Closed dimpase closed 2 years ago
looks good, thank you! my test was to check the sha1sums of the previously generated artefacts, then update all of them using the new code and, without surprise, sha1sums are the same 🥳
~/website/publications$ sha1sum *.html
04e430ca8d9bf23ef13e7c16f053a9dc0f89fa83 publications-combinat.html
3768809a2388b75933688cdca4538347eb1648bf publications-general.html
cefc1b2e8480261370e7e53be748aee2b355fe79 publications-mathscinet.html
1c6435038c9eb624ad3864e984e750980fbbe1b7 publications-mupad.html
~/website/publications$ make clean; make
[...]
~/website/publications$ sha1sum *.html
04e430ca8d9bf23ef13e7c16f053a9dc0f89fa83 publications-combinat.html
3768809a2388b75933688cdca4538347eb1648bf publications-general.html
cefc1b2e8480261370e7e53be748aee2b355fe79 publications-mathscinet.html
1c6435038c9eb624ad3864e984e750980fbbe1b7 publications-mupad.html
I tested my changes locally, by running make
after changing few journaltitle
to journal
in one of .bib files. So from now on, journal
should be fine.
Going forward, one should use a modern tool, e.g. Sphinx with a BibTeX extension, to produce property formatted webpages from .bib files.
Is there any reason not to use make4ht+bibtex to produce a MathJax-enabled html out of bibfiles? It will need a short .tex file to load all the bibfiles, but then one can get rid of all this crude Python stuff.
I'm not sure who you ask. I haven't written this tool, everything is possible I guess? From what I know about this, the basic constraints are to produce html code, which will then be embedded as it is on the sage website (i.e. no extras, just the plain list), and to be able to run this tool on Linux.
It would for sure be much simpler, and solve problems of displaying TeX properly.
e.g. I've done https://cddlib.github.io/polyhedral_faq/ this way. https://github.com/cddlib/polyhedral_faq
(even the deployment is done via GitHub's Actions, automatically)
This allows using
journal
withoutjournaltitle
, etc.