dist/cache: include paths in error contexts

[mathstuf]

Fixes: #885

---

In the interest of helping 0.4.0 get finalized :) .

[codecov-commenter]

Codecov Report

Base: 30.77% // Head: 30.84% // Increases project coverage by +0.06% 🎉

Coverage data is based on head (b02e044) compared to base (59e906d).
Patch coverage: 0.00% of modified lines in pull request are covered.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1611      +/-   ##
==========================================
+ Coverage   30.77%   30.84%   +0.06%     
==========================================
  Files          49       49              
  Lines       16611    16618       +7     
  Branches     8044     8049       +5     
==========================================
+ Hits         5112     5125      +13     
+ Misses       6217     6213       -4     
+ Partials     5282     5280       -2     

Impacted Files	Coverage Δ
src/dist/cache.rs	20.37% <0.00%> (-0.55%)	⬇️
src/test/tests.rs	33.33% <0.00%> (-0.69%)	⬇️
src/compiler/compiler.rs	36.03% <0.00%> (-0.08%)	⬇️
src/compiler/rust.rs	33.81% <0.00%> (-0.07%)	⬇️
src/lib.rs	11.37% <0.00%> (ø)
src/compiler/args.rs	62.43% <0.00%> (+0.17%)	⬆️
src/server.rs	31.08% <0.00%> (+0.20%)	⬆️
src/compiler/c.rs	38.16% <0.00%> (+0.22%)	⬆️
src/compiler/gcc.rs	54.75% <0.00%> (+0.30%)	⬆️
src/util.rs	37.50% <0.00%> (+1.04%)	⬆️
... and 2 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[Xuanwo]

Thanks for contribution, LGTM!

[sylvestre]

Would it be possible to have tests to cover these errors?

[mathstuf]

Maybe? What fs abstractions do we have available? FUSE is a possibility, but feels like a lot of code to test error paths. Also doesn't help with Windows at all.

[drahnr]

Is there a particular reason to not use fs-err instead?

[mathstuf]

That seemed like a wider-effecting change than the issue was asking for. And given that 0.4.0 is now "one" issue away by milestoned issues…I'd like to see 0.4.0 come out first personally.