XML import translators

[adamsmith]

I've mostly finished writing marcXML parsing as part of a library translator. It's written using DOMparser & and feeds the data to the marc translator.
I'm thinking that it may very well be worthwhile putting that code into a separate import translator, but I'm not user how to proceed. The documentation suggest using E4X for xml translators
http://www.zotero.org/support/dev/translators/coding#import_translators
which would  be easy, but doesn't work outside of FF. Should I just treat the file this as text with

var line;
while((line = Zotero.read()) !== false)) {

and then use 

var parser = new DOMParser();
var doc = parser.parseFromString(text, "text/xml");

Or is there a more elegant solution?

[Avram Lyon]

On Mon, Mar 12, 2012 at 12:49 AM, adamsmith <bst...@gmx.de> wrote:
> Or is there a more elegant solution?

Without code diving, I can't say whether there's a more elegant solution.

If Simon can chip in here, it'd be good to get a full listing of what
input formats can be specified, so we can update the docs accordingly.

Avram

[adamsmith]

For what it's worth, using Zotero.read and then parseFromString works smoothly, so unless there are compelling reasons against it I'll use that for now.

[Simon Kornblith]

We currently support:

rdf/xml (used in Bibliontology RDF.js, Zotero RDF.js, and RDF.js)
rdf/n3 (unused, same as rdf/xml but with N3 output)
xml/e4x (used in MODS and TEI)

In theory there's an xml/dom mode for which Zotero.getXML() returns a
DOM document, but it's quite likely that it's been broken since
Firefox 5, so DOMParser is probably the best option at the moment.

Simon