joostkremers
Possibility of upstreaming
I am wondering if contributing the parser upstream might be an option. Especially for parsing .bib files.
The built-in bibtex.el implements an incomplete regexp-based that cannot parse some more complex cases like described in https://debbugs.gnu.org/cgi/bugreport.cgi?bug=69266
It would be nice to have a dedicated parser that is more accurate available to core Emacs packages and ELPA packages. Specifically, Org mode (and maybe bibtex.el itself) could benefit.
WDYT?
I would be fine with upstreaming parsebib, though I imagine there could be some push-back from the maintainers: the bug you link to may not be convincing enough, given that BibTeX, biblatex and biber don't seem to agree on the proper .bib format... :thinking: I tend to follow biblatex/biber in such cases (unlike bibtex-mode, it seems).
I'm gonna ping @Hugo-Heagren on this, BTW, because the code that cleans TeX markup was originally contributed by him to Ebib. I moved it to parsebib to make it more generally available. The Git log doesn't mention him as contributor (I realise now I probably should have added some acknowledgement...) but it is his code and it's substantial enough that it requires a copyright assignment.
There are also contributions by others, but I haven't looked at the sizes yet.
I'm gonna ping @Hugo-Heagren on this, BTW, because the code that cleans TeX markup was originally contributed by him to Ebib. I moved it to parsebib to make it more generally available. The Git log doesn't mention him as contributor (I realise now I probably should have added some acknowledgement...) but it is his code and it's substantial enough that it requires a copyright assignment.
I'm flattered!
I'd be very happy for my code (or any part of parsebib including it, or anything which is built out of that) to be upstreamed into Emacs. I have actually contributed to Emacs core before, so I've done the copyright assignment already. Should be good to go.
Joost Kremers @.***> writes:
I would be fine with upstreaming parsebib, though I imagine there could be some push-back from the maintainers: the bug you link to may not be convincing enough, given that BibTeX, biblatex and biber don't seem to agree on the proper
.bibformat... :thinking: I tend to follow biblatex/biber in such cases (unlikebibtex-mode, it seems).
That's part of the point! Some .bib files are designed for biblatex/biber and won't work with bibtex. It would be nice if they were supported.
Moreover, having an actual recursive parser is much better than regexp-based.
Also, see https://debbugs.gnu.org/cgi/bugreport.cgi?bug=51621
There are also contributions by others, but I haven't looked at the sizes yet.
In the logs, the only other significant contributor is Rahguzar I think. Others are within non-trivial 15-20LOC limit.
-- Ihor Radchenko // yantar92, Org mode maintainer, Learn more about Org mode at https://orgmode.org/. Support Org development at https://liberapay.com/org-mode, or support my work at https://liberapay.com/yantar92
There are also contributions by others, but I haven't looked at the sizes yet. In the logs, the only other significant contributor is Rahguzar I think. Others are within non-trivial 15-20LOC limit.
Yeah, looks like.
So, pinging @aikrahguzar : would you be willing (or have you already) signed a copyright assignment form with the FSF to transfer copyright of your contribution to parsebib to the FSF?
So I guess the next question would be: what is the procedure for upstreaming, and who is gonna get the ball rolling?
So, pinging @aikrahguzar : would you be willing (or have you already) signed a copyright assignment form with the FSF to transfer copyright of your contribution to parsebib to the FSF?
Actually, reading through PR #25 , I noticed @aikrahguzar already filed the paperwork, so it looks like we're good to go!
Joost Kremers @.***> writes:
So I guess the next question would be: what is the procedure for upstreaming, and who is gonna get the ball rolling?
The procedure is simple: write to emacs-devel and CC me + bibtex maintainer (Roland Winkler). You can say that at least Org mode (`org-cite-basic--parse-bibtex') is interested in more robust parser.
We will go from there and see what Eli and others say.
-- Ihor Radchenko // yantar92, Org mode maintainer, Learn more about Org mode at https://orgmode.org/. Support Org development at https://liberapay.com/org-mode, or support my work at https://liberapay.com/yantar92