Read MARC records

RonBurk

When a web page contains a MARC record, easily identifiable by being inside a <pre> element, with no intervening tags of any kind, I would expect the "Create New Item from Current Page" button to use that MARC record. It does not.

erazlogo

I would prefer instead that in such cases Zotero read the type of the MARC record and displayed the appropriate icon in the location bar, instead of using "create new item feature." In some cases it may be necessary to save and cite the exact page appearing in the browser and the snapshot feature provides a way to do that.

RonBurk

Yeah, I framed that as a solution instead of a problem in order to make it fit the category "Feature Request" :-). In fact, I'm not partial to any solution, but the problem that there's a big ol' MARC record on the screen with no real parsing problems associated and no way to get Zotero to take advantage of it is really irritating.

OTOH, I guess the fact that Zotero cannot know from just the URL that there is a MARC record on the screen (except via the never-ending whack-a-mole game of adding yet another translator/URL match) does impose some constraints on the solution.

So I'll frame the problem like this: it should *always* be possible to punch a button to make Zotero use a plain-text MARC record that is displayed on the current web page. Simple things should be simple.

dstillman

https://www.zotero.org/trac/ticket/635

As I say on the ticket, ideal would probably be some sort of meta translator that runs on every page and does a regex for supported import formats and passes the content off to the appropriate format translator(s).

Since such a translator would have to run a complex regex on every page, this would need to be profiled to make sure there wasn't a performance hit.