Add test case as example of Git commit with invalid data

nvie · nvie · commit 0eafe201905d · 2016-05-30T16:32:26.000+02:00
This is a real commit from the microjs.com open source project, see microjs/microjs.com@7e8457c, which is declared to be encoded in UTF-8, but contains invalid bytes. This makes GitPython choke on it while decoding. Rather than choking, this should instead accept the error and replace the invalid bytes by the � (\x80) char.
diff --git a/git/test/fixtures/commit_invalid_data b/git/test/fixtures/commit_invalid_data
@@ -0,0 +1,6 @@
+tree 9f1a495d7d9692d24f5caedaa89f5c2c32d59368
+parent 492ace2ffce0e426ebeb55e364e987bcf024dd3b
+author E.Azer Ko�o�o�oculu <azer@kodfabrik.com> 1306710073 +0300
+committer E.Azer Ko�o�o�oculu <azer@kodfabrik.com> 1306710073 +0300
+
+add environjs
diff --git a/git/test/test_commit.py b/git/test/test_commit.py
@@ -306,6 +306,13 @@ def test_serialization_unicode_support(self):
         # it appears
         cmt.author.__repr__()
 
+    def test_invalid_commit(self):
+        cmt = self.rorepo.commit()
+        cmt._deserialize(open(fixture_path('commit_invalid_data'), 'rb'))
+
+        assert cmt.author.name == u'E.Azer Ko�o�o�oculu', cmt.author.name
+        assert cmt.author.email == 'azer@kodfabrik.com', cmt.author.email
+
     def test_gpgsig(self):
         cmt = self.rorepo.commit()
         cmt._deserialize(open(fixture_path('commit_with_gpgsig'), 'rb'))
