Use PathBuf instead of String where applicable #46335
Conversation
src/librustc_driver/driver.rs
Outdated
| } | ||
|
|
||
| pub fn source_name(input: &Input) -> String { | ||
| pub fn source_name(input: &Input) -> FileName { | ||
| match *input { | ||
| // FIXME (#9639): This needs to handle non-utf8 paths |
There was a problem hiding this comment.
Is this FIXME still relevant in this place?
There was a problem hiding this comment.
I think it's fixed with this PR, but there's no repro example on the issue
kennytm
added
the
S-waiting-on-review
oli-obk
force-pushed
the
cleanups
branch
2 times, most recently
from
503e433
to
12c9233
Compare
|
Travis likes it now |
|
☔ The latest upstream changes (presumably #46338) made this pull request unmergeable. Please resolve the merge conflicts. |
|
rebased |
|
Excellent, looks great! Will do detailed review tomorrow. |
src/libsyntax/parse/parser.rs
Outdated
| parser.directory.path = sess.codemap() | ||
| .span_to_unmapped_path(parser.span) | ||
| .to_string() | ||
| .into(); |
There was a problem hiding this comment.
Why do we need PathBuf -> String -> PathBuf here?
There was a problem hiding this comment.
lol. I have no idea what I did there. That was probably some intermediate step to get stuff to compile.
There was a problem hiding this comment.
Ah no, I remember why. This is FileName -> String -> PathBuf. I should probably change the path field to be a FileName
|
There are two |
|
I assumed (wrongly) that |
|
Makes sense! @bors r+ |
|
📌 Commit 7870500 has been approved by |
|
⌛ Testing commit 78705008cc10f1a0fc8495b79ae87561330db61f with merge 4fd6bd5f3faa692795a2907ff09115df2071fe62... |
|
💔 Test failed - status-travis |
|
Legit. Aux builder pretty-fulldeps |
kennytm
added
S-waiting-on-author
|
That was a weird one. |
|
@oli-obk That sounds like a rustbuild bug... |
kennytm
added
S-waiting-on-review
|
@kennytm I'm not sure, since it's explicitly done on the aux builder, this might be intentional |
|
@oli-obk The |
|
@bors r=jseyfried |
|
📌 Commit c273d45 has been approved by |
|
🔒 Merge conflict |
|
☔ The latest upstream changes (presumably #46605) made this pull request unmergeable. Please resolve the merge conflicts. |
|
@bors delegate+ |
|
✌️ @oli-obk can now approve this pull request |
|
@bors r=jseyfried |
|
📌 Commit d732da8 has been approved by |
Use PathBuf instead of String where applicable r? @jseyfried
|
☀️ Test successful - status-appveyor, status-travis |
|
@oli-obk RLS and Rustfmt have both been fixed (using your commits), could you send a PR to update them in the Rust repo and unbreak them for testing (if you only touch the submodules and Cargo.lock, then pre-emptive r=me, p=1). |
|
on it |
This fixes an accidental regression rust-lang#46335 where the behavior of `Path::ends_with` is different from `str::ends_with` (paths operate over components, strs operate over chars).
…on, r=estebank rustc: Spawn `cmd /c` for `.bat` scripts This fixes an accidental regression rust-lang#46335 where the behavior of `Path::ends_with` is different from `str::ends_with` (paths operate over components, strs operate over chars).
…on, r=estebank rustc: Spawn `cmd /c` for `.bat` scripts This fixes an accidental regression rust-lang#46335 where the behavior of `Path::ends_with` is different from `str::ends_with` (paths operate over components, strs operate over chars).
This fixes an accidental regression rust-lang#46335 where the behavior of `Path::ends_with` is different from `str::ends_with` (paths operate over components, strs operate over chars).
r? @jseyfried