in Content-Disposition into account.

[schmitch]

@Rob--W oh I thought it is cause of some issue's. However if' just appended a &filename="name.pdf" to my url I think that parsing it wouldn't be too trivial for my use case.

[jumbosushi]

Does this still need work? I would love to make this my first contribution to oss

[Rob--W]

@jumbosushi This still needs work indeed. A few comments back I have described the steps that you can follow to implement this feature. If there's anything unclear, feel free to ask.

[corblanc]

@Rob--W Would like to give this a try if it's still open.

[Rob--W]

@corblanc This is still open, thanks for giving it a try!

The invitation to ask questions is also still open.

[corblanc]

@Rob--W Great! Mind if I shoot you an email? Bit new to all of this.

[Rob--W]

@corblanc Go ahead. You can also join us at #pdfjs on irc.mozilla.org if you want to have a conversation.

(edit: irc.mozilla.org, not Freenode)

[david0603]

Any updates on this? Thanks!

[Koolstr]

This is still an issue, for me at least. Any updates? It's been over a year.

[Maiestas28]

Hey guys, im new to OSS and this would be my first contribution. The tag looks open. How can i contribute and any guidance is very much needed and much appreciated. Thank you.

[Rob--W]

See #6396 (comment)

I have recently written a simple yet complete Content-Disposition parser, its functionality can be reused: https://github.com/Rob--W/open-in-browser/blob/master/extension/content-disposition.js

[timvandermeij]

Note that there is already a pull request for this issue above (and it's almost finished), hence I would suggest working on another issue instead.

[Maiestas28]

Got it. Thank you.

…

On Sat, Jan 6, 2018 at 8:26 AM Tim van der Meij ***@***.***> wrote: Note that there is already a pull request for this issue above (and it's almost finished), hence I would suggest working on another issue instead. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#6396 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AKLkdZli8v_UDNN3Tkj50RGt4bna_zwgks5tH3R9gaJpZM4F0KLh> .

[timvandermeij]

Fixed in #9379.

[ddemoss222]

Has there been regression on this? I am trying to open a base64 pdf from an tag. I set the "download" attribute to the filename which correctly names the PDF on chrome, but with the pdfjs opener in firefox it shows only "document.pdf". If I set the:

 Content-Disposition: attachment; filename=name.pdf
 Content-Disposition: attachment; filename="name.pdf"

The above too is ignored by pdf.js and I just see:

Warning: getPdfFilenameFromUrl: ignore "data:"-URL for performance reasons.

Why does pdf.js still ignore the Content-Disposition and download attribute in the a tag. Is this regression? Are there any known workarounds for JSON base64 responses?

[Snuffleupagus]

@ddemoss222 For future reference: Note that it's always recommended to open a new issue, rather than commenting on one that was closed many years ago, since it's often very easy for comments on old issues to be overlooked/ignored.

Without a runnable test-case it's difficult (or rather impossible) to know for sure, but this is likely fixed by the PR referenced above.