Auto merge of #148597 - yotamofek:pr/search_index/crate_path_ref_map, r=<try>

Allow looking up "crate paths" in map without allocating a vector

Author: rust-bors[bot]

Cargo.lock

@@ -4793,11 +4793,13 @@ dependencies = [
  "askama",
  "base64",
  "expect-test",
+ "hashbrown",
  "indexmap",
  "itertools",
  "minifier",
  "pulldown-cmark-escape",
  "regex",
+ "rustc-hash 2.1.1",
  "rustdoc-json-types",
  "serde",
  "serde_json",

src/librustdoc/Cargo.toml

@@ -12,11 +12,13 @@ path = "lib.rs"
 arrayvec = { version = "0.7", default-features = false }
 askama = { version = "0.14", default-features = false, features = ["alloc", "config", "derive"] }
 base64 = "0.21.7"
+hashbrown = "0.15"
 indexmap = { version = "2", features = ["serde"] }
 itertools = "0.12"
 minifier = { version = "0.3.5", default-features = false }
 pulldown-cmark-escape = { version = "0.11.0", features = ["simd"] }
 regex = "1"
+rustc-hash = "2.1.1"
 rustdoc-json-types = { path = "../rustdoc-json-types" }
 serde = { version = "1.0", features = ["derive"] }
 serde_json = "1.0"

src/librustdoc/html/render/search_index.rs

@@ -2,15 +2,18 @@ pub(crate) mod encode;
 mod serde;
 
 use std::collections::BTreeSet;
-use std::collections::hash_map::Entry;
+use std::hash::Hash;
 use std::path::Path;
 
 use ::serde::de::{self, Deserializer, Error as _};
 use ::serde::ser::{SerializeSeq, Serializer};
 use ::serde::{Deserialize, Serialize};
+use hashbrown::hash_map::EntryRef;
+use hashbrown::{Equivalent, HashMap};
 use rustc_ast::join_path_syms;
 use rustc_data_structures::fx::{FxHashMap, FxHashSet, FxIndexMap};
 use rustc_data_structures::thin_vec::ThinVec;
+use rustc_hash::FxBuildHasher;
 use rustc_hir::attrs::AttributeKind;
 use rustc_hir::find_attr;
 use rustc_middle::ty::TyCtxt;
@@ -53,7 +56,46 @@ pub(crate) struct SerializedSearchIndex {
  generic_inverted_index: Vec<Vec<Vec<u32>>>,
  // generated in-memory backref cache
  #[serde(skip)]
- crate_paths_index: FxHashMap<(ItemType, Vec<Symbol>), usize>,
+ crate_paths_index: HashMap<CratePath, usize, FxBuildHasher>,
+}
+
+#[derive(Clone, Debug, PartialEq, Eq)]
+struct CratePath(ItemType, Vec<Symbol>);
+
+impl Hash for CratePath {
+ fn hash<H: std::hash::Hasher>(&self, state: &mut H) {
+ self.0.hash(state);
+ state.write_length_prefix(self.1.len());
+ for sym in &self.1 {
+ sym.hash(state);
+ }
+ }
+}
+
+struct CratePathRef<'sym>(ItemType, &'sym [&'sym [Symbol]]);
+
+impl<'a, 'sym> From<&'a CratePathRef<'sym>> for CratePath {
+ fn from(value: &'a CratePathRef<'sym>) -> Self {
+ Self(value.0, value.1.iter().copied().flatten().copied().collect())
+ }
+}
+
+impl<'sym> Equivalent<CratePath> for CratePathRef<'sym> {
+ fn equivalent(&self, key: &CratePath) -> bool {
+ self.0 == key.0 && self.1.iter().copied().flatten().eq(&key.1)
+ }
+}
+
+impl<'sym> Hash for CratePathRef<'sym> {
+ fn hash<H: std::hash::Hasher>(&self, state: &mut H) {
+ self.0.hash(state);
+ let iter = self.1.iter().copied().flatten();
+ let len = iter.clone().count();
+ state.write_length_prefix(len);
+ for sym in iter {
+ sym.hash(state);
+ }
+ }
 }
 
 impl SerializedSearchIndex {
@@ -176,7 +218,7 @@ impl SerializedSearchIndex {
  // generic_inverted_index is not the same length as other columns,
  // because it's actually a completely different set of objects
 
- let mut crate_paths_index: FxHashMap<(ItemType, Vec<Symbol>), usize> = FxHashMap::default();
+ let mut crate_paths_index = HashMap::default();
  for (i, (name, path_data)) in names.iter().zip(path_data.iter()).enumerate() {
  if let Some(path_data) = path_data {
  let full_path = if path_data.module_path.is_empty() {
@@ -186,7 +228,7 @@ impl SerializedSearchIndex {
  full_path.push(Symbol::intern(name));
  full_path
  };
- crate_paths_index.insert((path_data.ty, full_path), i);
+ crate_paths_index.insert(CratePath(path_data.ty, full_path), i);
  }
  }
 
@@ -216,15 +258,10 @@ impl SerializedSearchIndex {
  assert_eq!(self.names.len(), self.path_data.len());
  if let Some(path_data) = &path_data
  && let name = Symbol::intern(&name)
- && let fqp = if path_data.module_path.is_empty() {
- vec![name]
- } else {
- let mut v = path_data.module_path.clone();
- v.push(name);
- v
- }
- && let Some(&other_path) = self.crate_paths_index.get(&(path_data.ty, fqp))
- && self.path_data.get(other_path).map_or(false, Option::is_some)
+ && let Some(&other_path) = self
+ .crate_paths_index
+ .get(&CratePathRef(path_data.ty, &[&path_data.module_path, &[name]]))
+ && let Some(Some(_)) = self.path_data.get(other_path)
  {
  self.path_data.push(None);
  } else {
@@ -246,20 +283,22 @@ impl SerializedSearchIndex {
  ///
  /// The returned ID can be used to attach more data to the search result.
  fn add_entry(&mut self, name: Symbol, entry_data: EntryData, desc: String) -> usize {
- let fqp = if let Some(module_path_index) = entry_data.module_path {
- let mut fqp = self.path_data[module_path_index].as_ref().unwrap().module_path.clone();
- fqp.push(Symbol::intern(&self.names[module_path_index]));
- fqp.push(name);
- fqp
+ let interned;
+ let extra: [Symbol; 2];
+ let blah: &[Symbol] = &[name];
+ let fqp: &[&[Symbol]] = if let Some(module_path_index) = entry_data.module_path {
+ interned = Symbol::intern(&self.names[module_path_index]);
+ extra = [interned, name];
+ &[&self.path_data[module_path_index].as_ref().unwrap().module_path[..], &extra]
  } else {
- vec![name]
+ &[blah]
  };
  // If a path with the same name already exists, but no entry does,
  // we can fill in the entry without having to allocate a new row ID.
  //
  // Because paths and entries both share the same index, using the same
  // ID saves space by making the tree smaller.
- if let Some(&other_path) = self.crate_paths_index.get(&(entry_data.ty, fqp))
+ if let Some(&other_path) = self.crate_paths_index.get(&CratePathRef(entry_data.ty, fqp))
  && self.entry_data[other_path].is_none()
  && self.descs[other_path].is_empty()
  {
@@ -282,9 +321,9 @@ impl SerializedSearchIndex {
 
  fn get_id_by_module_path(&mut self, path: &[Symbol]) -> usize {
  let ty = if path.len() == 1 { ItemType::ExternCrate } else { ItemType::Module };
- match self.crate_paths_index.entry((ty, path.to_vec())) {
- Entry::Occupied(index) => *index.get(),
- Entry::Vacant(slot) => {
+ match self.crate_paths_index.entry_ref(&CratePathRef(ty, &[path])) {
+ EntryRef::Occupied(index) => *index.get(),
+ EntryRef::Vacant(slot) => {
  slot.insert(self.path_data.len());
  let (name, module_path) = path.split_last().unwrap();
  self.push_path(
@@ -301,16 +340,18 @@ impl SerializedSearchIndex {
  let mut skips = FxHashSet::default();
  for (other_pathid, other_path_data) in other.path_data.iter().enumerate() {
  if let Some(other_path_data) = other_path_data {
- let mut fqp = other_path_data.module_path.clone();
  let name = Symbol::intern(&other.names[other_pathid]);
- fqp.push(name);
+ let fqp = [&other_path_data.module_path[..], &[name]];
  let self_pathid = other_entryid_offset + other_pathid;
- let self_pathid = match self.crate_paths_index.entry((other_path_data.ty, fqp)) {
- Entry::Vacant(slot) => {
+ let self_pathid = match self
+ .crate_paths_index
+ .entry_ref(&CratePathRef(other_path_data.ty, &fqp))
+ {
+ EntryRef::Vacant(slot) => {
  slot.insert(self_pathid);
  self_pathid
  }
- Entry::Occupied(existing_entryid) => {
+ EntryRef::Occupied(existing_entryid) => {
  skips.insert(other_pathid);
  let self_pathid = *existing_entryid.get();
  let new_type_data = match (
@@ -1292,9 +1333,9 @@ pub(crate) fn build_index(
  let crate_doc =
  short_markdown_summary(&krate.module.doc_value(), &krate.module.link_names(cache));
  let crate_idx = {
- let crate_path = (ItemType::ExternCrate, vec![crate_name]);
- match serialized_index.crate_paths_index.entry(crate_path) {
- Entry::Occupied(index) => {
+ let crate_path = CratePathRef(ItemType::ExternCrate, &[&[crate_name]]);
+ match serialized_index.crate_paths_index.entry_ref(&crate_path) {
+ EntryRef::Occupied(index) => {
  let index = *index.get();
  serialized_index.descs[index] = crate_doc;
  for type_data in serialized_index.type_data.iter_mut() {
@@ -1343,7 +1384,7 @@ pub(crate) fn build_index(
  }
  index
  }
- Entry::Vacant(slot) => {
+ EntryRef::Vacant(slot) => {
  let krate = serialized_index.names.len();
  slot.insert(krate);
  serialized_index.push(
@@ -1384,9 +1425,12 @@ pub(crate) fn build_index(
  .or_else(|| check_external.then(|| cache.external_paths.get(&defid)).flatten())
  .map(|&(ref fqp, ty)| {
  let pathid = serialized_index.names.len();
- match serialized_index.crate_paths_index.entry((ty, fqp.clone())) {
- Entry::Occupied(entry) => *entry.get(),
- Entry::Vacant(entry) => {
+ match serialized_index
+ .crate_paths_index
+ .entry_ref(&CratePathRef(ty, &[&&fqp[..]]))
+ {
+ EntryRef::Occupied(entry) => *entry.get(),
+ EntryRef::Vacant(entry) => {
  entry.insert(pathid);
  let (name, path) = fqp.split_last().unwrap();
  serialized_index.push_path(
@@ -1533,46 +1577,47 @@ pub(crate) fn build_index(
  used_in_function_signature: &mut BTreeSet<isize>,
  ) -> RenderTypeId {
  let pathid = serialized_index.names.len();
- let pathid = match serialized_index.crate_paths_index.entry((ty, path.to_vec())) {
- Entry::Occupied(entry) => {
- let id = *entry.get();
- if serialized_index.type_data[id].as_mut().is_none() {
- serialized_index.type_data[id] = Some(TypeData {
- search_unbox,
- inverted_function_inputs_index: Vec::new(),
- inverted_function_output_index: Vec::new(),
- });
- } else if search_unbox {
- serialized_index.type_data[id].as_mut().unwrap().search_unbox = true;
+ let pathid =
+ match serialized_index.crate_paths_index.entry_ref(&CratePathRef(ty, &[path])) {
+ EntryRef::Occupied(entry) => {
+ let id = *entry.get();
+ if serialized_index.type_data[id].as_mut().is_none() {
+ serialized_index.type_data[id] = Some(TypeData {
+ search_unbox,
+ inverted_function_inputs_index: Vec::new(),
+ inverted_function_output_index: Vec::new(),
+ });
+ } else if search_unbox {
+ serialized_index.type_data[id].as_mut().unwrap().search_unbox = true;
+ }
+ id
  }
- id
- }
- Entry::Vacant(entry) => {
- entry.insert(pathid);
- let (name, path) = path.split_last().unwrap();
- serialized_index.push_type(
- name.to_string(),
- PathData {
- ty,
- module_path: path.to_vec(),
- exact_module_path: if let Some(exact_path) = exact_path
- && let Some((name2, exact_path)) = exact_path.split_last()
- && name == name2
- {
- Some(exact_path.to_vec())
- } else {
- None
+ EntryRef::Vacant(entry) => {
+ entry.insert(pathid);
+ let (name, path) = path.split_last().unwrap();
+ serialized_index.push_type(
+ name.to_string(),
+ PathData {
+ ty,
+ module_path: path.to_vec(),
+ exact_module_path: if let Some(exact_path) = exact_path
+ && let Some((name2, exact_path)) = exact_path.split_last()
+ && name == name2
+ {
+ Some(exact_path.to_vec())
+ } else {
+ None
+ },
  },
- },
- TypeData {
- inverted_function_inputs_index: Vec::new(),
- inverted_function_output_index: Vec::new(),
- search_unbox,
- },
- );
- pathid
- }
- };
+ TypeData {
+ inverted_function_inputs_index: Vec::new(),
+ inverted_function_output_index: Vec::new(),
+ search_unbox,
+ },
+ );
+ pathid
+ }
+ };
  used_in_function_signature.insert(isize::try_from(pathid).unwrap());
  RenderTypeId::Index(isize::try_from(pathid).unwrap())
  }

src/librustdoc/lib.rs

@@ -11,6 +11,7 @@
 #![feature(box_patterns)]
 #![feature(file_buffered)]
 #![feature(formatting_options)]
+#![feature(hasher_prefixfree_extras)]
 #![feature(if_let_guard)]
 #![feature(iter_advance_by)]
 #![feature(iter_intersperse)]