The hash changes are based on some tests with `arti` and various
specific queries, aimed at reducing the false positive rate.
Sorting the query elements so that generics always come first is
instead aimed at reducing the number of Map operations on mgens,
assuming if the bloom filter does find a false positive, it'll
be able to reject the row without having to track a mapping.
- https://hur.st/bloomfilter/?n=3&p=&m=96&k=6
Different functions have different amounts of inputs, and
unification isn't very slow anyway, so figuring out a single
ideal number of hash functions is nasty, but 6 keeps things
low even up to 10 inputs.
- https://web.archive.org/web/20210927123933/https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.72.2442&rep=rep1&type=pdf
This is the `h1` and `h2`, both derived from `h0`.
This commit adds ranking and quick filtering to type-based search,
improving performance and having it order results based on their
type signatures.
Motivation
----------
If I write a query like `str -> String`, a lot of functions come up.
That's to be expected, but `String::from_str` should come up on top, and
it doesn't right now. This is because the sorting algorithm is based
on the functions name, and doesn't consider the type signature at all.
`slice::join` even comes up above it!
To fix this, the sorting should take into account the function's
signature, and the closer match should come up on top.
Guide-level description
-----------------------
When searching by type signature, types with a "closer" match will
show up above types that match less precisely.
Reference-level explanation
---------------------------
Functions signature search works in three major phases:
* A compact "fingerprint," based on the [bloom filter] technique, is used to
check for matches and to estimate the distance. It sometimes has false
positive matches, but it also operates on 128 bit contiguous memory and
requires no backtracking, so it performs a lot better than real
unification.
The fingerprint represents the set of items in the type signature, but it
does not represent nesting, and it ignores when the same item appears more
than once.
The result is rejected if any query bits are absent in the function, or
if the distance is higher than the current maximum and 200
results have already been found.
* The second step performs unification. This is where nesting and true bag
semantics are taken into account, and it has no false positives. It uses a
recursive, backtracking algorithm.
The result is rejected if any query elements are absent in the function.
[bloom filter]: https://en.wikipedia.org/wiki/Bloom_filter
Drawbacks
---------
This makes the code bigger.
More than that, this design is a subtle trade-off. It makes the cases I've
tested against measurably faster, but it's not clear how well this extends
to other crates with potentially more functions and fewer types.
The more complex things get, the more important it is to gather a good set
of data to test with (this is arguably more important than the actual
benchmarking ifrastructure right now).
Rationale and alternatives
--------------------------
Throwing a bloom filter in front makes it faster.
More than that, it tries to take a tactic where the system can not only check
for potential matches, but also gets an accurate distance function without
needing to do unification. That way it can skip unification even on items
that have the needed elems, as long as they have more items than the
currently found maximum.
If I didn't want to be able to cheaply do set operations on the fingerprint,
a [cuckoo filter] is supposed to have better performance.
But the nice bit-banging set intersection doesn't work AFAIK.
I also looked into [minhashing], but since it's actually an unbiased
estimate of the similarity coefficient, I'm not sure how it could be used
to skip unification (I wouldn't know if the estimate was too low or
too high).
This function actually uses the number of distinct items as its
"distance function."
This should give the same results that it would have gotten from a Jaccard
Distance $1-\frac{|F\cap{}Q|}{|F\cup{}Q|}$, while being cheaper to compute.
This is because:
* The function $F$ must be a superset of the query $Q$, so their union is
just $F$ and the intersection is $Q$ and it can be reduced to
$1-\frac{|Q|}{|F|}.
* There are no magic thresholds. These values are only being used to
compare against each other while sorting (and, if 200 results are found,
to compare with the maximum match). This means we only care if one value
is bigger than the other, not what it's actual value is, and since $Q$ is
the same for everything, it can be safely left out, reducing the formula
to $1-\frac{1}{|F|} = \frac{|F|}{|F|}-\frac{1}{|F|} = |F|-1$. And, since
the values are only being compared with each other, $|F|$ is fine.
Prior art
---------
This is significantly different from how Hoogle does it.
It doesn't account for order, and it has no special account for nesting,
though `Box<t>` is still two items, while `t` is only one.
This should give the same results that it would have gotten from a Jaccard
Distance $1-\frac{|A\cap{}B|}{|A\cup{}B|}$, while being cheaper to compute.
Unresolved questions
--------------------
`[]` and `()`, the slice/array and tuple/union operators, are ignored while
building the signature for the query. This is because they match more than
one thing, making them ambiguous. Unfortunately, this also makes them
a performance cliff. Is this likely to be a problem?
Right now, the system just stashes the type distance into the
same field that levenshtein distance normally goes in. This means exact
query matches show up on top (for example, if you have a function like
`fn nothing(a: Nothing, b: i32)`, then searching for `nothing` will show it
on top even if there's another function with `fn bar(x: Nothing)` that's
technically a closer match in type signature.
Future possibilities
--------------------
It should be possible to adopt more sorting criteria to act as a tie breaker,
which could be determined during unification.
[cuckoo filter]: https://en.wikipedia.org/wiki/Cuckoo_filter
[minhashing]: https://en.wikipedia.org/wiki/MinHash
This function dates back to 9a45c9d7c6 and
seems to have been made obsolete when `addIntoResult` grew the ability to
check the levenshtein distance matching with commit
ba824ec52b.
Clean up variables in `search.js`
While reviewing https://github.com/rust-lang/rust/pull/118402, I saw a few small clean ups that were needed, mostly about variables creation.
r? ```@notriddle```
The `c === "="` was redundant when `isSeparatorCharacter` already
checks that.
The function `isStopCharacter` and `isEndCharacter` functions
did exactly the same thing and have synonymous names.
There doesn't seem much point in having both.
rustdoc: remove unused parameter `reversed` from onEach(Lazy)
This feature was added in edec5807ac to support JavaScript-based toggles that were later replaced with HTML `<details>`.
rustdoc: `div.where` instead of fmt-newline class
This is about equally readable, a lot more terse, and stops special-casing functions and methods.
```console
$ du -hs doc-old/ doc-new/
671M doc-old/
670M doc-new/
```
This is about equally readable, a lot more terse, and stops
special-casing functions and methods.
```console
$ du -hs doc-old/ doc-new/
671M doc-old/
670M doc-new/
```
This restriction made sense back when spaces separated function
parameters, but now that they separate path components, there's
no real ambiguity any more.
Additionally, the Rust language allows it.
The search sorting code already sorts by item type discriminant,
putting things with smaller discriminants first. There was
also a special case for sorting keywords and primitives earlier,
and this commit removes it by giving them lower discriminants.
The sorting code has another criteria where items with descriptions
appear earlier than items without, and that criteria has higher
priority than the item type. This shouldn't matter, though,
because primitives and keywords normally only appear in the
standard library, and it always gives them descriptions.
This computes the same result with less code by computing many of
the old checks at once:
* It won't enter the loop if clength > length, because then the
result of length - clength will be negative and the
loop conditional will fail.
* i + clength will never be greater than length, because it
starts out as i = length - clength, implying that i + clength
equals length, and it only goes down from there.
* The aborted variable is replaced with control flow.
This is significantly faster, because
- It allows the one-element fast path to kick in on multi-
element queries.
- It constructs intermediate data structures more lazily
than the old system did.
It's measurably faster than the old algo even without the fast path, but
that fast path still helps significantly.
Short queries, in addition to being common, are also the base
case for a lot of more complicated queries. We can avoid
most of the backtracking data structures, and use simple
recursive matching instead, by special casing them.
Profile output:
https://notriddle.com/rustdoc-html-demo-5/profile-3/index.html
This didn't show up in our local tests, because the problem is actually
caused by docs.rs rewritten HTML (which relocates the classes that this
code looked for from the body tag to a child div).
Fixes#117290
rustdoc: use JS to inline target type impl docs into alias
Preview docs:
- https://notriddle.com/rustdoc-html-demo-5/js-trait-alias/std/io/type.Result.html
- https://notriddle.com/rustdoc-html-demo-5/js-trait-alias-compiler/rustc_middle/ty/type.PolyTraitRef.html
This pull request also includes a bug fix for trait alias inlining across crates. This means more documentation is generated, and is why ripgrep runs slower (it's a thin wrapper on top of the `grep` crate, so 5% of its docs are now the Result type).
- Before, built with rustdoc 1.75.0-nightly (aa1a71e9e 2023-10-26), Result type alias method docs are missing: http://notriddle.com/rustdoc-html-demo-5/ripgrep-js-nightly/rg/type.Result.html
- After, built with this branch, all the methods on Result are shown: http://notriddle.com/rustdoc-html-demo-5/ripgrep-js-trait-alias/rg/type.Result.html
*Review note: This is mostly just reverting https://github.com/rust-lang/rust/pull/115201. The last commit has the new work in it.*
Fixes#115718
This is an attempt to balance three problems, each of which would
be violated by a simpler implementation:
- A type alias should show all the `impl` blocks for the target
type, and vice versa, if they're applicable. If nothing was
done, and rustdoc continues to match them up in HIR, this
would not work.
- Copying the target type's docs into its aliases' HTML pages
directly causes far too much redundant HTML text to be generated
when a crate has large numbers of methods and large numbers
of type aliases.
- Using JavaScript exclusively for type alias impl docs would
be a functional regression, and could make some docs very hard
to find for non-JS readers.
- Making sure that only applicable docs are show in the
resulting page requires a type checkers. Do not reimplement
the type checker in JavaScript.
So, to make it work, rustdoc stashes these type-alias-inlined docs
in a JSONP "database-lite". The file is generated in `write_shared.rs`,
included in a `<script>` tag added in `print_item.rs`, and `main.js`
takes care of patching the additional docs into the DOM.
The format of `trait.impl` and `type.impl` JS files are superficially
similar. Each line, except the JSONP wrapper itself, belongs to a crate,
and they are otherwise separate (rustdoc should be idempotent). The
"meat" of the file is HTML strings, so the frontend code is very simple.
Links are relative to the doc root, though, so the frontend needs to fix
that up, and inlined docs can reuse these files.
However, there are a few differences, caused by the sophisticated
features that type aliases have. Consider this crate graph:
```text
---------------------------------
| crate A: struct Foo<T> |
| type Bar = Foo<i32> |
| impl X for Foo<i8> |
| impl Y for Foo<i32> |
---------------------------------
|
----------------------------------
| crate B: type Baz = A::Foo<i8> |
| type Xyy = A::Foo<i8> |
| impl Z for Xyy |
----------------------------------
```
The type.impl/A/struct.Foo.js JS file has a structure kinda like this:
```js
JSONP({
"A": [["impl Y for Foo<i32>", "Y", "A::Bar"]],
"B": [["impl X for Foo<i8>", "X", "B::Baz", "B::Xyy"], ["impl Z for Xyy", "Z", "B::Baz"]],
});
```
When the type.impl file is loaded, only the current crate's docs are
actually used. The main reason to bundle them together is that there's
enough duplication in them for DEFLATE to remove the redundancy.
The contents of a crate are a list of impl blocks, themselves
represented as lists. The first item in the sublist is the HTML block,
the second item is the name of the trait (which goes in the sidebar),
and all others are the names of type aliases that successfully match.
This way:
- There's no need to generate these files for types that have no aliases
in the current crate. If a dependent crate makes a type alias, it'll
take care of generating its own docs.
- There's no need to reimplement parts of the type checker in
JavaScript. The Rust backend does the checking, and includes its
results in the file.
- Docs defined directly on the type alias are dropped directly in the
HTML by `render_assoc_items`, and are accessible without JavaScript.
The JSONP file will not list impl items that are known to be part
of the main HTML file already.
[JSONP]: https://en.wikipedia.org/wiki/JSONP
This is an attempt to balance three problems, each of which would
be violated by a simpler implementation:
- A type alias should show all the `impl` blocks for the target
type, and vice versa, if they're applicable. If nothing was
done, and rustdoc continues to match them up in HIR, this
would not work.
- Copying the target type's docs into its aliases' HTML pages
directly causes far too much redundant HTML text to be generated
when a crate has large numbers of methods and large numbers
of type aliases.
- Using JavaScript exclusively for type alias impl docs would
be a functional regression, and could make some docs very hard
to find for non-JS readers.
- Making sure that only applicable docs are show in the
resulting page requires a type checkers. Do not reimplement
the type checker in JavaScript.
So, to make it work, rustdoc stashes these type-alias-inlined docs
in a JSONP "database-lite". The file is generated in `write_shared.rs`,
included in a `<script>` tag added in `print_item.rs`, and `main.js`
takes care of patching the additional docs into the DOM.
The format of `trait.impl` and `type.impl` JS files are superficially
similar. Each line, except the JSONP wrapper itself, belongs to a crate,
and they are otherwise separate (rustdoc should be idempotent). The
"meat" of the file is HTML strings, so the frontend code is very simple.
Links are relative to the doc root, though, so the frontend needs to fix
that up, and inlined docs can reuse these files.
However, there are a few differences, caused by the sophisticated
features that type aliases have. Consider this crate graph:
```text
---------------------------------
| crate A: struct Foo<T> |
| type Bar = Foo<i32> |
| impl X for Foo<i8> |
| impl Y for Foo<i32> |
---------------------------------
|
----------------------------------
| crate B: type Baz = A::Foo<i8> |
| type Xyy = A::Foo<i8> |
| impl Z for Xyy |
----------------------------------
```
The type.impl/A/struct.Foo.js JS file has a structure kinda like this:
```js
JSONP({
"A": [["impl Y for Foo<i32>", "Y", "A::Bar"]],
"B": [["impl X for Foo<i8>", "X", "B::Baz", "B::Xyy"], ["impl Z for Xyy", "Z", "B::Baz"]],
});
```
When the type.impl file is loaded, only the current crate's docs are
actually used. The main reason to bundle them together is that there's
enough duplication in them for DEFLATE to remove the redundancy.
The contents of a crate are a list of impl blocks, themselves
represented as lists. The first item in the sublist is the HTML block,
the second item is the name of the trait (which goes in the sidebar),
and all others are the names of type aliases that successfully match.
This way:
- There's no need to generate these files for types that have no aliases
in the current crate. If a dependent crate makes a type alias, it'll
take care of generating its own docs.
- There's no need to reimplement parts of the type checker in
JavaScript. The Rust backend does the checking, and includes its
results in the file.
- Docs defined directly on the type alias are dropped directly in the
HTML by `render_assoc_items`, and are accessible without JavaScript.
The JSONP file will not list impl items that are known to be part
of the main HTML file already.
[JSONP]: https://en.wikipedia.org/wiki/JSONP
rustdoc-search: add impl disambiguator to duplicate assoc items
Preview (to see the difference, click the link and pay attention to the specific function that comes up):
| Before | After |
|--|--|
| [`simd<i64>, simd<i64> -> simd<i64>`](https://doc.rust-lang.org/nightly/std/?search=simd%3Ci64%3E%2C%20simd%3Ci64%3E%20-%3E%20simd%3Ci64%3E) | [`simd<i64>, simd<i64> -> simd<i64>`](https://notriddle.com/rustdoc-demo-html-3/impl-disambiguate-search/std/index.html?search=simd%3Ci64%3E%2C%20simd%3Ci64%3E%20-%3E%20simd%3Ci64%3E) |
| [`cow, vec -> bool`](https://doc.rust-lang.org/nightly/std/?search=cow%2C%20vec%20-%3E%20bool) | [`cow, vec -> bool`](https://notriddle.com/rustdoc-demo-html-3/impl-disambiguate-search/std/index.html?search=cow%2C%20vec%20-%3E%20bool)
Helps with #90929
This changes the search results, specifically, when there's more than one impl with an associated item with the same name. For example, the search queries `simd<i8> -> simd<i8>` and `simd<i64> -> simd<i64>` don't link to the same function, but most of the functions have the same names.
This change should probably be FCP-ed, especially since it adds a new anchor link format for `main.js` to handle, so that URLs like `struct.Vec.html#impl-AsMut<[T]>-for-Vec<T,+A>/method.as_mut` redirect to `struct.Vec.html#method.as_mut-2`. It's a strange design, but there are a few reasons for it:
* I'd like to avoid making the HTML bigger. Obviously, fixing this bug is going to add at least a little more data to the search index, but adding more HTML penalises viewers for the benefit of searchers.
* Breaking `struct.Vec.html#method.len` would also be a disappointment.
On the other hand:
* The path-style anchors might be less prone to link rot than the numbered anchors. It's definitely less likely to have URLs that appear to "work", but silently point at the wrong thing.
* This commit arranges the path-style anchor to redirect to the numbered anchor. Nothing stops rustdoc from doing the opposite, making path-style anchors the default and redirecting the "legacy" numbered ones.
### The bug
On the "Before" links, this example search calls for `i64`:
But if I click any of the results, I get `f64` instead.
The PR fixes this problem by adding enough information to the search result `href` to disambiguate methods with different types but the same name.
More detailed description of the problem at:
https://github.com/rust-lang/rust/pull/109422#issuecomment-1491089293
> When a struct/enum/union has multiple impls with different type parameters, it can have multiple methods that have the same name, but which are on different impls. Besides Simd, [Any](https://doc.rust-lang.org/nightly/std/any/trait.Any.html?search=any%3A%3Adowncast) also demonstrates this pattern. It has three methods named `downcast`, on three different impls.
>
> When that happens, it presents a challenge in linking to the method. Normally we link like `#method.foo`. When there are multiple `foo`, we number them like `#method.foo`, `#method.foo-1`, `#method.foo-2`, etc.
>
> It also presents a challenge for our search code. Currently we store all the variants in the index, but don’t have any way to generate unambiguous URLs in the results page, or to distinguish them in the SERP.
>
> To fix this, we need three things:
>
> 1. A fragment format that fully specifies the impl type parameters when needed to disambiguate (`#impl-SimdOrd-for-Simd<i64,+LANES>/method.simd_max`)
> 2. A search index that stores methods with enough information to disambiguate the impl they were on.
> 3. A search results interface that can display multiple methods on the same type with the same name, when appropriate OR a disambiguation landing section on item pages?
>
> For reviewers: it can be hard to see the new fragment format in action since it immediately gets rewritten to the numbered form.
This commit makes three changes for consistency and readability:
- It shows the sibling navigation on module pages. It's weird
that it didn't work before, and is inconsistent with everything
else (even Crates have sibling navigation with other Crates).
- It hides the "In [parent]" header if it's the same as the
current crate, and if there's no other header between them.
We need to keep it on modules and types, since they have
their own header and data between them, and we don't want
to show siblings under a header implying that they're children.
- It adds a margin to deal with the headers butting directly into
the branding lockup.
This should result in a layout for the actual standard library,
when built on CI, that looks like this:
_____
/ \ std
| R | 1.74.0-nightly
\_____/
(203c57dbe 2023-09-17)
Having the whole version as one string caused it to flex wrap,
because the sidebar isn't wide enough to fit the whole thing.
This commit changes the layout to something a bit less "look at my logo!!!111"
gigantic, and makes it clearer where clicking the logo will actually take you.
It also means the crate name is persistently at the top of the sidebar, even
when in a sub-item page, and clicking that name takes you back to the root.
| | Short crate name | Long crate name |
|---------|------------------|-----------------|
| Root | ![short-root] | ![long-root]
| Subpage | ![short-subpage] | ![long-subpage]
[short-root]: https://github.com/rust-lang/rust/assets/1593513/fe2ce102-d4b8-44e6-9f7b-68636a907f56
[short-subpage]: https://github.com/rust-lang/rust/assets/1593513/29501663-56c0-4151-b7de-d2637e167125
[long-root]: https://github.com/rust-lang/rust/assets/1593513/f6a385c0-b4c5-4a9c-954b-21b38de4192f
[long-subpage]: https://github.com/rust-lang/rust/assets/1593513/97ec47b4-61bf-4ebe-b461-0d2187b8c6cahttps://notriddle.com/rustdoc-html-demo-4/logo-lockup/image/index.htmlhttps://notriddle.com/rustdoc-html-demo-4/logo-lockup/crossbeam_channel/index.htmlhttps://notriddle.com/rustdoc-html-demo-4/logo-lockup/adler/struct.Adler32.htmlhttps://notriddle.com/rustdoc-html-demo-4/logo-lockup/crossbeam_channel/struct.Sender.html
This improves visual information density (the construct with the logo and
crate name is *shorter* than the logo on its own, because it's not
square) and navigation clarity (we can now see what clicking the Rust logo
does, specifically).
Compare this with the layout at [Phoenix's Hexdocs] (which is what this
proposal is closely based on), the old proposal on [Internals Discourse]
(which always says "Rust standard library" in the sidebar, but doesn't do the
side-by-side layout).
[Phoenix's Hexdocs]: https://hexdocs.pm/phoenix/1.7.7/overview.html
[Internals Discourse]: https://internals.rust-lang.org/t/poc-of-a-new-design-for-the-generated-rustdoc/11018
In newer versions of rustdoc, the crate name and version are always shown in
the sidebar, even in subpages. Clicking the crate name does the same thing
clicking the logo always did: return you to the crate root.
While this actually takes up less screen real estate than the old layout on
desktop, it takes up more HTML. It's also a bit more visually complex.
I could do what the Internals POC did and keep the vertically stacked layout
all the time, instead of doing a horizontal stack where possible. It would
take up more screen real estate, though.
This design is lifted almost verbatim from Hexdocs. It seems to work for them.
[`opentelemetry_process_propagator`], for example, has a long application name.
[`opentelemetry_process_propagator`]: https://hexdocs.pm/opentelemetry_process_propagator/OpentelemetryProcessPropagator.html
Has anyone written the rationale on why the Rust logo shows up on projects that
aren't the standard library? If we turned it off on non-standard crates by
default, it would line wrap crate names a lot less often.
Or maybe we should encourage crate authors to include their own logo more
often? It certainly helps give people a better sense of "place."
I'm not sure of anything that directly follows up this one. Plenty of other
changes could be made to improve the layout, like
* coming up with a less cluttered way to do disclosure (there's a lot of `[-]`
on the page)
* doing a better job of separating lateral navigation (vec::Vec links to
vec::IntoIter) and the table of contents (vec::Vec links to vec::Vec::new)
* giving readers more control of how much rustdoc hows them, and giving doc
authors more control of how much it generates
* better search that reduces the need to browse
But those are mostly orthogonal, not future possibilities unlocked by this change.
