Saving a PDF document

[remil19]

J'ai essayé d'ajouter un pdf qui s'était ouvert dans mon navigateur via pdf.js à poche, mais celui-ci l'enregistre comme du texte brut et sans titre. Ce serait bien qu'il enregistre un titre et qu'il l'ouvre dans un onglet quand on clique dessus.

[nicosomb]

Hello @remil19!
Sorry to answer so late.

I don't know if this feature has to be implemented in wallabag (our new name).
If anyone has an idea?

[tcitworld]

In fact, with a PDF file, no informations about the title are transmitted by HTTP. You can test it with any pdf document online with http://web-sniffer.net/.
So the only information we could get would be the filename, and sometimes they're not at all speaking to the user. But maybe if Wallabag v2 allows to change the title, something can be made.
However, a detection should be made (with the Content-Type : application/pdf http header) to say to Wallabag "this isn't a regular page".

Of course, we can also use an external library to read metadata informations.

[nicosomb]

I just added the "Plugin" label to this issue.

[remil19]

Sorry to answer so late but the first line of a PDF usually begin with %PDF so i guess we could look for this string when a entry is created (and i think it could be usefull to be implemented directly in Wallabag : many long and interesting documents are published in PDF and it may not require a lot of modification : maybe just a boolean in the database or a type attribute and you also could just use this viewer: https://github.com/mozilla/pdf.js).

[mariroz]

hi, @remil19 , yes, of course, pdf should be handled when entry is imported. But not by parsing itself, but 1 step ahead: by checking document http headers. (see related issue #444 about plain text handling). I hope, that now or later we will implement this. Anyway I will try :).

[tcitworld]

Yes, detection with parsing first bytes of files isn't really easy to made, compared to http headers detection. Although, I don't know if all servers serve pdf properly.

[tcitworld]

Going for v2.x.

[nicosomb]

Assigned to Tender discussion #4.

[j0k3r]

For now, instead of storing the pdf itself, we provide a text version of it: j0k3r/graby#16

[nicosomb]

Done by @j0k3r in graby 👍

[mdimura]

I tried saving PDF-url with wallabag v2.0.1, but I get " wallabag can't retrieve contents for this article. Please report this issue to us. " error. Would be great if wallabag downloaded the original PDF and stored it locally for future reading.

[nicosomb]

Can you open a new issue for that please?

[toobluescientist]

I tried saving PDF-url with wallabag v2.0.1, but I get " wallabag can't retrieve contents for this article. Please report this issue to us. " error. Would be great if wallabag downloaded the original PDF and stored it locally for future reading.

Hey, I have the same trouble as you. For me, I still cannot save PDF from a PDF url onto Wallabag. If it saves, sometimes the text is just unreadable. How is it now for you?