Fix hosted requirements rewrite of user direct refs (#542)

crates/socket-patch-cli/CLI_CONTRACT.md

@@ -1396,6 +1396,7 @@ Every `--json` invocation emits a single JSON object that follows the **unified
 | `redirect_vlt_lock_unsupported` | `redirect.warnings[]` (warning) | scan/get `--mode hosted` (vlt): `vlt-lock.json` has a `lockfileVersion` other than absent, `0` or `1` (decided on the raw JSON token), is not a JSON object, starts with a UTF-8 BOM, or its `nodes` section is not vlt's one-node-per-line layout. Nothing rewritten; also refuses a vendored → hosted takeover of a `flavor: "vlt"` entry before its revert (`redirect.skipped[].reason`). Exit 0. |
 | `redirect_takeover_kept_vendored` | `redirect.warnings[]` (warning) | scan/get `--mode hosted`: a vendored → hosted takeover the hosted rewrite would not pin was retracted (see **Staged takeover**): the package keeps its vendored wiring, ledger entry and artifact byte-identical and stays patched. The detail names the cause code, which is also the purl's `redirect.skipped[].reason`. Exit 0. Replaces v5.0-pre `redirect_takeover_unpatched`, which reported a package left unpatched in both modes and is no longer emitted. |
 | `redirect_takeover_not_pinned` | `redirect.skipped[].reason` | scan/get `--mode hosted`: the skip reason of a retracted takeover when no rewriter warning names the cause. |
+| `redirect_requirements_direct_reference` | `redirect.warnings[]` (warning) | scan/get `--mode hosted` (pypi / requirements.txt): the root `requirements.txt` installs the patched release from a PEP 508 direct reference the user wrote (`name @ <url>` whose archive names that release: a `files.pythonhosted.org` file, a private mirror, an internal fork build, a `file://` path), not from socket-patch's own hosted artifact. It is the user's own source choice, so the line is left byte-for-byte and the package is not redirected (#542). Before v5.0 the line was swapped for the hosted build, and rollback then wrote a plain `name==version` index pin. |
 | `redirect_requirements_takeover_unreachable` | `redirect.warnings[]` (warning) | scan/get `--mode hosted` (pypi / requirements.txt): a vendored → hosted takeover of a package that vendored mode wired through a pin in a `-r` include, or through a `(transitive)` line it appended to the root `requirements.txt`. Hosted mode only rewrites an existing pin in the root `requirements.txt`, so the staged takeover is retracted (wet and `--dry-run`): the vendored wiring, ledger entry and wheel stay byte-identical, the purl is skipped with this code as `redirect.skipped[].reason`, and nothing is redirected for it. Exit 0. The detail names the remedy and its reach: run `socket-patch vendor --revert` (it reverts EVERY vendored package in the project, not just this one), move the pin from the include into the root `requirements.txt` and delete it from the include (or, for a `(transitive)` line, add an exact `==` pin to the root file), then re-run `scan --mode hosted`. |
 | `redirect_uv_takeover_version_unreachable` | `redirect.warnings[]` (warning) | scan/get `--mode hosted` (pypi / uv): a vendored → hosted takeover of a package whose recorded pre-vendor `uv.lock` entry is at another version than the patch (vendored uv pins the lock entry down to the patch's version, for example when the lock resolved a newer release). Reverting would bring the lock's own version back, and hosted mode only pins the version the lock resolves, so the staged takeover is retracted (wet and `--dry-run`): the vendored wiring, ledger entry and wheel stay byte-identical, the purl is skipped with this code as `redirect.skipped[].reason`, and nothing is redirected for it. Exit 0. The detail names the remedy and its reach: run `socket-patch vendor --revert` (it reverts EVERY vendored package in the project, not just this one), make the project resolve the patch's version (for example an exact `==` requirement) and re-lock, then re-run `scan --mode hosted`. |
 | `redirect_vlt_missing_sha512` / `redirect_vlt_entry_not_found` / `redirect_vlt_entry_vendored` / `redirect_vlt_unsupported_lock_key` | `redirect.warnings[]` (warning) | scan/get `--mode hosted` (vlt): the grant has no sha512 / the lock has no default-registry node for `name@version` / the only match is a vendored `file` node under `.socket/vendor/npm/<uuid>/` / a default-registry instance is outside vlt's node-line grammar or still unpatched after the splice. Per dep; none of the dep's instances is written. `redirect_vlt_missing_sha512` and `redirect_vlt_unsupported_lock_key` refuse the dep: it is never confirmed, whichever lock drives (a sibling lock may still carry its rewritten URL). `redirect_vlt_entry_not_found` and `redirect_vlt_entry_vendored` only say `vlt-lock.json` does not wire it: while vlt drives it is not confirmed; otherwise a sibling lock's rules may confirm it. Exit 0. |

crates/socket-patch-core/src/patch/redirect/requirements.rs

@@ -128,8 +128,17 @@ fn without_hashes(text: &str) -> String {
 }
 
 enum RequirementVersion {
-    /// `==X` (PEP 440 equality), or a direct reference whose archive names X.
+    /// `==X` (PEP 440 equality), or socket-patch's own hosted direct
+    /// reference whose archive names X (a re-scan).
     Exact(String),
+    /// A direct reference to an archive naming release X that is NOT
+    /// socket-patch's hosted artifact: a public-index file, a private
+    /// mirror, an internal fork build, a `file://` path. That is the user's
+    /// own source choice, never rewritten (#542).
+    UserReference {
+        location: String,
+        version: String,
+    },
     /// `===X`: arbitrary equality, a plain string comparison.
     Arbitrary(String),
     Unpinned,
@@ -164,8 +173,17 @@ fn requirement_version(specifier: &str, name_re: &Regex, name: &str) -> Requirem
         return RequirementVersion::Unpinned;
     }
     if let Some(location) = tail.strip_prefix('@') {
-        return archive_version(location.trim(), name)
-            .map_or(RequirementVersion::Ambiguous, RequirementVersion::Exact);
+        let location = location.trim();
+        let Some(version) = archive_version(location, name) else {
+            return RequirementVersion::Ambiguous;
+        };
+        if crate::vendor::lock_inventory::pypi::socket_reference_coords(location).is_none() {
+            return RequirementVersion::UserReference {
+                location: location.to_string(),
+                version,
+            };
+        }
+        return RequirementVersion::Exact(version);
     }
     let (arbitrary, version) = match tail.strip_prefix("===") {
         Some(version) => (true, Some(version)),
@@ -264,6 +282,23 @@ pub(super) fn rewrite(
                     continue
                 }
                 RequirementVersion::Arbitrary(version) if version != dep.version => continue,
+                RequirementVersion::UserReference { version, .. }
+                    if !crate::utils::pep440::versions_equal(&version, &dep.version) =>
+                {
+                    continue
+                }
+                RequirementVersion::UserReference { location, .. } => {
+                    matched = true;
+                    result.warnings.push(RewriteWarning {
+                        code: "redirect_requirements_direct_reference".into(),
+                        detail: format!(
+                            "requirements.txt installs {}@{} from the direct reference {location}; \
+                             refusing to overwrite a user-authored source (not rewritten)",
+                            dep.name, dep.version
+                        ),
+                    });
+                    continue;
+                }
                 RequirementVersion::Exact(_) | RequirementVersion::Arbitrary(_) => {}
                 RequirementVersion::Unpinned
                     if row_counts.get(&target) == Some(&1)
@@ -755,11 +790,7 @@ mod tests {
             )
         );
         assert_eq!(result.edits.len(), 1);
-        for source in [
-            "requests @ https://files.pythonhosted.org/requests-2.28.1.tar.gz#sha256=old",
-            "requests ( == 2.28.1 )",
-            "requests===2.28.1",
-        ] {
+        for source in ["requests ( == 2.28.1 )", "requests===2.28.1"] {
             let result = rewrite_registry_redirect(&input(source), &[patch()]);
             assert_eq!(
                 result.files["requirements.txt"],
@@ -768,6 +799,46 @@ mod tests {
         }
     }
 
+    /// #542: a direct reference the user wrote — a public-index file, a
+    /// private mirror, an internal fork, a `file://` path — is their own
+    /// source choice: refused with a warning and left byte-for-byte, never
+    /// swapped for the hosted build (whose rollback would then write a
+    /// plain index pin, losing the original source).
+    #[test]
+    fn user_direct_references_are_refused_not_rewritten() {
+        for location in [
+            "https://files.pythonhosted.org/requests-2.28.1.tar.gz#sha256=old",
+            "http://127.0.0.1:18766/requests-2.28.1-py3-none-any.whl",
+            "file:///abs/private/requests-2.28.1-py3-none-any.whl",
+            "https://mirror.internal/requests-2.28.01-py3-none-any.whl",
+        ] {
+            let source = format!("idna==3.7\nrequests @ {location} ; python_version >= '3.7'\n");
+            let result = rewrite_registry_redirect(&input(&source), &[patch()]);
+            assert!(
+                result.files.is_empty() && result.edits.is_empty(),
+                "{location}"
+            );
+            let codes: Vec<_> = result.warnings.iter().map(|w| w.code.as_str()).collect();
+            assert_eq!(
+                codes,
+                ["redirect_requirements_direct_reference"],
+                "{location}"
+            );
+            assert!(result.warnings[0].detail.contains(location));
+            assert!(result.confirmed_requirements_uuids.is_empty());
+        }
+        // A user reference to another release is not this patch's entry.
+        let result = rewrite_registry_redirect(
+            &input("requests @ https://mirror.internal/requests-2.32.0-py3-none-any.whl\n"),
+            &[patch()],
+        );
+        assert!(result.files.is_empty());
+        assert_eq!(
+            result.warnings[0].code,
+            "redirect_requirements_entry_not_found"
+        );
+    }
+
     /// #376: an unhashed requirements file must stay unhashed. pip turns
     /// hash-checking mode on for the WHOLE install as soon as one line has a
     /// `--hash`, so pinning only the patched line breaks every other

crates/socket-patch-core/src/vendor/lock_inventory/pypi.rs

@@ -469,7 +469,7 @@ pub(super) fn is_public_pypi_url(url: &str) -> bool {
 /// `[./].socket/vendor/pypi/<uuid>/<wheel>` (coordinates from the shared
 /// vendored-leaf table, [`crate::vendor::path::leaf_to_purl`]). `None` for
 /// a user's own file/path reference.
-pub(super) fn socket_reference_coords(reference: &str) -> Option<(String, String)> {
+pub(crate) fn socket_reference_coords(reference: &str) -> Option<(String, String)> {
     if reference.contains("://") {
         let url = hosted_artifact_url(reference).ok()?;
         url.uuid_level.as_ref()?;

docs/testing/uv-compatibility.md

@@ -154,8 +154,11 @@ frozen, locked, and ordinary installation outcomes separately where supported.
   those cases requires marker-specific source mappings. Standalone PEP 751
   rewriting selects the exact package version; duplicate entries for the same
   name and version are refused when source selection is ambiguous.
-- Hosted requirements select exact `==`/`===` pins or identifiable archive URLs.
-  Other versions remain unchanged. A bare requirement is rewritten only when
+- Hosted requirements select exact `==`/`===` pins or socket-patch's own
+  hosted archive URLs. A user-authored direct reference to the patched release
+  (`name @ <url>` on any other origin, `files.pythonhosted.org` and `file://`
+  included) is refused with `redirect_requirements_direct_reference` and left
+  unchanged. Other versions remain unchanged. A bare requirement is rewritten only when
   one row and one override version identify the selection. Ranges, wildcard
   pins, opaque URLs, and ambiguous unpinned rows are reported as
   `redirect_requirements_version_ambiguous` and preserved.