"Find Available PDF" doesn't always work.

ryofurue
edited November 26, 2021
I was wondering why "Find Available PDF" doesn't work for the following paper:

doi: 10.1175/1520-0485(2000)030<1532:NSOTNA>2.0.CO;2

https://journals.ametsoc.org/view/journals/phoc/30/7/1520-0485_2000_030_1532_nsotna_2.0.co_2.xml

whereas it works for

doi: 10.1175/JPO-D-14-0171.1

https://journals.ametsoc.org/view/journals/phoc/45/6/jpo-d-14-0171.1.xml

Both are articles on the same journal and the "Download PDF" button works on the webpages of both.
  • ryofurue
    Is this something which can be addressed on the Zotero side? Or is the website broken in some way?
  • dstillman
    The latter, but maybe the former too. The page is invalid, but we'll see if we can put a workaround in place.
  • ryofurue
    @dstillman Thank you for your response! I wonder what you mean by "the page is invalid".

    If a detailed technical description of the problem is available, I could contact the publisher to ask them to fix it.
  • dstillman
    edited November 29, 2021
    https://validator.w3.org/nu/?doc=https://journals.ametsoc.org/view/journals/phoc/30/7/1520-0485_2000_030_1532_nsotna_2.0.co_2.xml

    Various problems, including a script between html and head, but the one that's breaking things for Zotero is "Bad start tag in img in noscript in head", which can cause HTML parsers to stop parsing the head and consider the rest to be the body. Zotero only looks for embedded metadata tags within head (which is where they were trying to put the tags on this page), so it doesn't find the meta tags when it parses the page and can't find the meta tag pointing to the PDF.

    (The Zotero Connector still works because browsers figure out what's going on and ignore the invalid img, but the Zotero app uses DOMParser, the standard web API for parsing HTML, and that doesn't work with the invalid structure.)
  • ryofurue
    @dstillman Thank you very much for the details! I've forwarded the information, as well as the link to this thread, to the publisher. I at least got their response that the information is forwarded to the technical department which produces their webpages.

    I hope they will fix their webpages. . . .
Sign In or Register to comment.