Fix Cargo removing the sparse+ prefix from sparse URLs in .crates.toml

Author: arlosi

src/cargo/core/source/source_id.rs

@@ -81,6 +81,9 @@ impl SourceId {
     ///
     /// The canonical url will be calculated, but the precise field will not
     fn new(kind: SourceKind, url: Url, name: Option<&str>) -> CargoResult<SourceId> {
+        if kind == SourceKind::SparseRegistry {
+            assert!(url.as_str().starts_with("sparse+"));
+        }
         let source_id = SourceId::wrap(SourceIdInner {
             kind,
             canonical_url: CanonicalUrl::new(&url)?,
@@ -152,7 +155,7 @@ impl SourceId {
                     .with_precise(Some("locked".to_string())))
             }
             "sparse" => {
-                let url = url.into_url()?;
+                let url = string.into_url()?;
                 Ok(SourceId::new(SourceKind::SparseRegistry, url, None)?
                     .with_precise(Some("locked".to_string())))
             }
@@ -721,6 +724,7 @@ impl<'a> fmt::Display for SourceIdAsUrl<'a> {
                 ref url,
                 ..
             } => {
+                // Sparse registry URL already includes the `sparse+` prefix
                 write!(f, "{url}")
             }
             SourceIdInner {
@@ -864,4 +868,14 @@ mod tests {
         assert_eq!(gen_hash(source_id), 17459999773908528552);
         assert_eq!(crate::util::hex::short_hash(&source_id), "6568fe2c2fab5bfe");
     }
+
+    #[test]
+    fn serde_roundtrip() {
+        let url = "sparse+https://my-crates.io/".into_url().unwrap();
+        let source_id = SourceId::for_registry(&url).unwrap();
+        let formatted = format!("{}", source_id.as_url());
+        let deserialized = SourceId::from_url(&formatted).unwrap();
+        assert_eq!(formatted, "sparse+https://my-crates.io/");
+        assert_eq!(source_id, deserialized);
+    }
 }

src/cargo/util/canonical_url.rs

@@ -1,4 +1,4 @@
-use crate::util::{errors::CargoResult, IntoUrl};
+use crate::util::errors::CargoResult;
 use std::hash::{self, Hash};
 use url::Url;
 
@@ -56,17 +56,6 @@ impl CanonicalUrl {
             url.path_segments_mut().unwrap().pop().push(&last);
         }
 
-        // Ignore the protocol specifier (if any).
-        if url.scheme().starts_with("sparse+") {
-            // NOTE: it is illegal to use set_scheme to change sparse+http(s) to http(s).
-            url = url
-                .to_string()
-                .strip_prefix("sparse+")
-                .expect("we just found that prefix")
-                .into_url()
-                .expect("a valid url without a protocol specifier should still be valid");
-        }
-
         Ok(CanonicalUrl(url))
     }