wallabag can't handle plain text

[lightonflux]

wallabag can't handle plain text

E.g. http://ibiblio.org/pub/linux/docs/ldpResearch/ldp-historic/LinuxNews.03A

[tcitworld]

This may be a little more difficult than it seems. We have to check the text/plain http header before starting to get all the content.

[mariroz]

this is next challenge, related to db structure expansion. Adding of "mime_type" field will allow wallabag to handle also other formats, for example we will be able to store pdf. This is good direction of growth, on my opinion. I hope, will be able to to test @nicosomb's install script next week and will try to think about accurate and secure db changes.

[nicosomb]

@mariroz you want to add a new column in entries table? To store mimetype? Good idea. But, I'm "chilly" to database modification ... wallabag is not well structured, so upgrades are quite sensitive.

[nicosomb]

@mariroz do you think this feature is possible for wallabag 1.x?

[mariroz]

I hope, that yes :)

[tcitworld]

If we modify database structure, could do it once and for all, like adding an is_administrator column in users, a comments column in entries, etc ?

[mariroz]

@tcitworld, yes, of course we will need more columns.
This is complex task: should be done together with wallabag upgrade (and maybe install) procedure modification.

[tcitworld]

I'll open a new issue specifically for this.

[tcitworld]

Going for v2.0.

[nicosomb]

I just tried the URL in the 1st post and this one for example (http://www.oracle.com/technetwork/java/readme-2-149793.txt) with wallabag v2. And it works.

Is this issue still mandatory?

[tcitworld]

And I heard someone saved pdfs. So I guess we're good, with MIME Type now stored.

[nicosomb]

mimetype is not yet stored.
About PDF, it's not the document content.

[tcitworld]

No, but the thing is : the url is saved, with the title.

[nicosomb]

yes 👍

about mimetype, I don't know if it's easy. Because you can't get the URL "mimetype", it doesn't exist ...

[tcitworld]

Of course you can, it's the Content-Type http header. ;-)

EDIT : https://stackoverflow.com/questions/2610713/get-mime-type-of-external-file-using-curl-and-php

[nicosomb]

oh yes, you're right.

[j0k3r]

Fixed in graby.