errors and missing fields in CSL output

[Robert Jäschke]

1. For the chapter Literatur recherchieren und verwalten (API) the title returned by CSL is "1" but should be "Literatur recherchieren und verwalten".
2. The intra hash and user name of a post are not available in separate fields but only concatenated in the field id, e.g., 9d87348e2ce9f970dfac29446b04729bjaeschke.
3. Although the output contains fields DOI and ISBN, they are always empty, cf. http://www.bibsonomy.org/bibtex/2797d40e05a48f4343d7695dac87b5870/jaeschke (API)
4. The field md5hash of the field documents is inconsistently named: the other fields use camel-case (e.g., userName, fileName, ...).
5. Furthermore, the field userName for the field documents would be redundant, when the user name would be added as separate field to the post output as suggested by 2.
6. Finally, the viewability of the post would be helpful, to decide whether it should be shown on public pages (when the posts are retrieved by the API, ALL posts are returned and it is not possible to filter by viewability status).

[Daniel Zoller]

1. already fixed in default
2. Any reason to drift from the standard CSL? Why do you need the username and hash? Can't you use the id?
3. What is the problem with an empty ISBNs/DOIs when the publication does not contain a ISBN/DOI? The checks for including ISBNs/DOIs should check if the value is not empty (https://github.com/citation-style-language/documentation/blob/master/specification.txt#L1516). The provided example contains ISBN and DOI.
4. This won't fix, if we would change this now, all clients would break.

[Daniel Zoller]

The userName attribute is not redundant, because for group exports this could be another user.

[Robert Jäschke]

1. perfect
2. To build proper links to BibSonomy (cf. http://www.kbs.uni-hannover.de/~jaeschke/test.html) or to download the documents for which one needs the intra hash and user name.
3. I have to check this. In the CSL I got they were missing. Maybe this is an error of the CSL processor I am using.
4. Too bad.
5. The documents of a post should always be owned by the post owner, right? So having the user name for the post would make the user name for the document obsolete.

[Daniel Zoller]

regarding 3): that was an error in the api; the dois and isbn should now be returned by the service

[Daniel Zoller]

except 6 all requests were implemented (release 3.6.0).

I would suggest to add a visibility filter to the api and database and not add the group field to the output.

[Daniel Zoller]

for point no. 6 see #2659

[Robert Jäschke]

until #2659 is implemented, having the attribute would still be helpful