fix(cache): preserve Windows source fingerprint cache hits

CHANGELOG.md

@@ -10,6 +10,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ### Bug Fixes
 
 - Avoid network false positives for bounded README examples that download sample images over HTTPS from Hugging Face.
+- Preserve Windows source-fingerprint cache hits while continuing to reject swapped or modified files.
 - Prevent Windows cache identity probes from creating locked temporary files inside scanned directories.
 - Preserve locked Windows cache probes reached through directory aliases while clearing stale scan results.
 - Place cross-volume Windows cache identity probes near the volume root instead of the nearest ancestor of the scanned path, so a probe can no longer appear inside a directory tree that a concurrent scan is walking.

modelaudit/cache/scan_results_cache.py

@@ -1715,6 +1715,20 @@ def _regular_file_identity_fingerprint(file_stat: os.stat_result) -> str:
         digest = hashlib.sha256(identity.encode()).hexdigest()
         return f"{_CALL_GRAPH_REGULAR_FILE_FINGERPRINT}:{digest}"
 
+    @staticmethod
+    def _cross_view_file_stat_identity(file_stat: os.stat_result) -> tuple[int, ...]:
+        """Compare path and descriptor views without Windows-only stat differences."""
+        identity = (
+            file_stat.st_dev,
+            file_stat.st_ino,
+            stat.S_IFMT(file_stat.st_mode),
+            file_stat.st_size,
+            file_stat.st_mtime_ns,
+        )
+        if os.name == "nt":
+            return identity
+        return (*identity, file_stat.st_ctime_ns)
+
     @staticmethod
     def _bounded_source_fingerprint(path: Path) -> str | None:
         flags = os.O_RDONLY | getattr(os, "O_BINARY", 0) | getattr(os, "O_NONBLOCK", 0)
@@ -1766,15 +1780,9 @@ def _bounded_source_fingerprint(path: Path) -> str | None:
                 path_stat = path.stat()
             except OSError as error:
                 raise ValueError("source fingerprint candidate path changed while being read") from error
-            path_identity = (
-                path_stat.st_dev,
-                path_stat.st_ino,
-                path_stat.st_mode,
-                path_stat.st_size,
-                path_stat.st_mtime_ns,
-                path_stat.st_ctime_ns,
-            )
-            if before_identity != after_identity or after_identity != path_identity:
+            if before_identity != after_identity or ScanResultsCache._cross_view_file_stat_identity(
+                after
+            ) != ScanResultsCache._cross_view_file_stat_identity(path_stat):
                 raise ValueError("source fingerprint candidate changed while being read")
             if is_extension:
                 return ScanResultsCache._regular_file_identity_fingerprint(after)
@@ -1873,7 +1881,20 @@ def _bounded_read_fingerprint(path: Path, read_limit: int, require_complete: boo
                 path_stat.st_mtime_ns,
                 path_stat.st_ctime_ns,
             )
-            if before_identity != after_identity or after_identity != path_identity:
+            initial_path_identity = (
+                path_before.st_dev,
+                path_before.st_ino,
+                path_before.st_mode,
+                path_before.st_size,
+                path_before.st_mtime_ns,
+                path_before.st_ctime_ns,
+            )
+            if (
+                before_identity != after_identity
+                or initial_path_identity != path_identity
+                or ScanResultsCache._cross_view_file_stat_identity(after)
+                != ScanResultsCache._cross_view_file_stat_identity(path_stat)
+            ):
                 raise ValueError("read fingerprint candidate changed while being read")
             if require_complete and len(source) > read_limit:
                 raise ValueError("read fingerprint budget exceeded")

tests/cache/test_cache_correctness.py

@@ -3,6 +3,7 @@
 import hashlib
 import json
 import os
+import stat
 import sys
 import tempfile
 import threading
@@ -22,7 +23,7 @@
     SourcelessFileLoader,
 )
 from pathlib import Path
-from types import FunctionType, ModuleType
+from types import FunctionType, ModuleType, SimpleNamespace
 from typing import Any, BinaryIO, cast
 from zipimport import zipimporter
 
@@ -3749,6 +3750,112 @@ def test_scan_cache_invalidates_replaced_large_extension_candidate(tmp_path: Pat
     assert cache.get_cached_result(str(file_path), version_context=version_context) is None
 
 
+@pytest.mark.parametrize("fingerprint_kind", ["source", "read"])
+def test_source_fingerprints_accept_windows_cross_view_stat_differences(
+    tmp_path: Path,
+    monkeypatch: pytest.MonkeyPatch,
+    fingerprint_kind: str,
+) -> None:
+    source_path = tmp_path / "helper.py"
+    source = b"def entrypoint():\n    return 1\n"
+    source_path.write_bytes(source)
+    original_fstat = os.fstat
+
+    def windows_descriptor_stat(file_descriptor: int) -> os.stat_result:
+        file_stat = original_fstat(file_descriptor)
+        return cast(
+            os.stat_result,
+            SimpleNamespace(
+                st_dev=file_stat.st_dev,
+                st_ino=file_stat.st_ino,
+                st_mode=file_stat.st_mode ^ 0o111,
+                st_size=file_stat.st_size,
+                st_mtime_ns=file_stat.st_mtime_ns,
+                st_ctime_ns=file_stat.st_ctime_ns + 1,
+            ),
+        )
+
+    with monkeypatch.context() as windows:
+        windows.setattr(os, "fstat", windows_descriptor_stat)
+        windows.setattr(os, "name", "nt")
+        if fingerprint_kind == "source":
+            fingerprint = ScanResultsCache._bounded_source_fingerprint(source_path)
+            expected = hashlib.sha256(source).hexdigest()
+        else:
+            fingerprint = ScanResultsCache._bounded_read_fingerprint(source_path, 64 * 1024, True)
+            expected = hashlib.sha256(b"file\0" + source).hexdigest()
+
+    assert fingerprint == expected
+
+
+@pytest.mark.parametrize("fingerprint_kind", ["source", "read"])
+def test_source_fingerprints_reject_windows_cross_view_file_replacement(
+    tmp_path: Path,
+    monkeypatch: pytest.MonkeyPatch,
+    fingerprint_kind: str,
+) -> None:
+    source_path = tmp_path / "helper.py"
+    source_path.write_bytes(b"def entrypoint():\n    return 1\n")
+    original_fstat = os.fstat
+
+    def replaced_descriptor_stat(file_descriptor: int) -> os.stat_result:
+        file_stat = original_fstat(file_descriptor)
+        return cast(
+            os.stat_result,
+            SimpleNamespace(
+                st_dev=file_stat.st_dev,
+                st_ino=file_stat.st_ino + 1,
+                st_mode=stat.S_IFMT(file_stat.st_mode) | 0o666,
+                st_size=file_stat.st_size,
+                st_mtime_ns=file_stat.st_mtime_ns,
+                st_ctime_ns=file_stat.st_ctime_ns + 1,
+            ),
+        )
+
+    with monkeypatch.context() as windows:
+        windows.setattr(os, "fstat", replaced_descriptor_stat)
+        windows.setattr(os, "name", "nt")
+        with pytest.raises(ValueError, match="changed while being read"):
+            if fingerprint_kind == "source":
+                ScanResultsCache._bounded_source_fingerprint(source_path)
+            else:
+                ScanResultsCache._bounded_read_fingerprint(source_path, 64 * 1024, True)
+
+
+@pytest.mark.skipif(os.name == "nt", reason="Windows ctime differs across stat views")
+@pytest.mark.parametrize("fingerprint_kind", ["source", "read"])
+def test_source_fingerprints_reject_posix_cross_view_ctime_change(
+    tmp_path: Path,
+    monkeypatch: pytest.MonkeyPatch,
+    fingerprint_kind: str,
+) -> None:
+    source_path = tmp_path / "helper.py"
+    source_path.write_bytes(b"def entrypoint():\n    return 1\n")
+    original_fstat = os.fstat
+
+    def changed_descriptor_stat(file_descriptor: int) -> os.stat_result:
+        file_stat = original_fstat(file_descriptor)
+        return cast(
+            os.stat_result,
+            SimpleNamespace(
+                st_dev=file_stat.st_dev,
+                st_ino=file_stat.st_ino,
+                st_mode=file_stat.st_mode,
+                st_size=file_stat.st_size,
+                st_mtime_ns=file_stat.st_mtime_ns,
+                st_ctime_ns=file_stat.st_ctime_ns + 1,
+            ),
+        )
+
+    monkeypatch.setattr(os, "fstat", changed_descriptor_stat)
+
+    with pytest.raises(ValueError, match="changed while being read"):
+        if fingerprint_kind == "source":
+            ScanResultsCache._bounded_source_fingerprint(source_path)
+        else:
+            ScanResultsCache._bounded_read_fingerprint(source_path, 64 * 1024, True)
+
+
 @pytest.mark.skipif(sys.platform == "win32", reason="Windows prevents replacing an open source file")
 def test_source_fingerprint_rejects_path_replacement_during_read(
     tmp_path: Path,