Crates in third party registries can override the cached source of other crates

@@ -501,6 +501,13 @@ mod test {
         assert_eq!(ident1, ident2);
     }
 
+    #[test]
+    fn test_canonicalize_idents_does_not_strip_dot_git_for_sparse() {
+        let ident1 = ident(&src("sparse+https://crates.io/fake-registry"));
+        let ident2 = ident(&src("sparse+https://crates.io/fake-registry.git"));
+        assert_ne!(ident1, ident2);
+    }
+
     fn src(s: &str) -> SourceId {
         SourceId::for_git(&s.into_url().unwrap(), GitReference::DefaultBranch).unwrap()
     }

@@ -198,7 +198,7 @@ use flate2::read::GzDecoder;
 use futures::FutureExt as _;
 use serde::Deserialize;
 use serde::Serialize;
-use tar::Archive;
+use tar::{Archive, EntryType};
 use tracing::debug;
 
 use crate::core::dependency::Dependency;
@@ -1019,6 +1019,14 @@ fn unpack(
             )
         }
 
+        // Prevent unpacking symlinks and other unexpected entry types
+        match entry.header().entry_type() {
+            EntryType::Regular | EntryType::Directory => {}
+            t => anyhow::bail!(
+                "invalid tarball downloaded, contains an entry at {entry_path:?} with invalid type {t:?}",
+            ),
+        }
+
         // Prevent unpacking the lockfile from the crate itself.
         if entry_path
             .file_name()

@@ -33,27 +33,31 @@ impl CanonicalUrl {
             url.path_segments_mut().unwrap().pop_if_empty();
         }
 
-        // For GitHub URLs specifically, just lower-case everything. GitHub
-        // treats both the same, but they hash differently, and we're gonna be
-        // hashing them. This wants a more general solution, and also we're
-        // almost certainly not using the same case conversion rules that GitHub
-        // does. (See issue #84)
-        if url.host_str() == Some("github.com") {
-            url = format!("https{}", &url[url::Position::AfterScheme..])
-                .parse()
-                .unwrap();
-            let path = url.path().to_lowercase();
-            url.set_path(&path);
-        }
+        // Perform further canonicalization specific to git registries, which
+        // do not contain a `+` specifier.
+        if !url.scheme().contains('+') {
+            // For GitHub URLs specifically, just lower-case everything. GitHub
+            // treats both the same, but they hash differently, and we're gonna be
+            // hashing them. This wants a more general solution, and also we're
+            // almost certainly not using the same case conversion rules that GitHub
+            // does. (See issue #84)
+            if url.host_str() == Some("github.com") {
+                url = format!("https{}", &url[url::Position::AfterScheme..])
+                    .parse()
+                    .unwrap();
+                let path = url.path().to_lowercase();
+                url.set_path(&path);
+            }
 
-        // Repos can generally be accessed with or without `.git` extension.
-        let needs_chopping = url.path().ends_with(".git");
-        if needs_chopping {
-            let last = {
-                let last = url.path_segments().unwrap().next_back().unwrap();
-                last[..last.len() - 4].to_owned()
-            };
-            url.path_segments_mut().unwrap().pop().push(&last);
+            // Repos can generally be accessed with or without `.git` extension.
+            let needs_chopping = url.path().ends_with(".git");
+            if needs_chopping {
+                let last = {
+                    let last = url.path_segments().unwrap().next_back().unwrap();
+                    last[..last.len() - 4].to_owned()
+                };
+                url.path_segments_mut().unwrap().pop().push(&last);
+            }
         }
 
         Ok(CanonicalUrl(url))

@@ -3276,8 +3276,7 @@ fn package_lock_inside_package_is_overwritten() {
 }
 
 #[cargo_test]
-fn package_lock_as_a_symlink_inside_package_is_overwritten() {
-    let registry = registry::init();
+fn package_lock_as_a_symlink_inside_package_is_invalid() {
     let p = project()
         .file(
             "Cargo.toml",
@@ -3300,21 +3299,23 @@ fn package_lock_as_a_symlink_inside_package_is_overwritten() {
         .symlink(".cargo-ok", "src/lib.rs")
         .publish();
 
-    p.cargo("check").run();
+    p.cargo("check")
+        .with_status(101)
+        .with_stderr_data(str![[r#"
+[UPDATING] `dummy-registry` index
+[LOCKING] 1 package to latest compatible version
+[DOWNLOADING] crates ...
+[DOWNLOADED] bar v0.0.1 (registry `dummy-registry`)
+[ERROR] failed to download replaced source registry `crates-io`
 
-    let id = SourceId::for_registry(registry.index_url()).unwrap();
-    let hash = cargo::util::hex::short_hash(&id);
-    let pkg_root = paths::cargo_home()
-        .join("registry")
-        .join("src")
-        .join(format!("-{}", hash))
-        .join("bar-0.0.1");
-    let ok = pkg_root.join(".cargo-ok");
-    let librs = pkg_root.join("src/lib.rs");
+Caused by:
+  failed to unpack package `bar v0.0.1 (registry `dummy-registry`)`
 
-    // Is correctly overwritten and doesn't affect the file linked to
-    assert_eq!(ok.metadata().unwrap().len(), 7);
-    assert_eq!(fs::read_to_string(librs).unwrap(), "pub fn f() {}");
+Caused by:
+  invalid tarball downloaded, contains an entry at "bar-0.0.1/.cargo-ok" with invalid type Symlink
+
+"#]])
+        .run();
 }
 
 #[cargo_test]
@@ -4770,13 +4771,7 @@ Caused by:
   failed to unpack package `bar v1.0.0 (registry `dummy-registry`)`
 
 Caused by:
-  failed to unpack entry at `bar-1.0.0/smuggled`
-
-Caused by:
-  failed to unpack `[ROOT]/home/.cargo/registry/src/-[HASH]/bar-1.0.0/smuggled`
-
-Caused by:
-  [..] when creating dir [ROOT]/home/.cargo/registry/src/-[HASH]/bar-1.0.0/smuggled
+  invalid tarball downloaded, contains an entry at "bar-1.0.0/smuggled" with invalid type Symlink
 
 "#]])
         .run();