Don't lint binary files (#10)

src/rapids_pre_commit_hooks/lint.py

@@ -17,12 +17,17 @@
 import contextlib
 import functools
 import itertools
+import warnings
 
 
 class OverlappingReplacementsError(RuntimeError):
     pass
 
 
+class BinaryFileWarning(Warning):
+    pass
+
+
 class Replacement:
     def __init__(self, pos, newtext):
         self.pos = pos
@@ -206,11 +211,18 @@ def __exit__(self, exc_type, exc_value, traceback):
         if exc_type:
             return
 
-        warnings = False
+        has_warnings = False
 
         for file in self.args.files:
             with open(file) as f:
-                content = f.read()
+                try:
+                    content = f.read()
+                except UnicodeDecodeError:
+                    warnings.warn(
+                        f"Refusing to run text linter on binary file {file}.",
+                        BinaryFileWarning,
+                    )
+                    continue
 
             linter = Linter(file, content)
             for check in self.checks:
@@ -224,9 +236,9 @@ def __exit__(self, exc_type, exc_value, traceback):
                         f.write(fix)
 
             if len(linter.warnings) > 0:
-                warnings = True
+                has_warnings = True
 
-        if warnings:
+        if has_warnings:
             exit(1)
 
 

test/rapids_pre_commit_hooks/test_lint.py

@@ -13,11 +13,16 @@
 # limitations under the License.
 
 import tempfile
-from unittest.mock import patch
+from unittest.mock import Mock, patch
 
 import pytest
 
-from rapids_pre_commit_hooks.lint import Linter, LintMain, OverlappingReplacementsError
+from rapids_pre_commit_hooks.lint import (
+    BinaryFileWarning,
+    Linter,
+    LintMain,
+    OverlappingReplacementsError,
+)
 
 
 class TestLinter:
@@ -129,6 +134,14 @@ def hello_file(self):
             f.seek(0)
             yield f
 
+    @pytest.fixture
+    def binary_file(self):
+        with tempfile.NamedTemporaryFile("wb+") as f:
+            f.write(b"\xDE\xAD\xBE\xEF")
+            f.flush()
+            f.seek(0)
+            yield f
+
     def the_check(self, linter, args):
         assert args.check_test
         linter.add_warning((0, 5), "say good bye instead").add_replacement(
@@ -281,3 +294,23 @@ def test_multiple_files(self, hello_world_file, hello_file, capsys):
 
 """
         )
+
+    def test_binary_file(self, binary_file, capsys):
+        mock_linter = Mock(wraps=Linter)
+        with patch(
+            "sys.argv",
+            [
+                "check-test",
+                "--check-test",
+                "--fix",
+                binary_file.name,
+            ],
+        ), patch("rapids_pre_commit_hooks.lint.Linter", mock_linter), pytest.warns(
+            BinaryFileWarning,
+            match=r"^Refusing to run text linter on binary file .*\.$",
+        ):
+            m = LintMain()
+            m.argparser.add_argument("--check-test", action="store_true")
+            with m.execute() as ctx:
+                ctx.add_check(self.the_check)
+        mock_linter.assert_not_called()