Nature translator is saving articles as web pages

danhicks

I think since Nature redesigned their site in February, every article I try to save to Zotero shows up as a web page. Here's an example article: https://www.nature.com/articles/d41586-018-06008-w

The connector lists the translator as "Nature Publishing Group." But the result in Zotero is a web page rather than a journal article. PDFs aren't attached, and the progress overlay doesn't even indicate that the connector is attempting to retrieve the PDF. Skimming the debug log, it looks like the connector tries to use the Nature Publishing Group translator, then for some reason switches over to Embedded Metadata.

Debug ID D1748822781
Chrome 68.0.3440.106 on Mac OS 10.13.6

adamsmith

I'm only seeing this for commentary. Research articles and letters work, e.g.
https://www.nature.com/articles/s41586-018-0440-4
https://www.nature.com/articles/s41586-018-0439-x

We can take a look at commentary, but I'm not seeing any systematic metadata on the page or downloadable from it, anything we do would be pretty fragile.

danhicks

I'll admit I've never taken a close look at how translators work, so maybe this is irrelevant. But, in the HTML source, all three have what looks like the same set of `meta` tags for, e.g., the PRISM metadata.

If those `meta` tags don't work, what if the first alternative were to extract the DOI and use the DOI translator ? If I try to add by commentary using the DOI (viz, 10.1038/d41586-018-06008-w), it seems to work as expected.