Match complete CI paths, including newline filenames

scripts/check_selected_ci_inputs.py

@@ -250,7 +250,8 @@ def glob_regex(pattern):
             require(pattern[i] not in '?+[]{}', 'Unsupported glob syntax: ' + pattern)
             out.append(re.escape(pattern[i]))
             i += 1
-    return re.compile('^' + ''.join(out) + '$')
+    # Paths match in full; globstars include LF, which Git permits in filenames.
+    return re.compile(r'\A' + ''.join(out) + r'\Z', re.DOTALL)
 
 
 def triggers(patterns, name):

tests/test_selected_ci_inputs.py

@@ -4,6 +4,7 @@
 import importlib.util
 import io
 import json
+import os
 from pathlib import Path
 import shutil
 import subprocess
@@ -341,6 +342,23 @@ def test_github_path_filter_semantics(self):
             with self.subTest(patterns=patterns, name=name):
                 self.assertEqual(guard.triggers(patterns, name), expected)
 
+    def test_github_path_filters_match_whole_paths_including_newlines(self):
+        cases = [
+            (['proof.md'], 'proof.md\n', False),
+            (['proof.md\n'], 'proof.md\n', True),
+            (['supplier/**'], 'supplier/line\nbreak.md', True),
+            (['supplier/**/proof.md'], 'supplier/line\nbreak/proof.md', True),
+            (['supplier/**/proof.md'], 'supplier/proof.md', True),
+            (['supplier/*'], 'supplier/line\nbreak.md', True),
+            (['supplier/*'], 'supplier/line\nbreak/proof.md', False),
+            (['supplier/**', '!supplier/*'], 'supplier/line\nbreak.md', False),
+            (['supplier/**', '!supplier/**', 'supplier/*'], 'supplier/line\nbreak.md', True),
+            (['supplier/**', '!supplier/proof.md'], 'supplier/proof.md\n', True),
+        ]
+        for patterns, name, expected in cases:
+            with self.subTest(patterns=patterns, name=name):
+                self.assertEqual(guard.triggers(patterns, name), expected)
+
     def test_unsupported_syntax_fails_closed(self):
         for pattern in ('research/[ab]/**', 'scripts/verify_?.py', 'research/CORE-23+.json', 'a/{b,c}'):
             with self.subTest(pattern=pattern), self.assertRaises(ValueError):
@@ -356,7 +374,7 @@ def test_event_reader(self):
         self.assertIsNone(guard.event_paths(text, 'schedule'))
 
 
-class BranchFilterCLI(unittest.TestCase):
+class TriggerFilterCLI(unittest.TestCase):
     def setUp(self):
         temporary = tempfile.TemporaryDirectory()
         self.addCleanup(temporary.cleanup)
@@ -408,6 +426,53 @@ def test_excluded_main_fails_only_when_triggers_are_required(self):
                         else:
                             self.assertNotIn('ADVISORY:', result.stdout)
 
+    @unittest.skipIf(os.name == 'nt', 'Windows filenames cannot contain LF')
+    def test_newline_paths_through_the_real_cli(self):
+        command = [sys.executable, '-B', str(ROOT / 'scripts/check_selected_ci_inputs.py'),
+                   '--root', str(self.root)]
+        workflow = self.root / '.github/workflows/replay.yml'
+        ordinary = ['scripts/**', 'certificates/**', 'certificate.json', 'manifest.json',
+                    '.github/workflows/**', 'supplier/proof.md']
+        original_manifest = (self.root / 'manifest.json').read_bytes()
+        cases = [
+            ('supplier/proof.md\n', 'supplier/proof.md', False),
+            ('supplier/line\nbreak.md', 'supplier/**', True),
+            ('supplier/line\nbreak/proof.md', 'supplier/**/proof.md', True),
+        ]
+        for name, pattern, covered in cases:
+            path = self.root / name
+            path.parent.mkdir(parents=True, exist_ok=True)
+            path.write_bytes(b'inventoried input\n')
+            manifest = json.loads(original_manifest)
+            manifest['files'][name] = sha256(path.read_bytes()).hexdigest()
+            (self.root / 'manifest.json').write_text(json.dumps(manifest), encoding='utf-8')
+            try:
+                for event in guard.EVENTS:
+                    text = 'on:\n'
+                    for kind in guard.EVENTS:
+                        text += '  ' + kind + ':\n'
+                        if kind == event:
+                            text += '    paths: ' + json.dumps(ordinary + [pattern]) + '\n'
+                    text += 'jobs:\n  replay:\n    steps:\n      - run: python scripts/replay.py\n'
+                    workflow.write_text(text, encoding='utf-8')
+                    for required in (False, True):
+                        with self.subTest(name=name, event=event, required=required):
+                            result = subprocess.run(command + (['--require-triggers'] if required else []),
+                                                    check=False, capture_output=True, text=True, timeout=30)
+                            self.assertEqual(result.returncode, int(required and not covered),
+                                             result.stdout + result.stderr)
+                            report, _ = json.JSONDecoder().raw_decode(result.stdout)
+                            self.assertEqual(report['record_failures'], {})
+                            self.assertIsNone(report['coverage_error'])
+                            self.assertEqual(report['uncovered'],
+                                             {kind: [name] if kind == event and not covered else []
+                                              for kind in guard.EVENTS})
+                            marker = 'FAIL_SELECTED_CI_INPUTS' if required and not covered else 'PASS_SELECTED_CI_INPUTS'
+                            self.assertEqual(result.stdout.splitlines()[-1], marker)
+            finally:
+                path.unlink()
+                (self.root / 'manifest.json').write_bytes(original_manifest)
+
 
 if __name__ == '__main__':
     unittest.main()