XPather incompatible with Firefox 18 - HWZT++

deanmcurran · January 30, 2013

Hi,

I am new to writing translators, and wanted to create one to extract the citations from this page: http://cns.asu.edu/cns-library/type/#report

I started with HWZT++ but am having problems building an XPath. XPather, which integrates with the DOM inspector recommended by HWZT++, is incompatible with Firefox 18. Bridging is impossible, and I've read that downgrading Firefox could jeopardize my data.

Any add-on suggestions that would build an XPath, and might work with DOM inspector? Or my my approach entirely wrong?

Maybe with such a simple site, I'd be better off working with existing translators...

Thanks for your help

Gracile · January 30, 2013

HWZT++ is rather old now (but not outdated AFAIK).
The translator framework is enough in many cases: http://www.zotero.org/support/dev/translators/coding#translator_framework

You can use the XCpath bookmarklet: http://dl.dropbox.com/u/848981/it/xp/xp.html
or Firebug with Firepath.

Gracile · January 30, 2013

I didn't click on your link before... I don't think you've chosen an easy webpage for your first translator; even with some experience, this might be hard to scrape the metadata there but others will know better than me.
(and the translator framework is of no help)

adamsmith · January 30, 2013

I'd also recommend the bookmarklet as an xpath builder, for looking at xpaths, Firefox's "Inspect Element" function is now just as useful as Firebug, though (edit: ah, didn't see the reference to Firepath - yes, that would do more than FF's built in function).

I'd go further than Gracile and say that it's basically impossible to write a translator for the site you link to, regardless of your skills as a programmer.
Parsing formatted citation data is very complex and all but impossible to do reliably. You can take a look at this: http://www.zotero.org/support/kb/importing_formatted_bibliographies for tools specifically designed to do that.

Beyond that, you could recommend to the folks maintaining the site that they embed citation info, e.g. in COinS that Zotero could read automatically:
http://www.zotero.org/support/dev/exposing_metadata#using_an_open_standard_for_exposing_metadata