adam3smith / translators

We're still seeing some reports that the new PubMed isn't working, so
trigger another update for clients that failed to pull the update for
some reason.

They seem to have gotten rid of <meta name="ncbi_article_id"> in favor
of <meta name="uid">?

https://forums.zotero.org/discussion/83218/chrome-connector-is-not-working-correctly-in-pubmed

Ingenta - Fixed RIS url and multiple detection.
Update tests & lint

The language section is rewritten. Now it should handle all situations well.

Pages are rendered with JS and are broken with our current snapshots

de-facto re-write:

* Retreinving data from theses.fr .rdf files
* Considering supervisors as contributors
* Adding multiples imports for organization/supervisors/doctors pages
* Making a distinction between "Thèses de doctorat" and "Thèse en préparation"
* Adding extra information such as laboratory, graduate schools, jury members
* Adding licence for theses.fr data
* linting

clinicaltrials.gov: Fix authors

And use field mode; fixes 
https://forums.zotero.org/discussion/83162/saving-item-from-clinicaltrials-gov/p1

Translator depends on dynamic rendering

Various adjustment to make the code work, mainly:

  - Change Element ID from N1 to N0 at multiple occurrences
  - Add DOI
  - Adjust textVariables (metadata names). Remove some obsolete ones.
  - Optimized the way author names are handled.

DOI: Find DOIs in links, and other improvements

Aosic fix

Translator is currently broken and needs to be updated, though

BibTeX: Add newline to end of exported file

Fix author names for new(ish) layout (update tests; no more abstracts, but these were always of questionable quality)

take abstract from RIS and pages(article numbers) from RIS for electronic journals

Also add a list of electronic only journals for which we use the RIS page number

Multiples, collections, bibliography; also update tests & remove invalid 
URLs

using report item type for now; 
test for multiple  fails, so not adding, but tested successfully in Firefox

…tero#2169)

and update tests (no more ISBN in metadata)
Fixes zotero#2168

```
curl -L -v \
    -H "Accept: application/vnd.datacite.datacite+json, application/vnd.crossref.unixref+xml, application/vnd.citationstyles.csl+json" \
    'https://doi.org/10.7916/d8959hr1'
```

The JSON for https://doi.org/10.7916/d8959hr1 didn't have many of the
fields expected by the translator, so I made them optional.

The new test fails because of zotero#2018.

zotero#2018 (comment)

This is a temporary fix until the DataCite JSON includes
"http://datacite.org/schema" again or we can check the Content-Type
response.

Fixes zotero#2162

added multivol support to TEI.js and fix linting