#Failed to parse biblatex: wrong number of digits
47 messages Β· Page 1 of 1 (latest)
https://bib-it.sourceforge.net/help/fieldsAndEntryTypes.php
The only one a bit ambiguous is the volume. But they are clear about the year field in bibtex format
It's possible that, in the future, typst allows more flexible year specification (an others)
thanks very much for this!
although i can probably edit the volume sections to only include numbers or just leave the info out altogether
the year line has to be able to include no date, to appear, in prep, etc
ill try to send it on as a suggestion
there's quite a number of core sources in my field that have no date or are perpetually forthcoming, so ill have to wait it out till it's possible!
I dont know if youre using something already, but I greatly encourage you to try Zotero https://www.zotero.org/
It comes with an app for windows (and works in mac and linux too) + a plugin for chrome or firefox to easily add new entries to your bibliography (and has cloud services too, so you can sync between computers)
The advantage of this is that it parses the fields of each entry (thus being type checked). You can later export the bibliography to bibtex format (or many others)
You can of course add entries manually
Zotero is a free, easy-to-use tool to help you collect, organize, cite, and share research.
yes, ive an extensive zotero library, thanks! ive been building it for some six or seven years
they allow non-numerals in the year and volume section without a problem
maybe i could change some settings there to do some data checking though
i dont regularly add books, so that may be why i never had that problem
and i tested typst with my 70 long bibliography
i got lucky i guess π€£
I ran into this today @gentle briar did you figure anything out? Is there a Zotero setting to force the year as numeric?
@errant nest i didnt bother tryin unfortunately
was hopin itd be resolved by the dev team before i need to write up my next paper tbh lmao π
sorry though!!
No worries! Hope it gets sorted soon! thanks for the reply
There has been discussion previously about moving to some (more standard if I understand correctly) biblatex parser due to the very strict nature of the current one. It's causing issues for me as well, I have to manually edit dozens of sources and they end up being incorrect π
ive confirmation now that the problem has been added to the issue tracker
no word on when itll be addressed but the devs have been so fast so far and all
Honestly, it's such a chore for me that I have half a mind to open a PR for it
But I have done so many PRs already
And I am currently focusing on a image manipulation PR and then I need a proper glossary PR
where I am honestly thinking of just modifying the existing Terms system
the problem is so serious for me that i cant use typst for my work
ive many many sources that are unpublished (not uncommon for linguists specialised in the americas!)
so anything to make that easier will be welcomed by me! π
Once the figure PR I am working on is merged, I may take a stab at it, at least a fork that works for me and perhaps for you too
Since I need it too
But I probably won't be the one to turn it into a PR
I need to write my thesis damn it
When having to do stuff for Uni all other things are so much more fun to do all of a sudden π
Indeed π
We plan to rework the bibliography system, probably with citeproc. But that will probably take some time. I hadn't planned to spend time on the old system (as it's going to be replaced), but if this bug is so severe, maybe we should first relax the current system a bit and then work on the new one.
@solar granite relaxing it for now would already help a lot of people I think
We'll do that
(Okay, we should do it anyway as I was just instructed. Even if we're moving to citeproc, we'll continue to need the biblatex parser.)
Really? I am quite surpised that you need the biblatex parser for citeproc, doesn't it come with its own?
Ah yes I see!
It takes a special JSON format as an input
so you'd need to convert bibtex to that JSON format
Oof the API is not pretty
structs with 29 fields? oof
and the project seems dead
no updates in two years
Mind you it is WASM compatible which is nice
That's also why we're a bit underwhelmed by the idea of moving to it