array::try_map

[eopb]

TI: #79711

pub fn try_map<F, R, E, U>(self, mut f: F) -> Result<[U; N], E> where F: FnMut(T) -> R, R: Try<Ok = U, Error = E> { ... }

Thanks to @lcnr for helping me on this PR.

Much of the code was copied from array::map.

[rust-highfive]

r? @withoutboats

(rust-highfive has picked a reviewer for you, use r? to override)

[withoutboats]

Thanks for the PR. I haven't been involved in the decisions to add any of the relevant methods for evaluating this one, so I want to consult with the rest of the libs teams on this. Specifically, I think we need to develop more policy around these questions:

1. What guidelines do we use to determine if its worthwhile to add a try_ variant of a higher order function?
2. What guidelines do we use to determine if its worthwhile to add iterator-like combinators directly to array?
3. This uses the Try bound, which I notice, so do the stable Iterator try methods; however, it necessarily also returns a Result, rather than R, because its not possible to return R, due to lack of HKT. Should we have the R parameter at all, or just limit this to Result? (fwiw I was somewhat surprised to discover the Iterator methods use a Try bound and I'd like some consensus around that even without this issue)

@rust-lang/libs

[scottmcm]

A few off-the-cuff musings,

1. I think it would make sense for there to be at least one "fallible constructor" kind of thing for arrays. Whether that's try_map, I don't know -- it could be Add 'core::array::from_fn' and 'core::array::try_from_fn' #75644try_generate instead, or something else, or it might be fine to have both. It's straight-forward to do generate as <[(); N]>::map or map as generate-with-mutable-closure, so multiple doesn't seem strictly necessary.
2. A rule similar to whether to have them on Option seems plausible to me -- which I guess is only an argument for map, not for try_map. Hmm, try_map is similar to .map().transpose() on Option, so perhaps transpose: [Result<T, E>; N] -> Result<[T; N], E> could be argued for, but I guess that doesn't quite work the same as it has different short-circuiting.
3. I agree it would be undesirable for a closure producing Option<U>s to end up with a Result<[U; N], NoneError> from this method. Looks like the GAT need has come up on the Iterator::try_find tracking issue too, and something like that would be nice, but it's probably far off so only using Result would probably be better near-term.

[bors]

☔ The latest upstream changes (presumably #79451) made this pull request unmergeable. Please resolve the merge conflicts.

Note that reviewers usually do not review pull requests until merge conflicts are resolved! Once you resolve the conflicts, you should change the labels applied by bors to indicate that your PR is ready for review. Post this as a comment to change the labels:

@rustbot modify labels: +S-waiting-on-review -S-waiting-on-author

[rust-log-analyzer]

The job x86_64-gnu-llvm-9 failed! Check out the build log: (web)(plain)

Click to see the possible cause of the failure (guessed by this bot)

extracting /checkout/obj/build/cache/2020-11-19/rustfmt-nightly-x86_64-unknown-linux-gnu.tar.xz
error: failed to read `/checkout/src/tools/cargo/crates/credential/cargo-credential-1password/Cargo.toml`
Caused by:
No such file or directory (os error 2)
Build completed unsuccessfully in 0:00:15
make: *** [prepare] Error 1
Makefile:60: recipe for target 'prepare' failed
Command failed. Attempt 2/5:
Command failed. Attempt 2/5:
error: failed to read `/checkout/src/tools/cargo/crates/credential/cargo-credential-1password/Cargo.toml`
Caused by:
No such file or directory (os error 2)
Build completed unsuccessfully in 0:00:00
make: *** [prepare] Error 1
Makefile:60: recipe for target 'prepare' failed
Command failed. Attempt 3/5:
Command failed. Attempt 3/5:
error: failed to read `/checkout/src/tools/cargo/crates/credential/cargo-credential-1password/Cargo.toml`
Caused by:
No such file or directory (os error 2)
Build completed unsuccessfully in 0:00:00
make: *** [prepare] Error 1
Makefile:60: recipe for target 'prepare' failed
Command failed. Attempt 4/5:
Command failed. Attempt 4/5:
error: failed to read `/checkout/src/tools/cargo/crates/credential/cargo-credential-1password/Cargo.toml`
Caused by:
No such file or directory (os error 2)
Build completed unsuccessfully in 0:00:00
make: *** [prepare] Error 1
Makefile:60: recipe for target 'prepare' failed
Command failed. Attempt 5/5:
Command failed. Attempt 5/5:
error: failed to read `/checkout/src/tools/cargo/crates/credential/cargo-credential-1password/Cargo.toml`
Caused by:
No such file or directory (os error 2)
Build completed unsuccessfully in 0:00:00
make: *** [prepare] Error 1
Makefile:60: recipe for target 'prepare' failed
The command has failed after 5 attempts.

[bors]

☔ The latest upstream changes (presumably #80566) made this pull request unmergeable. Please resolve the merge conflicts.

[JohnCSimon]

Ping from triage - @ethanboxx
can you please resolve the merge conflict and build failure?

@rustbot label: -S-waiting-on-review +S-waiting-on-author

[JohnCSimon]

Triage:
@ethanboxx - unfortunately I'm closing this as inactive. Thanks for contributing. If you're willing to continue on this please open a new PR.