From 5143869cf7ab89df25e29249c275e12a2817dcf8 Mon Sep 17 00:00:00 2001
From: Charlie Marsh <charlie.r.marsh@gmail.com>
Date: Fri, 1 Mar 2024 17:53:04 -0500
Subject: [PATCH] Drop hash fragment when downloading from registry

---
 crates/uv-client/src/registry_client.rs       | 12 ++++++++--
 .../src/distribution_database.rs              | 16 ++++++++------
 crates/uv-distribution/src/source/mod.rs      | 22 +++++++++----------
 3 files changed, 29 insertions(+), 21 deletions(-)

diff --git a/crates/uv-client/src/registry_client.rs b/crates/uv-client/src/registry_client.rs
index be511b8e8ef4d..cd25d04e11c00 100644
--- a/crates/uv-client/src/registry_client.rs
+++ b/crates/uv-client/src/registry_client.rs
@@ -225,6 +225,7 @@ impl RegistryClient {
     ) -> Result<Result<OwnedArchive<SimpleMetadata>, CachedClientError<Error>>, Error> {
         // Format the URL for PyPI.
         let mut url: Url = index.clone().into();
+        url.set_fragment(None);
         url.path_segments_mut()
             .unwrap()
             .pop_if_empty()
@@ -386,7 +387,10 @@ impl RegistryClient {
             .as_ref()
             .is_some_and(pypi_types::DistInfoMetadata::is_available)
         {
-            let url = Url::parse(&format!("{url}.metadata")).map_err(ErrorKind::UrlParseError)?;
+            // Like `pip`, drop the fragment, if it exists.
+            let mut url = url.clone();
+            url.set_fragment(None);
+            url.set_path(&format!("{}.metadata", url.path()));
 
             let cache_entry = self.cache.entry(
                 CacheBucket::Wheels,
@@ -455,6 +459,10 @@ impl RegistryClient {
             Connectivity::Offline => CacheControl::AllowStale,
         };
 
+        // Like `pip`, drop the fragment, if it exists.
+        let mut url = url.clone();
+        url.set_fragment(None);
+
         let client = self.client_raw.clone();
         let req = self
             .client
@@ -521,7 +529,7 @@ impl RegistryClient {
         }
 
         // Stream the file, searching for the METADATA.
-        let reader = self.stream_external(url).await?;
+        let reader = self.stream_external(&url).await?;
         read_metadata_async_stream(filename, url.to_string(), reader).await
     }
 
diff --git a/crates/uv-distribution/src/distribution_database.rs b/crates/uv-distribution/src/distribution_database.rs
index b79bb3526c9e7..a60f62e0bca44 100644
--- a/crates/uv-distribution/src/distribution_database.rs
+++ b/crates/uv-distribution/src/distribution_database.rs
@@ -104,7 +104,7 @@ impl<'a, Context: BuildContext + Send + Sync> DistributionDatabase<'a, Context>
                     return Err(Error::NoBinary);
                 }
 
-                let url = match &wheel.file.url {
+                let mut url = match &wheel.file.url {
                     FileLocation::RelativeUrl(base, url) => {
                         pypi_types::base_url_join_relative(base, url)?
                     }
@@ -152,6 +152,9 @@ impl<'a, Context: BuildContext + Send + Sync> DistributionDatabase<'a, Context>
                     }
                 };
 
+                // Like `pip`, drop the fragment, if it exists.
+                url.set_fragment(None);
+
                 // Create an entry for the wheel itself alongside its HTTP cache.
                 let wheel_entry = self.cache.entry(
                     CacheBucket::Wheels,
@@ -244,12 +247,11 @@ impl<'a, Context: BuildContext + Send + Sync> DistributionDatabase<'a, Context>
                     .instrument(info_span!("download", wheel = %wheel))
                 };
 
-                let req = self
-                    .client
-                    .cached_client()
-                    .uncached()
-                    .get(wheel.url.raw().clone())
-                    .build()?;
+                // Like `pip`, drop the fragment, if it exists.
+                let mut url = wheel.url.raw().clone();
+                url.set_fragment(None);
+
+                let req = self.client.cached_client().uncached().get(url).build()?;
                 let cache_control = match self.client.connectivity() {
                     Connectivity::Online => CacheControl::from(
                         self.cache
diff --git a/crates/uv-distribution/src/source/mod.rs b/crates/uv-distribution/src/source/mod.rs
index 078283e4e0a58..26188440b7ff8 100644
--- a/crates/uv-distribution/src/source/mod.rs
+++ b/crates/uv-distribution/src/source/mod.rs
@@ -263,6 +263,10 @@ impl<'a, T: BuildContext> SourceDistCachedBuilder<'a, T> {
             Connectivity::Offline => CacheControl::AllowStale,
         };
 
+        // Like `pip`, drop the fragment, if it exists.
+        let mut url = url.clone();
+        url.set_fragment(None);
+
         let download = |response| {
             async {
                 // At this point, we're seeing a new or updated source distribution. Initialize a
@@ -280,12 +284,7 @@ impl<'a, T: BuildContext> SourceDistCachedBuilder<'a, T> {
             .boxed()
             .instrument(info_span!("download", source_dist = %source_dist))
         };
-        let req = self
-            .client
-            .cached_client()
-            .uncached()
-            .get(url.clone())
-            .build()?;
+        let req = self.client.cached_client().uncached().get(url).build()?;
         let manifest = self
             .client
             .cached_client()
@@ -366,6 +365,10 @@ impl<'a, T: BuildContext> SourceDistCachedBuilder<'a, T> {
             Connectivity::Offline => CacheControl::AllowStale,
         };
 
+        // Like `pip`, drop the fragment, if it exists.
+        let mut url = url.clone();
+        url.set_fragment(None);
+
         let download = |response| {
             async {
                 // At this point, we're seeing a new or updated source distribution. Initialize a
@@ -383,12 +386,7 @@ impl<'a, T: BuildContext> SourceDistCachedBuilder<'a, T> {
             .boxed()
             .instrument(info_span!("download", source_dist = %source_dist))
         };
-        let req = self
-            .client
-            .cached_client()
-            .uncached()
-            .get(url.clone())
-            .build()?;
+        let req = self.client.cached_client().uncached().get(url).build()?;
         let manifest = self
             .client
             .cached_client()