[update_lib] fast date lookup for todo

youknowone · youknowone · commit f36614ab53e6 · 2026-03-01T23:38:53.000+09:00
diff --git a/scripts/update_lib/cmd_todo.py b/scripts/update_lib/cmd_todo.py
@@ -597,17 +597,16 @@ def format_all_todo(
             else get_module_diff_stat(item["name"], cpython_prefix, lib_prefix)
         )
 
-    # Add last_updated to displayed test items (verbose only - slow)
-    if verbose:
-        for tests in test_by_lib.values():
-            for test in tests:
-                test["last_updated"] = get_test_last_updated(
-                    test["name"], cpython_prefix, lib_prefix
-                )
-        for test in no_lib_tests:
+    # Add last_updated to displayed test items
+    for tests in test_by_lib.values():
+        for test in tests:
             test["last_updated"] = get_test_last_updated(
                 test["name"], cpython_prefix, lib_prefix
             )
+    for test in no_lib_tests:
+        test["last_updated"] = get_test_last_updated(
+            test["name"], cpython_prefix, lib_prefix
+        )
 
     # Format lib todo with embedded tests
     lines.extend(format_todo_list(lib_todo, test_by_lib, limit, verbose))
diff --git a/scripts/update_lib/deps.py b/scripts/update_lib/deps.py
@@ -1109,6 +1109,83 @@ def _count_path_diff(path_a: pathlib.Path, path_b: pathlib.Path) -> int:
     return 0
 
 
+@functools.cache
+def _bulk_last_updated() -> dict[str, str]:
+    """Get last git commit dates for all paths under Lib/ in one git call.
+
+    Keys are Lib/-relative paths (e.g. "re/__init__.py", "test/test_os.py",
+    "os.py"), plus directory rollups (e.g. "re", "test/test_zoneinfo").
+
+    Returns:
+        Dict mapping Lib/-relative path to date string.
+    """
+    file_map: dict[str, str] = {}
+    try:
+        result = subprocess.run(
+            ["git", "log", "--format=%cd", "--date=short", "--name-only", "--",
+             "Lib/"],
+            capture_output=True,
+            text=True,
+            timeout=30,
+        )
+        if result.returncode != 0:
+            return file_map
+    except Exception:
+        return file_map
+
+    current_date = None
+    for line in result.stdout.splitlines():
+        line = line.strip()
+        if not line:
+            continue
+        # Date lines are YYYY-MM-DD format
+        if len(line) == 10 and line[4] == "-" and line[7] == "-":
+            current_date = line
+        elif current_date and line.startswith("Lib/"):
+            # Strip "Lib/" prefix to get Lib-relative key
+            rel = line[4:]
+            if rel and rel not in file_map:
+                file_map[rel] = current_date
+
+    # Pre-compute directory rollups
+    dir_map: dict[str, str] = {}
+    for filepath, date in file_map.items():
+        parts = filepath.split("/")
+        for i in range(1, len(parts)):
+            dirpath = "/".join(parts[:i])
+            if dirpath not in dir_map or date > dir_map[dirpath]:
+                dir_map[dirpath] = date
+
+    dir_map.update(file_map)
+    return dir_map
+
+
+@functools.cache
+def _lib_prefix_stripped(lib_prefix: str) -> str:
+    """Get the normalized prefix to strip from paths, with trailing /."""
+    # e.g. "Lib" -> "Lib/", "../Lib" -> "../Lib/"
+    return lib_prefix.rstrip("/") + "/"
+
+
+def _lookup_last_updated(paths: list[str], lib_prefix: str) -> str | None:
+    """Look up the most recent date among paths from the bulk cache."""
+    cache = _bulk_last_updated()
+    prefix = _lib_prefix_stripped(lib_prefix)
+    latest = None
+    for p in paths:
+        # Strip lib_prefix to get Lib-relative key
+        # e.g. "Lib/test/test_os.py" -> "test/test_os.py"
+        #      "../Lib/re" -> "re"
+        if p.startswith(prefix):
+            key = p[len(prefix):]
+        else:
+            key = p
+        date = cache.get(key)
+        if date and (latest is None or date > latest):
+            latest = date
+    return latest
+
+
 def get_module_last_updated(
     name: str, cpython_prefix: str, lib_prefix: str
 ) -> str | None:
@@ -1126,18 +1203,7 @@ def get_module_last_updated(
             continue
     if not local_paths:
         return None
-    try:
-        result = subprocess.run(
-            ["git", "log", "-1", "--format=%cd", "--date=short", "--"] + local_paths,
-            capture_output=True,
-            text=True,
-            timeout=10,
-        )
-        if result.returncode == 0 and result.stdout.strip():
-            return result.stdout.strip()
-    except Exception:
-        pass
-    return None
+    return _lookup_last_updated(local_paths, lib_prefix)
 
 
 def get_module_diff_stat(name: str, cpython_prefix: str, lib_prefix: str) -> int:
@@ -1167,18 +1233,7 @@ def get_test_last_updated(
     local_path = _get_local_test_path(cpython_path, lib_prefix)
     if not local_path.exists():
         return None
-    try:
-        result = subprocess.run(
-            ["git", "log", "-1", "--format=%cd", "--date=short", "--", str(local_path)],
-            capture_output=True,
-            text=True,
-            timeout=10,
-        )
-        if result.returncode == 0 and result.stdout.strip():
-            return result.stdout.strip()
-    except Exception:
-        pass
-    return None
+    return _lookup_last_updated([str(local_path)], lib_prefix)
 
 
 def get_test_dependencies(
