Use hasher to compute resolution hash (#5495)

## Summary

Addressing one TODO. This should be more efficient.

crates/cache-key/src/digest.rs

@@ -1,22 +1,34 @@
-use std::hash::Hasher;
-
 use crate::cache_key::{CacheKey, CacheKeyHasher};
+use seahash::SeaHasher;
+use std::hash::{Hash, Hasher};
 
 /// Compute a hex string hash of a `CacheKey` object.
 ///
-/// The value returned by [`digest`] should be stable across releases and platforms.
-pub fn digest<H: CacheKey>(hashable: &H) -> String {
+/// The value returned by [`cache_digest`] should be stable across releases and platforms.
+pub fn cache_digest<H: CacheKey>(hashable: &H) -> String {
+    /// Compute a u64 hash of a [`CacheKey`] object.
+    fn cache_key_u64<H: CacheKey>(hashable: &H) -> u64 {
+        let mut hasher = CacheKeyHasher::new();
+        hashable.cache_key(&mut hasher);
+        hasher.finish()
+    }
+
     to_hex(cache_key_u64(hashable))
 }
 
+/// Compute a hex string hash of a hashable object.
+pub fn hash_digest<H: Hash>(hashable: &H) -> String {
+    /// Compute a u64 hash of a hashable object.
+    fn hash_u64<H: Hash>(hashable: &H) -> u64 {
+        let mut hasher = SeaHasher::new();
+        hashable.hash(&mut hasher);
+        hasher.finish()
+    }
+
+    to_hex(hash_u64(hashable))
+}
+
 /// Convert a u64 to a hex string.
 fn to_hex(num: u64) -> String {
     hex::encode(num.to_le_bytes())
 }
-
-/// Compute a u64 hash of a [`CacheKey`] object.
-fn cache_key_u64<H: CacheKey>(hashable: &H) -> u64 {
-    let mut hasher = CacheKeyHasher::new();
-    hashable.cache_key(&mut hasher);
-    hasher.finish()
-}

crates/cache-key/src/lib.rs

@@ -1,6 +1,6 @@
 pub use cache_key::{CacheKey, CacheKeyHasher};
 pub use canonical_url::{CanonicalUrl, RepositoryUrl};
-pub use digest::digest;
+pub use digest::{cache_digest, hash_digest};
 
 mod cache_key;
 mod canonical_url;

crates/distribution-types/src/any.rs

@@ -6,6 +6,7 @@ use crate::{InstalledMetadata, InstalledVersion, Name};
 
 /// A distribution which is either installable, is a wheel in our cache or is already installed.
 #[derive(Debug, Clone)]
+#[allow(clippy::large_enum_variant)]
 pub enum LocalDist {
     Cached(CachedDist),
     Installed(InstalledDist),

crates/distribution-types/src/file.rs

@@ -19,7 +19,7 @@ pub enum FileConversionError {
 
 /// Internal analog to [`pypi_types::File`].
 #[derive(
-    Debug, Clone, Serialize, Deserialize, rkyv::Archive, rkyv::Deserialize, rkyv::Serialize,
+    Debug, Clone, Hash, Serialize, Deserialize, rkyv::Archive, rkyv::Deserialize, rkyv::Serialize,
 )]
 #[archive(check_bytes)]
 #[archive_attr(derive(Debug))]
@@ -67,7 +67,7 @@ impl File {
 
 /// While a registry file is generally a remote URL, it can also be a file if it comes from a directory flat indexes.
 #[derive(
-    Debug, Clone, Serialize, Deserialize, rkyv::Archive, rkyv::Deserialize, rkyv::Serialize,
+    Debug, Clone, Hash, Serialize, Deserialize, rkyv::Archive, rkyv::Deserialize, rkyv::Serialize,
 )]
 #[archive(check_bytes)]
 #[archive_attr(derive(Debug))]
@@ -147,13 +147,14 @@ impl Display for FileLocation {
 #[derive(
     Debug,
     Clone,
+    PartialEq,
+    Eq,
+    Hash,
     Serialize,
     Deserialize,
     rkyv::Archive,
     rkyv::Deserialize,
     rkyv::Serialize,
-    PartialEq,
-    Eq,
 )]
 #[archive(check_bytes)]
 #[archive_attr(derive(Debug))]
@@ -185,7 +186,7 @@ impl From<UrlString> for String {
     }
 }
 
-impl fmt::Display for UrlString {
+impl Display for UrlString {
     fn fmt(&self, f: &mut Formatter<'_>) -> fmt::Result {
         fmt::Display::fmt(&self.0, f)
     }

crates/distribution-types/src/installed.rs

@@ -16,7 +16,7 @@ use uv_normalize::PackageName;
 use crate::{DistributionMetadata, InstalledMetadata, InstalledVersion, Name, VersionOrUrlRef};
 
 /// A built distribution (wheel) that is installed in a virtual environment.
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 pub enum InstalledDist {
     /// The distribution was derived from a registry, like `PyPI`.
     Registry(InstalledRegistryDist),
@@ -30,14 +30,14 @@ pub enum InstalledDist {
     LegacyEditable(InstalledLegacyEditable),
 }
 
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 pub struct InstalledRegistryDist {
     pub name: PackageName,
     pub version: Version,
     pub path: PathBuf,
 }
 
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 pub struct InstalledDirectUrlDist {
     pub name: PackageName,
     pub version: Version,
@@ -47,21 +47,21 @@ pub struct InstalledDirectUrlDist {
     pub path: PathBuf,
 }
 
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 pub struct InstalledEggInfoFile {
     pub name: PackageName,
     pub version: Version,
     pub path: PathBuf,
 }
 
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 pub struct InstalledEggInfoDirectory {
     pub name: PackageName,
     pub version: Version,
     pub path: PathBuf,
 }
 
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 pub struct InstalledLegacyEditable {
     pub name: PackageName,
     pub version: Version,

crates/distribution-types/src/lib.rs

@@ -126,14 +126,14 @@ impl std::fmt::Display for InstalledVersion<'_> {
 /// Either a built distribution, a wheel, or a source distribution that exists at some location.
 ///
 /// The location can be an index, URL or path (wheel), or index, URL, path or Git repository (source distribution).
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 pub enum Dist {
     Built(BuiltDist),
     Source(SourceDist),
 }
 
 /// A wheel, with its three possible origins (index, url, path)
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 #[allow(clippy::large_enum_variant)]
 pub enum BuiltDist {
     Registry(RegistryBuiltDist),
@@ -142,7 +142,7 @@ pub enum BuiltDist {
 }
 
 /// A source distribution, with its possible origins (index, url, path, git)
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 #[allow(clippy::large_enum_variant)]
 pub enum SourceDist {
     Registry(RegistrySourceDist),
@@ -153,15 +153,15 @@ pub enum SourceDist {
 }
 
 /// A built distribution (wheel) that exists in a registry, like `PyPI`.
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 pub struct RegistryBuiltWheel {
     pub filename: WheelFilename,
     pub file: Box<File>,
     pub index: IndexUrl,
 }
 
 /// A built distribution (wheel) that exists in a registry, like `PyPI`.
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 pub struct RegistryBuiltDist {
     /// All wheels associated with this distribution. It is guaranteed
     /// that there is at least one wheel.
@@ -191,7 +191,7 @@ pub struct RegistryBuiltDist {
 }
 
 /// A built distribution (wheel) that exists at an arbitrary URL.
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 pub struct DirectUrlBuiltDist {
     /// We require that wheel urls end in the full wheel filename, e.g.
     /// `https://example.org/packages/flask-3.0.0-py3-none-any.whl`
@@ -203,7 +203,7 @@ pub struct DirectUrlBuiltDist {
 }
 
 /// A built distribution (wheel) that exists in a local directory.
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 pub struct PathBuiltDist {
     pub filename: WheelFilename,
     /// The path to the wheel.
@@ -213,7 +213,7 @@ pub struct PathBuiltDist {
 }
 
 /// A source distribution that exists in a registry, like `PyPI`.
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 pub struct RegistrySourceDist {
     pub name: PackageName,
     pub version: Version,
@@ -230,7 +230,7 @@ pub struct RegistrySourceDist {
 }
 
 /// A source distribution that exists at an arbitrary URL.
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 pub struct DirectUrlSourceDist {
     /// Unlike [`DirectUrlBuiltDist`], we can't require a full filename with a version here, people
     /// like using e.g. `foo @ https://github.com/org/repo/archive/master.zip`
@@ -244,7 +244,7 @@ pub struct DirectUrlSourceDist {
 }
 
 /// A source distribution that exists in a Git repository.
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 pub struct GitSourceDist {
     pub name: PackageName,
     /// The URL without the revision and subdirectory fragment.
@@ -256,7 +256,7 @@ pub struct GitSourceDist {
 }
 
 /// A source distribution that exists in a local archive (e.g., a `.tar.gz` file).
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 pub struct PathSourceDist {
     pub name: PackageName,
     /// The resolved, absolute path to the distribution which we use for installing.
@@ -270,7 +270,7 @@ pub struct PathSourceDist {
 }
 
 /// A source distribution that exists in a local directory.
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 pub struct DirectorySourceDist {
     pub name: PackageName,
     /// The resolved, absolute path to the distribution which we use for installing.

crates/distribution-types/src/resolution.rs

@@ -71,7 +71,7 @@ impl Resolution {
     }
 }
 
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
 pub enum ResolutionDiagnostic {
     MissingExtra {
         /// The distribution that was requested with a non-existent extra. For example,

crates/distribution-types/src/resolved.rs

@@ -12,7 +12,8 @@ use crate::{
 /// A distribution that can be used for resolution and installation.
 ///
 /// Either an already-installed distribution or a distribution that can be installed.
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Hash)]
+#[allow(clippy::large_enum_variant)]
 pub enum ResolvedDist {
     Installed(InstalledDist),
     Installable(Dist),

crates/pypi-types/src/direct_url.rs

@@ -1,4 +1,4 @@
-use std::collections::HashMap;
+use std::collections::BTreeMap;
 use std::path::PathBuf;
 
 use serde::{Deserialize, Serialize};
@@ -7,7 +7,7 @@ use url::Url;
 /// Metadata for a distribution that was installed via a direct URL.
 ///
 /// See: <https://packaging.python.org/en/latest/specifications/direct-url-data-structure/>
-#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
+#[derive(Debug, Clone, PartialEq, Eq, Hash, Serialize, Deserialize)]
 #[serde(rename_all = "snake_case", untagged)]
 pub enum DirectUrl {
     /// The direct URL is a local directory. For example:
@@ -41,23 +41,23 @@ pub enum DirectUrl {
     },
 }
 
-#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
+#[derive(Debug, Clone, PartialEq, Eq, Hash, Serialize, Deserialize)]
 #[serde(rename_all = "snake_case")]
 pub struct DirInfo {
     #[serde(skip_serializing_if = "Option::is_none")]
     pub editable: Option<bool>,
 }
 
-#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
+#[derive(Debug, Clone, PartialEq, Eq, Hash, Serialize, Deserialize)]
 #[serde(rename_all = "snake_case")]
 pub struct ArchiveInfo {
     #[serde(skip_serializing_if = "Option::is_none")]
     pub hash: Option<String>,
     #[serde(skip_serializing_if = "Option::is_none")]
-    pub hashes: Option<HashMap<String, String>>,
+    pub hashes: Option<BTreeMap<String, String>>,
 }
 
-#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
+#[derive(Debug, Clone, PartialEq, Eq, Hash, Serialize, Deserialize)]
 #[serde(rename_all = "snake_case")]
 pub struct VcsInfo {
     pub vcs: VcsKind,
@@ -67,7 +67,7 @@ pub struct VcsInfo {
     pub requested_revision: Option<String>,
 }
 
-#[derive(Debug, Copy, Clone, PartialEq, Eq, Serialize, Deserialize)]
+#[derive(Debug, Copy, Clone, PartialEq, Eq, Hash, Serialize, Deserialize)]
 #[serde(rename_all = "snake_case")]
 pub enum VcsKind {
     Git,

crates/pypi-types/src/simple_json.rs

@@ -92,6 +92,7 @@ impl CoreMetadata {
     Clone,
     PartialEq,
     Eq,
+    Hash,
     Serialize,
     Deserialize,
     rkyv::Archive,

crates/uv-cache/src/wheel.rs

@@ -2,7 +2,7 @@ use std::path::{Path, PathBuf};
 
 use url::Url;
 
-use cache_key::{digest, CanonicalUrl};
+use cache_key::{cache_digest, CanonicalUrl};
 use distribution_types::IndexUrl;
 
 /// Cache wheels and their metadata, both from remote wheels and built from source distributions.
@@ -30,19 +30,19 @@ impl<'a> WheelCache<'a> {
             WheelCache::Index(IndexUrl::Pypi(_)) => WheelCacheKind::Pypi.root(),
             WheelCache::Index(url) => WheelCacheKind::Index
                 .root()
-                .join(digest(&CanonicalUrl::new(url))),
+                .join(cache_digest(&CanonicalUrl::new(url))),
             WheelCache::Url(url) => WheelCacheKind::Url
                 .root()
-                .join(digest(&CanonicalUrl::new(url))),
+                .join(cache_digest(&CanonicalUrl::new(url))),
             WheelCache::Path(url) => WheelCacheKind::Path
                 .root()
-                .join(digest(&CanonicalUrl::new(url))),
+                .join(cache_digest(&CanonicalUrl::new(url))),
             WheelCache::Editable(url) => WheelCacheKind::Editable
                 .root()
-                .join(digest(&CanonicalUrl::new(url))),
+                .join(cache_digest(&CanonicalUrl::new(url))),
             WheelCache::Git(url, sha) => WheelCacheKind::Git
                 .root()
-                .join(digest(&CanonicalUrl::new(url)))
+                .join(cache_digest(&CanonicalUrl::new(url)))
                 .join(sha),
         }
     }

crates/uv-client/src/flat_index.rs

@@ -141,7 +141,7 @@ impl<'a> FlatIndexClient<'a> {
         let cache_entry = self.cache.entry(
             CacheBucket::FlatIndex,
             "html",
-            format!("{}.msgpack", cache_key::digest(&url.to_string())),
+            format!("{}.msgpack", cache_key::cache_digest(&url.to_string())),
         );
         let cache_control = match self.client.connectivity() {
             Connectivity::Online => CacheControl::from(

crates/uv-git/src/resolver.rs

@@ -63,7 +63,7 @@ impl GitResolver {
         fs::create_dir_all(&lock_dir).await?;
         let repository_url = RepositoryUrl::new(url.repository());
         let _lock = LockedFile::acquire(
-            lock_dir.join(cache_key::digest(&repository_url)),
+            lock_dir.join(cache_key::cache_digest(&repository_url)),
             &repository_url,
         )?;