Make is_sub_path faster (#17962)

See #17948

- 1.01x faster on clean
- 1.06x faster on long
- 1.04x faster on openai
- 1.26x faster on openai incremental

Author: hauntsaninja

mypy/build.py

@@ -59,7 +59,7 @@
     get_mypy_comments,
     hash_digest,
     is_stub_package_file,
-    is_sub_path,
+    is_sub_path_normabs,
     is_typeshed_file,
     module_prefix,
     read_py_file,
@@ -3528,10 +3528,9 @@ def is_silent_import_module(manager: BuildManager, path: str) -> bool:
     if manager.options.no_silence_site_packages:
         return False
     # Silence errors in site-package dirs and typeshed
-    return any(
-        is_sub_path(path, dir)
-        for dir in manager.search_paths.package_path + manager.search_paths.typeshed_path
-    )
+    if any(is_sub_path_normabs(path, dir) for dir in manager.search_paths.package_path):
+        return True
+    return any(is_sub_path_normabs(path, dir) for dir in manager.search_paths.typeshed_path)
 
 
 def write_undocumented_ref_info(

mypy/modulefinder.py

@@ -668,10 +668,13 @@ def mypy_path() -> list[str]:
 def default_lib_path(
     data_dir: str, pyversion: tuple[int, int], custom_typeshed_dir: str | None
 ) -> list[str]:
-    """Return default standard library search paths."""
+    """Return default standard library search paths. Guaranteed to be normalised."""
+
+    data_dir = os.path.abspath(data_dir)
     path: list[str] = []
 
     if custom_typeshed_dir:
+        custom_typeshed_dir = os.path.abspath(custom_typeshed_dir)
         typeshed_dir = os.path.join(custom_typeshed_dir, "stdlib")
         mypy_extensions_dir = os.path.join(custom_typeshed_dir, "stubs", "mypy-extensions")
         versions_file = os.path.join(typeshed_dir, "VERSIONS")
@@ -711,7 +714,7 @@ def default_lib_path(
 
 @functools.lru_cache(maxsize=None)
 def get_search_dirs(python_executable: str | None) -> tuple[list[str], list[str]]:
-    """Find package directories for given python.
+    """Find package directories for given python. Guaranteed to return absolute paths.
 
     This runs a subprocess call, which generates a list of the directories in sys.path.
     To avoid repeatedly calling a subprocess (which can be slow!) we
@@ -773,6 +776,7 @@ def compute_search_paths(
         root_dir = os.getenv("MYPY_TEST_PREFIX", None)
         if not root_dir:
             root_dir = os.path.dirname(os.path.dirname(__file__))
+        root_dir = os.path.abspath(root_dir)
         lib_path.appendleft(os.path.join(root_dir, "test-data", "unit", "lib-stub"))
     # alt_lib_path is used by some tests to bypass the normal lib_path mechanics.
     # If we don't have one, grab directories of source files.
@@ -829,6 +833,7 @@ def compute_search_paths(
     return SearchPaths(
         python_path=tuple(reversed(python_path)),
         mypy_path=tuple(mypypath),
+        # package_path and typeshed_path must be normalised and absolute via os.path.abspath
         package_path=tuple(sys_path + site_packages),
         typeshed_path=tuple(lib_path),
     )

mypy/util.py

@@ -6,7 +6,6 @@
 import io
 import json
 import os
-import pathlib
 import re
 import shutil
 import sys
@@ -419,9 +418,26 @@ def replace_object_state(
             pass
 
 
-def is_sub_path(path1: str, path2: str) -> bool:
-    """Given two paths, return if path1 is a sub-path of path2."""
-    return pathlib.Path(path2) in pathlib.Path(path1).parents
+def is_sub_path_normabs(path: str, dir: str) -> bool:
+    """Given two paths, return if path is a sub-path of dir.
+
+    Moral equivalent of: Path(dir) in Path(path).parents
+
+    Similar to the pathlib version:
+    - Treats paths case-sensitively
+    - Does not fully handle unnormalised paths (e.g. paths with "..")
+    - Does not handle a mix of absolute and relative paths
+    Unlike the pathlib version:
+    - Fast
+    - On Windows, assumes input has been slash normalised
+    - Handles even fewer unnormalised paths (e.g. paths with "." and "//")
+
+    As a result, callers should ensure that inputs have had os.path.abspath called on them
+    (note that os.path.abspath will normalise)
+    """
+    if not dir.endswith(os.sep):
+        dir += os.sep
+    return path.startswith(dir)
 
 
 if sys.platform == "linux" or sys.platform == "darwin":