Feature: BibTeX support

[GoogleCodeExporter]

I suggest that some pages, or delimited parts of pages, could be written in
BibTeX. Gitit could then run BibTeX over the references in order to convert
them to HTML.

This would have a couple of benefits:

* A wide variety of bibliography management software (Zotero, CiteULike,
etc.) can export to BibTeX.

* The display style of academic references could be customised and made
uniform (per reference type) across the wiki, using existing BibTeX styles
and BibTeX style generation software.

Original issue reported on code.google.com by gree...@gmail.com on 5 Feb 2009 at 10:02

[GoogleCodeExporter]

The best way to do this would be to use the experimental pandoc support for 
Andrea
Rossato's hs-citeproc.  This would allow citations to be written in a 
markdownish
format and exported not just to HTML but to all of pandoc's output formats. 
hs-citeproc also has bindings to bibutils, which allows it to read 
bibliographies in
many different formats. For details, see http://code.haskell.org/citeproc-hs/.

Original comment by fiddloso...@gmail.com on 5 Feb 2009 at 7:14

• Added labels: Type-Enhancement
• Removed labels: Type-Defect

[GoogleCodeExporter]

So, John, what would adding support for hs-citeproc mean from the standpoint of
gitit? I presume you'd need to make a few tweaks to enable this?

Original comment by bdar...@gmail.com on 12 Feb 2009 at 3:09

[GoogleCodeExporter]

The main bit of work is figuring out how to specify a bibliography file.  If 
the file
is xml or bibtex, it could just live in the repository as an uploaded file 
(though
this would make it visible in the wiki - perhaps that's not always desirable). 
But
there would still need to be some convention for specifying which bibliography 
to use
for a given page.  Ideas welcome.

Original comment by fiddloso...@gmail.com on 12 Feb 2009 at 3:23

[GoogleCodeExporter]

I guess the best balance of ease of implementation and use and generality is to 
have
a single global bibliography file, which could optionally be marked private? I 
could
imagine some useful UI stuff could later be built on top of that using 
Javascript.

BTW, might be nice if Andrea can at some point add optional RDFa support, so 
that the
HTML content can be tagged for easy extraction.

Original comment by bdar...@gmail.com on 12 Feb 2009 at 4:49