Zotero: which export format should I use?

Question

Which of Zotero's export format would you recommend regarding
- the portability with similar programs
- possibility of reading and adding new entries with a Perl script?

Answer 1

Much of this depends on what other software you will be working with. Any flexible read/write connection to Zotero should probably use the server API; there are already pretty strong client libraries in Python and PHP that you can explore, and it would be reasonable to write one in Perl.

If you just need read access, or read access in addition to write access, there is a Python library, libzotero, that's provided by the wonderful qnotero tool. It opens a read-only connection to a local Zotero installation's underlying sqlite database. If you need quick read access and searching, that library or its approach will serve you well.

Without using the server API, it's also possible to use the Firefox extension MozRepl with the MozRepl CPAN module to get programmatic access to a running local Zotero instance. This is pretty powerful, but it means that you need to send JavaScript to MozRepl. This approach is used with elisp to implement Zotero access for org-mode, zotero-plain.

If you certainly want export, the most expressive option is Bibliontology RDF, but not much out there understands it. MODS export from Zotero is also pretty solid, and it can be converted into pretty much anything else, using the superb bibutils package.

And the main place for questions like this is the mailing list zotero-dev, where you'll find just about everyone who works on programming in the broader Zotero ecosystem, so it may be worth stopping by there as well.