automatic bibtex (was Re: [Skim-app-users] Updating references)

Mike Marchywka marchywka at
Fri Feb 21 20:47:45 CET 2020

On Fri, Feb 21, 2020 at 01:19:37PM -0500, Thomas Schneider via texhax wrote:
> Jason:
> > For a lot of the manuscript and grant applications I have to include
> > the pmid and PMC numbers that pubmed provides. I've added the Pmc
> > filed as a default for all references. This works just fine for new
> > references, but is blank in my previous references. I really don't
> > want to go through the processes of re-importing, dealing with
> > duplicates etc. because that would take a long time. How can I
> > simply update my references?
> Since this is not related to Skim display, this discussion would be
> better held on the texhax news group (texhax at
> so I am ccing there and
> suggest that after this we remove the Skim references.
> I have a tcsh script (!) that calls a Pascal (!) program:
> Given a PubMed Id (pmid) together they grab the reference format and
> create a Bibtex entry.  Then they look in my bibliography file to see
> if the same key has been used before (first author, last author, year
> which is nearly unique in PubMed).  If so, they are compared using
> sdiff.  If the new entry has a doi, that is provided in a file.  I
> then edit my bibliography file BY HAND to upgrade the reference.  I do
> it by hand to avoid disasters of programs doing unexpected things.

I have a bunch of scripts that try to extract bibtex
from any url on the clipboard. See the previous complaints and maybe
the "bomtex" messages :) For pubmed, my original target, I just make
the citation the pubmed or pmc id number prefixed with pmid or pmc. 
At first, this was nice because they could be recognized as missing
and fetched during the build. My bibtex often includes various url's
and if one of them points to pubmed I guess it would be easy to 
extract the info. I also had a concern about bibliography management.
I remember I wrote some c++ code to parse and play with entries
but I'm sure it is not a standard parse on corner or bad bibtex.
IIRC the ideal was something like a code diff/merge that would have
automated and manual options. 
This did give me some possibility to look for possible dups and
errors. Ideally it would let me merge and clean up and finally integrate
URL's or non-standard fields ( again see the "bomtex" that I have
dropped for now).

My interest is still just getting all sites- I've noticed wikipedia
does offer some bibtex now but still things like news sites and
clinical trials or other research sites do not - to have a button
to put bibtex on the clipboard or at least make it available from all relevant

> Tom
>   Thomas D. Schneider, Ph.D.
>   Senior Investigator
>   National Institutes of Health
>   National Cancer Institute
>   Center for Cancer Research
>   RNA Biology Laboratory
>   Biological Information Theory Group
>   Frederick, Maryland  21702-1201
>   schneidt at


mike marchywka
306 charles cox
canton GA 30115
USA, Earth 
marchywka at
ORCID: 0000-0001-9237-455X

More information about the texhax mailing list.