Zulip Chat Archive
Stream: Zulip meta
Topic: PhysLean linkifier broken
Joscha Mennicken (Aug 25 2025 at 11:23):
The PhysLean#123 linkifier is linking to https://github.com/leanprover-community/PhysLean/pull/123 (which does not exist) instead of https://github.com/HEPLean/PhysLean/pull/123.
Yaël Dillies (Aug 25 2025 at 11:27):
In fact blahblahblah#123 links to (https://github.com/leanprover-community/blahblahblah/pull/123 for most values of blahblahblah. It's not so much that the linkifier is broken, rather than it doesn't exist!
Joscha Mennicken (Aug 25 2025 at 11:30):
I see, that makes sense. I didn't expect a linkifier this general :D (and I didn't check the list of linkifiers before making this post)
In that case, a custom PhysLean linkifier might be useful.
Joscha Mennicken (Aug 25 2025 at 11:32):
Looking through the list of linkifiers, HEPLean/PhysLean#123 already works, but requires you to type HEPLean/ every time.
Yaël Dillies (Aug 25 2025 at 11:36):
I too am also slightly annoyed that I need to type YaelDillies/LeanAPAP#8 every time, but unfortunately the linkifier is already taken for the docs instead: LeanAPAP#conv
Joscha Mennicken (Aug 25 2025 at 11:49):
Could you not have two linkifiers, one that matches on LeanAPAP#<number> for PRs and one that matches on LeanAPAP#<one non-digit character><arbitrary characters> for the docs?
Yaël Dillies (Aug 25 2025 at 11:59):
That's a good idea!
Kim Morrison (Aug 25 2025 at 12:51):
I am happy to set up xyz#123 linkifiers on request, or to install more complicated linkifiers if provided with candidate regexes.
Joscha Mennicken (Aug 25 2025 at 13:40):
These should work (tested with https://regex101.com/ set to "Golang", which uses the same regex engine that zulip does according to the docs, namely Google's re2):
LeanAPAP#(?P<id>[0-9]+)→https://github.com/YaelDillies/LeanAPAP/pull/{id}LeanAPAP#(?P<decl>[a-zA-Z_'-](?:[\w_.'-]*[\w_'-])?)→https://yaeldillies.github.io/LeanAPAP/docs/find/?pattern={decl}#docPhysLean#(?P<id>[0-9]+)→https://github.com/HEPLean/PhysLean/pull/{id}
Yaël Dillies (Aug 25 2025 at 13:42):
And could we please the same two for YaelDillies/Toric ?
Joscha Mennicken (Aug 25 2025 at 13:43):
That should be
Toric#(?P<id>[0-9]+)→https://github.com/YaelDillies/Toric/pull/{id}Toric#(?P<decl>[a-zA-Z_'-](?:[\w_.'-]*[\w_'-])?)→https://yaeldillies.github.io/Toric/docs/find/?pattern={decl}#doc
Bryan Gin-ge Chen (Aug 25 2025 at 13:44):
Test:
Joscha Mennicken (Aug 25 2025 at 13:45):
I suspect the specific linkifiers must come before the general ones in the linkifier list.
Bryan Gin-ge Chen (Aug 25 2025 at 13:48):
Yeah, it's a bit slow to drag them through the list. Should work now: PhysLean#123
Violeta Hernández (Sep 12 2025 at 19:40):
Kim Morrison said:
I am happy to set up xyz#123 linkifiers on request, or to install more complicated linkifiers if provided with candidate regexes.
Hi! Think we could get a CGT#123 linkifier for vihdzp/combinatorial-games?
Bryan Gin-ge Chen (Sep 12 2025 at 19:56):
test: CGT#123
Kevin Buzzard (Sep 12 2025 at 22:10):
and indeed we discover that they have PRs about surreal birthdays
Kevin Buzzard (Sep 12 2025 at 22:11):
My 50th was a pretty surreal birthday -- I got Real.pi as a present.
Aaron Liu (Sep 12 2025 at 22:27):
Kevin Buzzard said:
and indeed we discover that they have PRs about surreal birthdays
that PR was tricky since I couldn't find any reference to it anywhere online
Aaron Liu (Sep 12 2025 at 22:27):
I had to prove it myself and it took like a week
Last updated: Dec 20 2025 at 21:32 UTC