Better docs for the str primitive #30057
Conversation
|
(rust_highfive has picked a reviewer for you, use r? to override) |
steveklabnik
force-pushed
the
doc_str
branch
from
0c13e4b to
a4ee9f5
Compare
| /// Returns the length of `self` in bytes. | ||
| /// Returns the length of `self`. | ||
| /// | ||
| /// This length is in bytes, not [`char`]s, or graphemes. In other words, |
There was a problem hiding this comment.
I think the comma after chars is potentially confusing.
steveklabnik
force-pushed
the
doc_str
branch
from
a4ee9f5 to
a4142e9
Compare
| /// | ||
| /// # Safety | ||
| /// | ||
| /// Caller must check both UTF-8 sequence boundaries and the boundaries | ||
| /// of the entire slice as well. | ||
| /// Callers of this function are responsible that three sets of bounds are |
There was a problem hiding this comment.
I usually call these things “preconditions”. Is there any pretext to use “sets of bounds” instead of “preconditions” or “requirements” or another familiar synonym?
There was a problem hiding this comment.
There isn't, actually. It would be good to check what we use elsewhere, hmmm
steveklabnik
force-pushed
the
doc_str
branch
from
a4142e9 to
d989834
Compare
|
Okay, I think I've gotten all of the review questions except what to do about pointing to the unicode crate out of the way. |
steveklabnik
force-pushed
the
doc_str
branch
from
d989834 to
7b30f5c
Compare
| /// | ||
| /// The pattern can be a simple `char`, or a closure that determines | ||
| /// The pattern can be a `&str`, [`char`], or a closure that determines |
There was a problem hiding this comment.
I don't think the pattern can be a &str: https://play.rust-lang.org/?gist=85340c83e50bc58d899b&version=stable
There was a problem hiding this comment.
@CamJN in the future, can you file a new issue rather than making a comment on an old PR? This comment will get lost in the sands of time. Thanks!
(Also, I think you're right here)
Part of #29338