Closed rish987 closed 3 years ago
Merging #119 (dd1a73a) into main (d0307a1) will decrease coverage by
0.55%
. The diff coverage is82.35%
.
@@ Coverage Diff @@
## main #119 +/- ##
==========================================
- Coverage 94.73% 94.18% -0.56%
==========================================
Files 31 31
Lines 1690 1702 +12
==========================================
+ Hits 1601 1603 +2
- Misses 89 99 +10
Impacted Files | Coverage Δ | |
---|---|---|
lua/lean/ft.lua | 71.42% <71.42%> (-28.58%) |
:arrow_down: |
lua/lean/lean3.lua | 77.77% <90.00%> (-22.23%) |
:arrow_down: |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update d0307a1...dd1a73a. Read the comment docs.
@gebner I think there might be a slight MacOS bug with leanprover/elan#36, see test failures -- since our default toolchain in CI is Lean 3 (and the error mistook lean
for lean3
), I have a suspicion that it has to do with leanprover/elan#36 checking the path case-sensitively, while MacOS has a case-insensitive (all-lowercase) filesystem.
I have a suspicion that it has to do with leanprover/elan#36 checking the path case-sensitively, while MacOS has a case-insensitive (all-lowercase) filesystem.
This is a potential explanation. But I think we need someone with access to a MacOS machine to debug this (and test the fix).
I have a mountain of things on my plate now for the next few days so I'm going to be quite slow at many things, but will see if I can circle back when I get a moment.
Nevermind Gabriel your code is good (also tested it at rish987/elan@6658cab, let me know if that's worth upstreaming). I fixed this by schedule_wrap
ing our elan
- based filetype detection.
Can you separate (and feel free to merge separately) the unrelated changes here so it's easier to see what this does?
Closes #88,
failed Lean 3 stdlib ft test pending leanprover/elan#36. Also fixed the stdlib filetype tests.