Use Matchit to Resolve Per-File Settings

Cargo.toml

@@ -58,13 +58,15 @@ libcst = { version = "1.1.0", default-features = false }
 log = { version = "0.4.17" }
 lsp-server = { version = "0.7.6" }
 lsp-types = { version = "0.95.0", features = ["proposed"] }
+matchit = { version = "0.8.1" }
 memchr = { version = "2.7.1" }
 mimalloc = { version = "0.1.39" }
 natord = { version = "1.0.9" }
 notify = { version = "6.1.1" }
 num_cpus = { version = "1.16.0" }
 once_cell = { version = "1.19.0" }
 path-absolutize = { version = "3.1.1" }
+path-slash = { version = "0.2.1" }
 pathdiff = { version = "0.2.1" }
 pep440_rs = { version = "0.6.0", features = ["serde"] }
 pretty_assertions = "1.3.0"

crates/ruff_workspace/Cargo.toml

@@ -28,9 +28,11 @@ ignore = { workspace = true }
 is-macro = { workspace = true }
 itertools = { workspace = true }
 log = { workspace = true }
+matchit = { workspace = true }
 glob = { workspace = true }
 globset = { workspace = true }
 path-absolutize = { workspace = true }
+path-slash = { workspace = true }
 pep440_rs = { workspace = true, features = ["serde"] }
 regex = { workspace = true }
 rustc-hash = { workspace = true }

crates/ruff_workspace/src/resolver.rs

@@ -2,7 +2,6 @@
 //! filesystem.
 
 use std::cmp::Ordering;
-use std::collections::BTreeMap;
 use std::ffi::OsStr;
 use std::path::{Path, PathBuf};
 use std::sync::RwLock;
@@ -13,7 +12,9 @@ use globset::{Candidate, GlobSet};
 use ignore::{WalkBuilder, WalkState};
 use itertools::Itertools;
 use log::debug;
+use matchit::{InsertError, Match, Router};
 use path_absolutize::path_dedot;
+use path_slash::PathExt;
 use rustc_hash::{FxHashMap, FxHashSet};
 
 use ruff_linter::fs;
@@ -86,29 +87,32 @@ pub enum Relativity {
 }
 
 impl Relativity {
-    pub fn resolve(self, path: &Path) -> PathBuf {
+    pub fn resolve(self, path: &Path) -> &Path {
         match self {
             Relativity::Parent => path
                 .parent()
-                .expect("Expected pyproject.toml file to be in parent directory")
-                .to_path_buf(),
-            Relativity::Cwd => path_dedot::CWD.clone(),
+                .expect("Expected pyproject.toml file to be in parent directory"),
+            Relativity::Cwd => &path_dedot::CWD,
         }
     }
 }
 
 #[derive(Debug)]
 pub struct Resolver<'a> {
     pyproject_config: &'a PyprojectConfig,
-    settings: BTreeMap<PathBuf, Settings>,
+    /// All [`Settings`] that have been added to the resolver.
+    settings: Vec<Settings>,
+    /// A router from path to index into the `settings` vector.
+    router: Router<usize>,
 }
 
 impl<'a> Resolver<'a> {
     /// Create a new [`Resolver`] for the given [`PyprojectConfig`].
     pub fn new(pyproject_config: &'a PyprojectConfig) -> Self {
         Self {
             pyproject_config,
-            settings: BTreeMap::new(),
+            settings: Vec::new(),
+            router: Router::new(),
         }
     }
 
@@ -140,19 +144,31 @@ impl<'a> Resolver<'a> {
     }
 
     /// Add a resolved [`Settings`] under a given [`PathBuf`] scope.
-    fn add(&mut self, path: PathBuf, settings: Settings) {
-        self.settings.insert(path, settings);
+    fn add(&mut self, path: &Path, settings: Settings) {
+        self.settings.push(settings);
+
+        // normalize the path to use `/` separators and escape the '{' and '}' characters,
+        // which matchit uses for routing parameters
+        let path = path.to_slash_lossy().replace('{', "{{").replace('}', "}}");
+
+        match self
+            .router
+            .insert(format!("{path}/{{*filepath}}"), self.settings.len() - 1)
+        {
+            Ok(()) => {}
+            Err(InsertError::Conflict { .. }) => {}
+            Err(_) => unreachable!("file paths are escaped before being inserted in the router"),
+        }
     }
 
     /// Return the appropriate [`Settings`] for a given [`Path`].
     pub fn resolve(&self, path: &Path) -> &Settings {
         match self.pyproject_config.strategy {
             PyprojectDiscoveryStrategy::Fixed => &self.pyproject_config.settings,
             PyprojectDiscoveryStrategy::Hierarchical => self
-                .settings
-                .iter()
-                .rev()
-                .find_map(|(root, settings)| path.starts_with(root).then_some(settings))
+                .router
+                .at(path.to_slash_lossy().as_ref())
+                .map(|Match { value, .. }| &self.settings[*value])
                 .unwrap_or(&self.pyproject_config.settings),
         }
     }
@@ -196,7 +212,7 @@ impl<'a> Resolver<'a> {
 
     /// Return an iterator over the resolved [`Settings`] in this [`Resolver`].
     pub fn settings(&self) -> impl Iterator<Item = &Settings> {
-        std::iter::once(&self.pyproject_config.settings).chain(self.settings.values())
+        std::iter::once(&self.pyproject_config.settings).chain(self.settings.iter())
     }
 }
 
@@ -257,7 +273,7 @@ fn resolve_configuration(
         let options = pyproject::load_options(&path)?;
 
         let project_root = relativity.resolve(&path);
-        let configuration = Configuration::from_options(options, Some(&path), &project_root)?;
+        let configuration = Configuration::from_options(options, Some(&path), project_root)?;
 
         // If extending, continue to collect.
         next = configuration.extend.as_ref().map(|extend| {
@@ -285,14 +301,14 @@ fn resolve_configuration(
 
 /// Extract the project root (scope) and [`Settings`] from a given
 /// `pyproject.toml`.
-fn resolve_scoped_settings(
-    pyproject: &Path,
+fn resolve_scoped_settings<'a>(
+    pyproject: &'a Path,
     relativity: Relativity,
     transformer: &dyn ConfigurationTransformer,
-) -> Result<(PathBuf, Settings)> {
+) -> Result<(&'a Path, Settings)> {
     let configuration = resolve_configuration(pyproject, relativity, transformer)?;
     let project_root = relativity.resolve(pyproject);
-    let settings = configuration.into_settings(&project_root)?;
+    let settings = configuration.into_settings(project_root)?;
     Ok((project_root, settings))
 }