Nature Communications article number

For online only journals like Nature Communications, the trends seems to be to use a single article number for "Pages".This works for PLOS journals but right now Nature Communications loads 1-x in "Pages" where x is the number of PDFs pages in the article.It seems this is coming from the metadata and causes a failure in the test for Nature Communications in "Nature Publishing Group.js".

I don't have much experience fixing translators, but it looks like the only places the article number is noted in the HTML is in an element called "article-number" in the



Any thoughts here how to fix this?

  • This information can be stored in Extra like this:
    Number: 45

    @adamsmithwould need to comment on the feasibility of extracting this from the Nature page as it currently is presented.
  • It could be yes, but right now its stored in "Pages" for other online only journals.This also is more convenient for creating correct citations since the article number is what should be replace "Pages" in most citation styles.
  • I agree generally.Some styles like APA ask for different formatting for pages versus article numbers.I frankly think that it is stupid, but it is something to bear in mind.
  • yeah, I'm pretty sure this is _possible_ but given what Jeremy points out, I'm not sure it's desirable
  • frankly, its bizarre that NPG doesn't have this "article number" in the metadata!its the number they recommend using for citing Nat Comm articles.
  • @adamsmithIt seems like we should at the least store the number in the pages field, rather than the number of pages in the PDF.
  • edited April 13, 2020
    Many publishers that use an article number instead of a journal page range nonetheless include 1-n as a page range with the article metadata.The article webpage clearly includes an article number but it isn't in the metadata (DOI, header tags, or RIS)..188BET靠谱吗I'd love for someone to identify an easier way to get the article number into the pages field of Zotero.

    A similar issue is with article numbers that are provided in publisher metadata.Some publishers (BMJ Group, and some Springer and Wiley journals) preface their article number with an "e".Others, I'm primarily of Elsevier and BMC as I write this but there are many others) that only provide a number.188BET靠谱吗Would it be a reasonable request to offer a Zotero right-click option for the program to insert an "e" before the article number?That is a far less tedious way to accomplish this than carefully placing the cursor at the beginning of the article number and entering the "e" manually.

    edited to fix typos
  • For Nature Communications, where I looked the article number is in the SP tag of the RIS, which seems reasonable.It's not in a standard metatag on the page, but it is in a span with a data-test="article-number" attribute, which would also be parseable, albeit as a custom solution.

    Their metatags are otherwise excellent -- I'm not sure if the 1-x page range actually has some benefit for google scholar or otherwise.
  • @adamsmithI suspect that eventually, if not sooner, the 1-n page range will be replaced with a number of pages metadata item much like with books and technical reports.I know that if I'm collecting potential sources for a review, I'm likely to want to read a 5 to 10 page journal article than one that is a mere 2 pages.But that isn't useful as a locator.

    Right now, for the purposes of citation (and in my case, perfecting a database record that someone can use for citing a source) the article/item number is more important as a locator than an indication of the number of pages an article encompasses.The number of pages doesn't help an information seeker to locate the article.

    Maybe, the English "e" isn't the best character but several major publishers are already using it.
  • 188BET靠谱吗I understand that 1-9 isn't useful for citations and I completely agree it's not useful in Zotero's page field.

    But the principal goal of the google highwire metatags isn't accurate referencing but proper indexing in google scholar and there may be good reasons for Nature to display the data as they do for that purpose (e.g., as you suggest, to indicate length).
  • @adamsmithWhat are your thoughts on my idea of an easy way to accurately preface an article ID number with a character through a right-click option?188BET靠谱吗I have no concept of the difficulty of implementing this in Zotero.
  • It seems to me that the NPG translator was using the RIS file in the past.I have old saved items where the article number was properly saved to the Pages field.As adamsmith noted, the RIS files list the article number with the SP tag.

    Now the Firefox connector throws this error message: "An error occurred while saving with Nature Publishing Group.Attempting to save using Embedded Metadata instead."

    Here is some debug output from the connector:
    [JavaScript Error: "HTTP GEThttps://www.nature.com/articles/ncomms14906.ris
    failed with status code 0"]
    (3)(+0000000): Translate: resolved tohttps://www.nature.com/articles/ncomms14906.ris
    188BET靠谱吗(3)(+0000000): Zotero.HTTP.doGet is deprecated.188BET靠谱吗Use Zotero.HTTP.request
    (3)(+0000001): Using content XHR
    (3)(+0000000): HTTP GEThttps://www.nature.com/articles/ncomms14906.ris
    (3)(+0000001): Translate: WARNING:
    188BET靠谱吗Zotero.done() called after translator completion with error


    I'm now getting a Number entry in Extra, but it's not the article number.It could be a duplicate of the Issue number.
  • I tried to have a look in the Chrome developer tools.There's a X-Frame-Options: DENYin the response header for the RIS file above.Could this be a problem?188BET靠谱吗(I understand little about how Zotero's translators work.)
    See here: https://stackoverflow.com/a/27359031.
    Article page: https://www.nature.com/articles/ncomms14906.
    Possibly related issue: 188BET靠谱吗https://github.com/zotero/translators/issues/2126.
  • yeah, the RIS seems fine.I use the chrome connector and it doesn't give an error message and seems to not use the RIS and instead uses the metadata.
  • The same issue occurs with other NPG online-only journals that use article numbers.Here are some example articles from such journals:

    scientific reports:
    https://www.nature.com/articles/s41598-020-63484-9
    light: science & applications:
    https://www.nature.com/articles/s41377-020-0298-8
    npj science of food:
    https://www.nature.com/articles/s41538-020-0064-6
  • ah, great.I have a simple "fix" for this here: 188BET靠谱吗https://github.com/zotero/translators/pull/2165

    and can include other NPG journals that are online only.
  • This is now live -- Thanks @qqbbfor mentioning the other journals.The code now contains a list (currently just those for) that's easy to add to.
  • @jeremyvancleve, @adamsmith: Thanks for working on this.I tried to use the new NPG translator, but the issue seems not fixed for Firefox when importing https://www.nature.com/articles/ncomms14906.

    Were you able to identify the issue with the NPG translator?If it's the way the RIS files are served, maybe there's a way to ask the publisher to change this.Given that there are about 150 Nature Research journals according to [1], it would be great if there could be a more general solution that doesn't rely on a list of journals that use article numbers.I also noticed that Nature Energy has changed from article numbers [2,3] to pages [4].

    [1] https://en.wikipedia.org/wiki/Nature_Research#Journals
    [2] https://www.nature.com/nenergy/volumes/1/issues/1
    [3] https://www.nature.com/articles/nenergy20152
    [4] https://www.nature.com/articles/s41560-020-0557-1
  • The fix works with the Chrome connector.(But not with Firefox, see above.)
  • glad to know its working for chrome.I don't have much experience with the firefox connector though...:-/.
  • I've fixed the save problem in Firefox in the latest 188BET靠谱吗Zotero Connector beta.188BET靠谱吗The fix will be included in Zotero Connector 5.0.67 soon.(See 188BET靠谱吗zotero-connectors#313if you're curious about the cause.)
Sign Inor Registerto comment.