Add AsRef, as_slice, as_str to more things #65901
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
(rust_highfive has picked a reviewer for you, use r? to override) |
rust-highfive
added
the
S-waiting-on-review
jonas-schievink
added
relnotes
clarfonthey
force-pushed
the
as_ref
branch
5 times, most recently
from
12a83c7
to
26b6aec
Compare
hellow554
reviewed
Oct 28, 2019
| @@ -42,7 +42,7 @@ __pycache__/ | |||
| /src/libcore/unicode/SpecialCasing.txt | |||
| /src/libcore/unicode/UnicodeData.txt | |||
| /src/libcore/unicode/downloaded | |||
| /target/ | |||
| target/ | |||
There was a problem hiding this comment.
Yes; I should have put it in a separate commit. RLS puts the target/ directory in src/libcore for me.
|
The job Click to expand the log.I'm a bot! I can only do what humans tell me to, so if this was not helpful or you have suggestions for improvements, please ping or otherwise contact |
|
Ping from triage. Thank you! |
|
This is libs territory, sorry, thought I'd already done this :) r? @dtolnay |
Centril
reviewed
Nov 7, 2019
dtolnay
reviewed
Nov 10, 2019
There was a problem hiding this comment.
Thanks for the PR! I am going to close because I don't feel that any of the methods or impls added here are sufficiently motivated.
I agree with you that adding associated methods to permutations of iterator adaptors is a rabbit hole.
Regarding the various AsRef impls, I guess I don't understand the use case. It's pretty uncommon for callers to need the genericness of AsRef in writing clear code. Do we really need someone to be able to make a CharIndices and pass it to fs::write in order to express what they mean in the clearest way? Not that permitting it would be bad, it's just not a need that people have and so it's not something that is worth designing for.
In my view the bar for adding surface area to the standard library isn't whether the behavior would be unambiguous to a reasonable observer ("never ultimately lead to unexpected behaviour" in your words), but whether people can get done what they need to get done with an effort that is proportional to how common their individual use case is. Instead of adding the most amount of surface area that remains unambiguous, we need to add the least amount of surface area that makes a usable library.
Centril
removed
the
relnotes
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This is somewhat related to #58957.
When writing some code I noticed the disparity between
CharIndices<'_>andEnumerate<Iter<'_, T>>; while you can get the original slice from the former, you can't from the latter.Essentially, I chose a relatively conservative list of adapters that would reasonably add this functionality where this would never ultimately lead to unexpected behaviour. For example, having
as_refforFilterorStepBywould likely lead to items being shown that aren't part of the iterator, which would be confusing. However,Clonedcan safely assumed to iterate its items in order, and addingas_refto that wouldn't be to weird.Ultimately, this fills in a few holes that seemed reasonable to fill, namely:
str::Byteshas anas_bytesmethod now and implementsAsRef<[u8]>,str::Charsandstr::CharIndicesnow implementAsRef<str>andAsRef<[u8]>. (unsure ifOsStrandPathshould be included, or[u8]should be excluded)iter::{Copied, Cloned, Enumerate}will have anas_slicemethod when wrapped around aslice::Iterorslice::IterMut, and anAsRef<[T]>implementation when wrappingslice::Iter.This honestly seems like an ever-deepening rabbit hole and I'm not sure where it should stop. That said, I feel that adding these methods to the above is shallow enough it shouldn't be too controversial.
Curious to see what people's thoughts are.