Add check for illegal Widnows names

Fixes pre-commit#589

.pre-commit-hooks.yaml

@@ -40,6 +40,11 @@
     language: python
     types: [text, executable]
     stages: [commit, push, manual]
+-   id: check-illegal-windows-names
+    name: Check for illegal Windows names
+    description: Check for files that cannot be created on Windows.
+    entry: check-illegal-windows-names
+    language: python
 -   id: check-json
     name: check json
     description: checks json files for parseable syntax.

README.md

@@ -51,6 +51,9 @@ Checks for a common error of placing code before the docstring.
 #### `check-executables-have-shebangs`
 Checks that non-binary executables have a proper shebang.
 
+#### `check-illegal-windows-names`
+Check for files that cannot be created on Windows.
+
 #### `check-json`
 Attempts to load all json files to verify syntax.
 

pre_commit_hooks/check_illegal_windows_names.py

@@ -0,0 +1,64 @@
+import argparse
+import os.path
+from pathlib import Path
+from typing import Iterable, Iterator, Optional, Sequence, Set
+
+from pre_commit_hooks.util import added_files
+
+
+def lower_set(iterable: Iterable[str]) -> Set[str]:
+    return {x.lower() for x in iterable}
+
+
+def parents(file: str) -> Iterator[str]:
+    file = os.path.dirname(file)
+    while file:
+        yield file
+        file = os.path.dirname(file)
+
+
+def directories_for(files: Set[str]) -> Set[str]:
+    return {parent for file in files for parent in parents(file)}
+
+
+# https://docs.microsoft.com/en-us/windows/win32/fileio/naming-a-file
+ILLEGAL_NAMES = {
+    "CON",
+    "PRN",
+    "AUX",
+    "NUL",
+    *(f"COM{i}" for i in range(1, 10)),
+    *(f"LPT{i}" for i in range(1, 10)),
+}
+
+
+def find_illegal_windows_names(filenames: Sequence[str]) -> int:
+    relevant_files = set(filenames) | added_files()
+    relevant_files |= directories_for(relevant_files)
+    retv = 0
+
+    for filename in relevant_files:
+        root = Path(filename)
+        while root.suffix:
+            root = root.with_suffix("")
+        if root.name.lower() in lower_set(ILLEGAL_NAMES):
+            print(f"Illegal name {filename}")
+            retv = 1
+
+    return retv
+
+
+def main(argv: Optional[Sequence[str]] = None) -> int:
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "filenames",
+        nargs="*",
+        help="Filenames pre-commit believes are changed.",
+    )
+
+    args = parser.parse_args(argv)
+    return find_illegal_windows_names(args.filenames)
+
+
+if __name__ == "__main__":
+    exit(main())

setup.cfg

@@ -37,6 +37,7 @@ console_scripts =
     check-case-conflict = pre_commit_hooks.check_case_conflict:main
     check-docstring-first = pre_commit_hooks.check_docstring_first:main
     check-executables-have-shebangs = pre_commit_hooks.check_executables_have_shebangs:main
+    check-illegal-windows-names = pre_commit_hooks.check_illegal_windows_names:main
     check-json = pre_commit_hooks.check_json:main
     check-merge-conflict = pre_commit_hooks.check_merge_conflict:main
     check-shebang-scripts-are-executable = pre_commit_hooks.check_shebang_scripts_are_executable:main

tests/check_illegal_windows_names_test.py

@@ -0,0 +1,78 @@
+import sys
+
+import pytest
+
+from pre_commit_hooks.check_illegal_windows_names import (
+    find_illegal_windows_names,
+    main,
+    parents,
+)
+from pre_commit_hooks.util import cmd_output
+
+skip_win32 = pytest.mark.skipif(
+    sys.platform == "win32",
+    reason="case conflicts between directories and files",
+)
+
+
+def test_parents():
+    assert set(parents("a")) == set()
+    assert set(parents("a/b")) == {"a"}
+    assert set(parents("a/b/c")) == {"a/b", "a"}
+    assert set(parents("a/b/c/d")) == {"a/b/c", "a/b", "a"}
+
+
+def test_nothing_added(temp_git_dir):
+    with temp_git_dir.as_cwd():
+        assert find_illegal_windows_names(["f.py"]) == 0
+
+
+def test_adding_something(temp_git_dir):
+    with temp_git_dir.as_cwd():
+        temp_git_dir.join("f.py").write("print('hello world')")
+        cmd_output("git", "add", "f.py")
+
+        assert find_illegal_windows_names(["f.py"]) == 0
+
+
+@skip_win32  # pragma: win32 no cover
+def test_adding_something_with_illegal_filename(temp_git_dir):
+    with temp_git_dir.as_cwd():
+        temp_git_dir.join("CoM3.py").write("print('hello world')")
+        cmd_output("git", "add", "CoM3.py")
+
+        assert find_illegal_windows_names(["CoM3.py"]) == 1
+
+
+@skip_win32  # pragma: win32 no cover
+def test_adding_files_with_illegal_directory(temp_git_dir):
+    with temp_git_dir.as_cwd():
+        temp_git_dir.mkdir("lpt2").join("x").write("foo")
+        cmd_output("git", "add", "-A")
+
+        assert find_illegal_windows_names([]) == 1
+
+
+@skip_win32  # pragma: win32 no cover
+def test_adding_files_with_illegal_deep_directories(temp_git_dir):
+    with temp_git_dir.as_cwd():
+        temp_git_dir.mkdir("x").mkdir("y").join("pRn").write("foo")
+        cmd_output("git", "add", "-A")
+
+        assert find_illegal_windows_names([]) == 1
+
+
+@skip_win32  # pragma: win32 no cover
+def test_integration(temp_git_dir):
+    with temp_git_dir.as_cwd():
+        assert main(argv=[]) == 0
+
+        temp_git_dir.join("f.py").write("print('hello world')")
+        cmd_output("git", "add", "f.py")
+
+        assert main(argv=["f.py"]) == 0
+
+        temp_git_dir.join("CON.py").write("print('hello world')")
+        cmd_output("git", "add", "CON.py")
+
+        assert main(argv=["CON.py"]) == 1