I have two issues with the section "Deref and method calls" of the book's chapter "Deref coercions".
- (Minor) It says "In other words, these are the same two things in Rust:", followed by a code block in which no two things seem similar, much less the same. Presumably this sentence made more sense in a previous revision.
- The next paragraph conflates two concepts which, imho, should kept separate. They are
- deref coercion, i.e. inserting as many `*` as necessary and
- implicitly referencing the receiver, i.e. inserting a single `&` to satisfy the method's `self` parameter type.
I appreciate that with the proposed changes the example becomes very contrived, even for a foo-bar-baz one. However, the current exmplanation is just wrong.
This adds missing `?` marks to productions for loops and break/continue.
It also adds missing option label to while let loop.
Note that '[' foo ']' means grouping in BNF, and '?' is used for possible missing items.
r? @steveklabnik
this resolves type-variables early in assemble_candidates and
bails out quickly if the self type is an inference variable (which would
fail anyway because of `assemble_candidates_from_projected_tys`).
In both these cases, `assemble_candidates_from_impls` would try to go
over all impls and match them, leading to O(`n*m`) performance. Fixing this
improves rustc type-checking performance by 10%. As type-checking is only
is 5% of compilation, this doesn't impact bootstrap times, but *does*
improve type-error-detection time which is nice.
Crates that have many dependencies and contain significant amounts of
generic functions could see a bigger perf boost. As a microbenchmark,
the crate generated by
```
echo '#![feature(rustc_private)]'
echo 'extern crate rustc_driver;'
for i in {1..1000}; do cat << _EOF_
pub fn foo$i<T>() {
let mut v = Vec::new();
let _w = v.clone();
v.push("");
}
_EOF_
done
```
sees performance improve from 7.2 to 1.4 seconds. I imagine many crates
would fall somewhere in-between.
r? @nikomatsakis
Nothing too big, a few needless returns and a few closures eliminated (the latter may improve performance in some cases, at least compilation should be a bit faster).
And replace more `0 as *const T`/`0 as *mut T`s with `null()`/`null_mut()`s
I'm not sure what is the general policy about making simple functions `const`, but `null()` and `null_mut()` seem to be good candidates.
Originally in an example it reads as follows:
```rust
fn inverse<T>() -> T
// this is using ConvertTo as if it were "ConvertFrom<i32>"
where i32: ConvertTo<T> {
42.convert()
}
```
There was no mention of `ConvertFrom` elsewhere in the page other than in this comment. Is this supposed to be `ConvertTo<i64>` ?
I'm confused by this example.
Conventionally in C `*mut T` is a transfer of ownership where `*const T` is a
loan, so `*mut T` is likely the more appropriate return type for these
functions. Additionally, this more closely mirrors the APIs on `Box` for this
sort of functionality.
cc #27769
