[openbiblio-dev] First cut of AsyncUpload branch

Fri Feb 10 10:06:00 UTC 2012

On 10 February 2012 10:57, Peter Murray-Rust <pm286 at cam.ac.uk> wrote:
> I agree. I am not a Pythonista but I can feel the technology gathering
> critical mass. We do the same sort of thing in out chemistry tools - import
> calls parse. But you have to expect poeple to submit all sorts of
> mislabelled stuff and non-standard stuff so the error trapping and reporting
> has to be clear.

The plan is to capture a list of errors encountered along the steps in
the pipeline and store it with the ticket, so the submitter can always
view what went wrong, and submit bug reports to the parser owner, or
website.

I think supporting non-Python parsers is actually quite important.
There is a large amount of Java tools out there that we could leverage
for example. If all we say is: make a self-contained executable
jar-file that can spit out BibJSON, and you can parse anything you
want. (or Ruby/PHP/shellscripts/XSLT pipelines etc. )

Etienne