I use the bookmarklet to import from my web browser to Bookends browser extensively. Sadly quite a few sites fail to import from the bookends browser (they fail to support COiNs or use all sorts of non-standard metadata etc.). Also even if COiNs is supported, the import often fails to have an abstract or PMID or other critical (to me at least) information and requires more fiddling and searching, and with many refs to work thorough this has a substantial workflow impact.
Anyway, if i could select the title or DOI and just run a Pubmed search on the selection within the bookends browser immediately it would really make this process faster. A new bookmarklet to do this from within my own browser would be even more wonderful!!!
Thanks Jon, defaulting to Pubmed would certainly help. One other problem with Bookends browser is its speed. I assume it doesn't use a cache because it it much slower at loading pages than my browser. Sometimes it seems to be waiting for resources and until everything is in, including all the advertising, one has to wait before one can click to import. Then depending on the site (lots of COiN span elements or other metadata to parse), Bookends "hangs" without any feedback (the cursor turns to a static watch), until all the potential links are scanned, which can take a much longer time. The Pubmed search interface on the other hand is blazingly faster.
I assume you already have the difficult work done in using the bookmarks:// URL system in place, and instead of the encoded URL, you use the encoded selected text.
This bookmarklet shows the selected text as an alert:
COinS are rarely used. The vast majority of "hits" are DOIs. Bookends verifies each DOI and gets the title before it shows them to you. Some pages have dozens of DOIs, and Bookends has to verify them all. Depending on your internet latency, this can take a few seconds. I believe that you can halt the process with Command-. if it's taking a long time, and Bookends will display what it has found up until that point. This thread is veering off course. Please contact me directly if you want to discuss anything further.
Thank you whoever posted this - this is the reason I have not purchased Bookends yet. Please consider the following feature requests that I hope would benefit many:
1. Selected PDF text could be searched (via right-click) using PubMed.
2. Selected PDF text could be assigned (via right-click) to reference title, journal abbreviated or long name, article title, authors, volume, issue, start page and end page.
3. If the PMID (PubMed ID) field would be set, search would use PubMed as default and update fields (PubMed keywords, for example, may be post-edited).
4. If items such as journal name, volume, issue, etc. would be set, the PubMed search text should optionally restrict the search to those journals and volumes. For example, articles may have a generic title such as "Reply" for their published correspondence and this would help limit the number of PubMed hits.
5. Ideally, multiple search results should show as a list, e.g. presented as a table, in order for the user to select the correct reference.
I currently use EndNote. To accomplish the above, I have written a program (sorry, it is for Windows, but works when ENdNote is running on MacOS X and the program is running under Windows using Parallels) that runs in the background does an optionally restricted search (restrictions are based on journal, year, and/or volume) for journal titles, and if a hit is made, it audio-signals success and replaced the clipboard contents with the PubMed ID. I then paste the clipboard in EndNote's "Accession Number" field and when update the reference, EndNote can always correctly update the reference using PubMed.
Jon: If you want to look at that program and source (Delphi & VCL framework), you would be welcome.
This thread is getting confused (or at least I am). The original poster was talking about a browser. You're talking about PDFs (in Bookends I assume, but that's not clear to me). Assuming it is, you can do what you're describing (at least some of it) when using Autocomplete Paper.
Thanks for the reply - yes you are correct, I was referring to PDFs. I did not grasp the "Autocomplete Paper ..." function's logic. It appears to create a new detached reference (shows two paperclips but no attachment - please see attachment).
The Autocomplete Paper function's workflow might be improved if selected PDF text could be transferred to specific search fields using a contextual popup menu choice.
Thanks again,
Navid
Screen Shot 2016-02-20 at 6.24.24 PM.jpg (227.85 KiB) Viewed 13083 times
That is not a picture of Autocomplete Paper. That is the browser view (in the lower pane), which shows you the reference online (if you have the URL in the URL field). Once there, you can navigate to the PDF and click the paperclip icon to download it and attach it to that reference (you an also attach the web page as a webarchive). So this is useful, among other things, when you have the metadata but not the PDF.
Autocomplete Paper is = the opposite -- you use it when you have the PDF but not the metadata. It has a mechanism for putting the selected text to a specific field (for searching, or in the database) via the Action popup menu (although typically its' easier to type in a few words, that's often enough). To see this window, select a reference with a PDF and use Refs -> Autocomplete Paper. All the better if there is no metadata in the reference, just the attachment (the PDF).
You would benefit from the Bookends tutorials (in the Help menu), both PDF and video. And peruse the TOC of the user guide (check out "Autocomplete Paper"). If you are coming to Bookends from another app you're expectations will naturally be that it works the same way you're used to, and quite often that's not the case.
Yes - I knew that was the browser view. I started with an unrecognized PDF, from where I went to the Autocomplete Paper dialog and imported the relevant citation. However, I ended up with the unexpected result that I showed in the image.
In other words, I started with a PDF containing no metadata. I used Autocomplete Paper, but I lost the PDF attachment in the process of importing. Reading your prompt response (I really appreciate it!) and having already viewed the help resources I don't think I carried out any incorrect steps.
Autocomplete Paper doesn't add an attachment, it imports the metadata, so something is being lost in the translation, so-to-speak. Let's follow this up off-forum. Please send me screen snaps of exactly what you did. You might start with a fresh new database so that there is no ambiguity about what is going on.
