rustdoc: refactor rustdoc syntax highlighting for a more flexible API
Clients can now use the rustdoc syntax highlighter to classify tokens, then use that info to put together there own HTML (or whatever), rather than just having static HTML output.
Since these impls can be scattered around quite a bit, it is nice
to be able to jump to the location where individual methods and
trait impls are defined.
Fixes: #30416
Clients can now use the rustdoc syntax highlighter to classify tokens, then use that info to put together there own HTML (or whatever), rather than just having static HTML output.
Especially in cases like the one in the test file, this can blow
up the docs big time if string constants in the code contain HTML.
But also other constants can contain special chars (e.g. `&` as an
operator in constant expressions).
rustdoc: Cleanup ABI rendering
Use a common method for rendering `extern "<abi>"`.
This now consistently shows `extern fn` rather than `extern "C" fn`.
Make some fatal lexer errors recoverable
I've kept the changes to a minimum since I'm not really sure if this approach is a acceptable.
fixes#12834
cc @nrc
rustdoc: Handle concurrent mkdir requests
It's likely that `rustdoc` as a tool is run concurrently in the same output
(e.g. documenting multiple crates as Cargo does), in which case it needs to
handle concurrent calls to `fs::create_dir`.
rustdoc: inline all the impls
This used to be done to avoid inlining impls referencing private items, but is now unnecessary since we actually check that impls do not reference non-doc-reachable items.
fixes#32881fixes#33025fixes#33113
r? @alexcrichton
It's likely that `rustdoc` as a tool is run concurrently in the same output
(e.g. documenting multiple crates as Cargo does), in which case it needs to
handle concurrent calls to `fs::create_dir`.
Due to inlining it is possible to visit the same module multiple times
during `<Cache as DocFolder>::fold_crate`, so we keep track of the
modules we've already visited.
rustdoc: refine cross-crate impl inlining
This changes the current rule that impls within `doc(hidden)` modules aren't inlined, to only inlining impls where the implemented trait and type are reachable in documentation.
fixes#14586fixes#31948
.. and also applies the reachability checking to cross-crate links.
fixes#28480
r? @alexcrichton
This changes the current rule that impls within `doc(hidden)` modules
aren't inlined, to only inlining impls where the implemented
trait and type are reachable in documentation.
Implement `pub(restricted)` privacy (RFC 1422)
This implements `pub(restricted)` privacy from RFC 1422 (cc #32409) behind a feature gate.
`pub(restricted)` paths currently cannot use re-exported modules both for simplicity of implementation and for future compatibility with RFC 1560 (cf #31783).
r? @nikomatsakis
We don't want to render default item docs but previously
`doctraititem` naively delegated to the trait definition in those
cases.
Updated tests to also check that this doesn't strip default item
docs from the trait definition.
Like the comment on `Context` explains, `Context` is supposed to be
lightweight, so we're putting everything that's immutable after
creation of the Context behind an `Arc<SharedContext>`.
Instead of hardcoding knowledge about the strip-private pass into the
rendering process we represent (some) stripped items as `ItemEnum::StrippedItem`.
Rustdoc will, for example, generate redirect pages for public items
contained in private modules which have been re-exported to somewhere
externally reachable - this will now not only work for the `strip-private`
pass, but for other passes as well, such as the `strip-hidden` pass.
Adding these "known" values to the table of used ids is only required
when embedding markdown into a rustdoc html page and may yield
unexpected results when rendering a standalone `*.md` file.
