Add check for text file encodings

.pre-commit-hooks.yaml

@@ -32,6 +32,12 @@
     entry: check-docstring-first
     language: python
     types: [python]
+-   id: check-encoding
+    name: check text files have desired encoding
+    description: checks that text files decode cleanly using an encoding.
+    entry: check-encoding
+    language: python
+    types: [text]
 -   id: check-executables-have-shebangs
     name: check that executables have shebangs
     description: ensures that (non-binary) executables have a shebang.

README.md

@@ -49,6 +49,11 @@ Check for files with names that would conflict on a case-insensitive filesystem
 #### `check-docstring-first`
 Checks for a common error of placing code before the docstring.
 
+#### `check-encoding`
+Checks that text files have desired encoding.
+  - `--encoding` - specify encoding to assert; if not specified, default is
+    [platform dependent](https://docs.python.org/3/library/functions.html?highlight=encoding#open)
+
 #### `check-executables-have-shebangs`
 Checks that non-binary executables have a proper shebang.
 

pre_commit_hooks/check_encoding.py

@@ -0,0 +1,29 @@
+import argparse
+from typing import Optional
+from typing import Sequence
+
+
+def main(argv: Optional[Sequence[str]] = None) -> int:
+    parser = argparse.ArgumentParser()
+    parser.add_argument('filenames', nargs='*', help='Filenames to check.')
+    parser.add_argument('--encoding', help='Encoding to assert.')
+    args = parser.parse_args(argv)
+
+    retval = 0
+    for filename in args.filenames:
+        try:
+            with open(filename, encoding=args.encoding) as f:
+                f.read()
+        except LookupError as exc:
+            # Unknown encoding, don't bother with the rest
+            print(f'{__file__}: {exc}')
+            retval = 2
+            break
+        except Exception as exc:
+            print(f'{filename}: {exc}')
+            retval = 1
+    return retval
+
+
+if __name__ == '__main__':
+    raise SystemExit(main())

setup.cfg

@@ -41,6 +41,7 @@ console_scripts =
     check-byte-order-marker = pre_commit_hooks.check_byte_order_marker:main
     check-case-conflict = pre_commit_hooks.check_case_conflict:main
     check-docstring-first = pre_commit_hooks.check_docstring_first:main
+    check-encoding = pre_commit_hooks.check_encoding:main
     check-executables-have-shebangs = pre_commit_hooks.check_executables_have_shebangs:main
     check-json = pre_commit_hooks.check_json:main
     check-merge-conflict = pre_commit_hooks.check_merge_conflict:main

tests/check_encoding_test.py

@@ -0,0 +1,17 @@
+import pytest
+
+from pre_commit_hooks.check_encoding import main
+
+
+@pytest.mark.parametrize(
+    ('content', 'encoding', 'expected'),
+    (
+        (b'Hello!', 'ascii', 0),
+        (b'Hello!', 'unknown-encoding', 2),
+        ('Hello ☃!'.encode(), 'ascii', 1),
+    ),
+)
+def test_has_encoding(content, encoding, expected, tmpdir):
+    path = tmpdir.join('path')
+    path.write(content, 'wb')
+    assert main(('--encoding', encoding, str(path))) == expected