BUG: Safari BMJ citation drags in other elements

Bren777 · July 18, 2017

When using Safari to add a citation for a BMJ article (http://blogs.bmj.com/bmj/2017/07/18/social-prescribing-offers-huge-potential-but-requires-a-nuanced-evidence-base/) Zotero added other elements to the collection. Some were blank elements with attached links, which look like they were drawn from the web page. Others were named: "Disqus Comments", "Disqus Realtime Notification", "Facebook Cross-Domain Messaging helper". It also added the citation correctly.

adamsmith · July 20, 2017

I can't replicate this but I don't have Safari. Are any of the folks with Mac here seeing this?

DWL-SDCA · July 20, 2017

The behavior Bren777 described is likely more related to the blog source instead of Safari. I get similar behavior with BMJ Group blog posts when I use Chrome or Firefox. When I attempt to download a regular article from any BMJ Group journal things work as they should. However, when I try to download an online letter or blog post I can get nothing but a mostly empty webpage record or I get a webpage with not only what is visible on the webpage but social media bizarreness, adverts (including some that were not displayed on the page I viewed) and all sorts of usage tables and graphs. Sometimes online-only letters have DOIs. Zotero will sometimes successfully import those records but with less than full metadata.

Blog posts and online-only letters that do not have DOIs import as a webpage but sometimes with crazy metadata. When I look at the header for embedded metadata for these blog posts there seems to be all sorts of random stuff within the regular tags. The "stuff" can include snippets from various parts of the text on the page, text from what is displayed behind images, mouse-over text, text from the footer that is repeated several times, etc.