The span on a inner doc-comment would point to the next token, e.g. the span for the `a` line points to the `b` line, and the span of `b` points to the `fn`.
```rust
//! a
//! b
fn bar() {}
```
Fix#13965.
This commit adopts the second strategy I outlined in #13965, where the bulk of the code is still "smoke tested" (in the sense that rustdoc attempts to run it, sending all of the generated output into a locally allocated `MemWriter`). The part of the code that is ignored (but included in the presentation) is isolated to a three-line `main` function that invokes the core rendering routine.
In the generated rustdoc output, this leads to a small break between the two code blocks, but I do not think this is a large issue.
for `~str`/`~[]`.
Note that `~self` still remains, since I forgot to add support for
`Box<self>` before the snapshot.
How to update your code:
* Instead of `~EXPR`, you should write `box EXPR`.
* Instead of `~TYPE`, you should write `Box<Type>`.
* Instead of `~PATTERN`, you should write `box PATTERN`.
[breaking-change]
Fix#13965.
There is a dance here between the `main` that actually runs versus the
`main` that is printed in the output documentation. We don't run the
latter `main`, but we do at least compile (and thus type-check) it.
It is still the responsibility of the documenter to ensure that the
signatures of `fn render` are kept in sync across the blocks.
Namely:
* Added conversion traits both to and from the various vector types,
analogous to how `str::MaybeOwned` works with `str::IntoMaybeOwned`
and `str::Str`. This led me to add the `FixedLen` variant of
`MaybeOwnedVector` for interoperability with `std::slice`.
* Revised client example code to make use of `into_maybe_owned`
* Added implementations of `Show` and `CloneableVector` for
`MaybeOwnedVector`.
* As suggested by kballard, added `into_vec` method that is analogous
to `CloneableVector::into_owned` except it produces a `Vec` rather
than a `~[T]`.
