Don't trace non-encodable file names. #891

CHANGES.rst

@@ -30,13 +30,19 @@ Unreleased
   information about the config files read now shows absolute paths to the
   files.
 
+- The handling of source files with non-encodable file names has changed.
+  Previously, if a file name could not be encoded as UTF-8, an error occurred,
+  as described in `issue 891`_.  Now, those files will not be measured, since
+  their data would not be recordable.
+
 - A new warning ("dynamic-conflict") is issued if two mechanisms are trying to
   change the dynamic context.  Closes `issue 901`_.
 
 - ``coverage run --debug=sys`` would fail with an AttributeError. This is now
   fixed (`issue 907`_).
 
 .. _issue 890: https://github.com/nedbat/coveragepy/issues/890
+.. _issue 891: https://github.com/nedbat/coveragepy/issues/891
 .. _issue 901: https://github.com/nedbat/coveragepy/issues/901
 .. _issue 907: https://github.com/nedbat/coveragepy/issues/907
 

coverage/inorout.py

@@ -18,7 +18,7 @@
 from coverage.disposition import FileDisposition, disposition_init
 from coverage.files import TreeMatcher, FnmatchMatcher, ModuleMatcher
 from coverage.files import prep_patterns, find_python_files, canonical_filename
-from coverage.misc import CoverageException
+from coverage.misc import CoverageException, contract
 from coverage.python import source_for_file, source_for_morf
 
 
@@ -296,6 +296,7 @@ def nope(disp, reason):
 
         return disp
 
+    @contract(filename='unicode')
     def check_include_omit_etc(self, filename, frame):
         """Check a file name against the include, omit, etc, rules.
 
@@ -333,6 +334,12 @@ def check_include_omit_etc(self, filename, frame):
         if self.omit_match and self.omit_match.match(filename):
             return "is inside an --omit pattern"
 
+        # No point tracing a file we can't later write to SQLite.
+        try:
+            filename.encode("utf8")
+        except UnicodeEncodeError:
+            return "non-encodable filename"
+
         # No reason found to skip this file.
         return None
 

tests/test_oddball.py

@@ -3,6 +3,7 @@
 
 """Oddball cases for testing coverage.py"""
 
+import os.path
 import sys
 
 from flaky import flaky
@@ -562,6 +563,18 @@ def test_correct_filename(self):
         self.assertEqual(statements, [31])
         self.assertEqual(missing, [])
 
+    def test_unencodable_filename(self):
+        # https://github.com/nedbat/coveragepy/issues/891
+        if env.PYVERSION < (3, 0):
+            self.skipTest("Python 2 can't seem to compile the file.")
+        self.make_file("bug891.py", r"""exec(compile("pass", "\udcff.py", "exec"))""")
+        cov = coverage.Coverage()
+        self.start_import_stop(cov, "bug891")
+        # Saving would fail trying to encode \udcff.py
+        cov.save()
+        files = [os.path.basename(f) for f in cov.get_data().measured_files()]
+        assert "bug891.py" in files
+
 
 class MockingProtectionTest(CoverageTest):
     """Tests about protecting ourselves from aggressive mocking.