Add python 3 support to cpplint and cpplint_unittest

cpplint/cpplint.py

@@ -53,6 +53,8 @@ import sys
 import unicodedata
 import sysconfig
 
+import six
+
 try:
   xrange          # Python 2
 except NameError:
@@ -952,7 +954,7 @@ class _CppLintState(object):
 
   def PrintErrorCounts(self):
     """Print a summary of errors by category, and the total."""
-    for category, count in self.errors_by_category.iteritems():
+    for category, count in six.iteritems(self.errors_by_category):
       sys.stderr.write('Category \'%s\' errors found: %d\n' %
                        (category, count))
     sys.stdout.write('Total errors found: %d\n' % self.error_count)
@@ -4286,7 +4288,7 @@ def GetLineWidth(line):
     The width of the line in column positions, accounting for Unicode
     combining characters and wide characters.
   """
-  if isinstance(line, unicode):
+  if isinstance(line, six.text_type):
     width = 0
     for uc in unicodedata.normalize('NFC', line):
       if unicodedata.east_asian_width(uc) in ('W', 'F'):
@@ -4622,7 +4624,7 @@ def _GetTextInside(text, start_pattern):
 
   # Give opening punctuations to get the matching close-punctuations.
   matching_punctuation = {'(': ')', '{': '}', '[': ']'}
-  closing_punctuation = set(matching_punctuation.itervalues())
+  closing_punctuation = set(six.itervalues(matching_punctuation))
 
   # Find the position to start extracting text.
   match = re.search(start_pattern, text, re.M)
@@ -5570,7 +5572,7 @@ def CheckForIncludeWhatYouUse(filename, clean_lines, include_state, error,
 
   # include_dict is modified during iteration, so we iterate over a copy of
   # the keys.
-  header_keys = include_dict.keys()
+  header_keys = list(six.iterkeys(include_dict))
   for header in header_keys:
     (same_module, common_path) = FilesBelongToSameModule(abs_filename, header)
     fullpath = common_path + header
@@ -6225,10 +6227,13 @@ def main():
 
   # Change stderr to write with replacement characters so we don't die
   # if we try to print something containing non-ASCII characters.
-  sys.stderr = codecs.StreamReaderWriter(sys.stderr,
+  # https://docs.python.org/3/library/sys.html#sys.stderr
+  sys.stderr = codecs.StreamReaderWriter(
+      sys.stderr if sys.version_info.major < 3 else sys.stderr.buffer,
       codecs.getreader('utf8'),
       codecs.getwriter('utf8'),
-                                         'replace')
+      'replace'
+      )
 
   _cpplint_state.ResetErrorCounts()
   for filename in filenames:

cpplint/cpplint_unittest.py

@@ -3071,7 +3071,7 @@ class CpplintTest(CpplintTestBase):
       error_collector = ErrorCollector(self.assert_)
       cpplint.ProcessFileData(
           'foo.cc', 'cc',
-          unicode(raw_bytes, 'utf8', 'replace').split('\n'),
+          raw_bytes.decode('utf-8', errors='replace').split('\n'),
           error_collector)
       # The warning appears only once.
       self.assertEquals(
@@ -3081,12 +3081,12 @@ class CpplintTest(CpplintTestBase):
               ' (or Unicode replacement character).'
               '  [readability/utf8] [5]'))
 
-    DoTest(self, 'Hello world\n', False)
+    DoTest(self, b'Hello world\n', False)
-    DoTest(self, '\xe9\x8e\xbd\n', False)
+    DoTest(self, b'\xe9\x8e\xbd\n', False)
-    DoTest(self, '\xe9x\x8e\xbd\n', True)
+    DoTest(self, b'\xe9x\x8e\xbd\n', True)
     # This is the encoding of the replacement character itself (which
     # you can see by evaluating codecs.getencoder('utf8')(u'\ufffd')).
-    DoTest(self, '\xef\xbf\xbd\n', True)
+    DoTest(self, b'\xef\xbf\xbd\n', True)
 
   def testBadCharacters(self):
     # Test for NUL bytes only
@@ -3104,7 +3104,7 @@ class CpplintTest(CpplintTestBase):
     cpplint.ProcessFileData(
         'nul_utf8.cc', 'cc',
         ['// Copyright 2014 Your Company.',
-         unicode('\xe9x\0', 'utf8', 'replace'), ''],
+        b'\xe9x\0'.decode('utf-8', errors='replace'), ''],
         error_collector)
     self.assertEquals(
         error_collector.Results(),
@@ -5723,8 +5723,9 @@ class QuietTest(unittest.TestCase):
 
   def testNonQuietWithErrors(self):
     # This will fail: the test header is missing a copyright and header guard.
-    (return_code, output) = self._runCppLint()
+    (return_code, output_bytes) = self._runCppLint()
     self.assertEquals(1, return_code)
+    output = output_bytes.decode('utf-8')
     # Always-on behavior: Print error messages as they come up.
     self.assertIn("[legal/copyright]", output)
     self.assertIn("[build/header_guard]", output)
@@ -5734,7 +5735,8 @@ class QuietTest(unittest.TestCase):
 
   def testQuietWithErrors(self):
     # When there are errors, behavior is identical to not passing --quiet.
-    (return_code, output) = self._runCppLint('--quiet')
+    (return_code, output_bytes) = self._runCppLint('--quiet')
+    output = output_bytes.decode('utf-8')
     self.assertEquals(1, return_code)
     self.assertIn("[legal/copyright]", output)
     self.assertIn("[build/header_guard]", output)
@@ -5744,9 +5746,10 @@ class QuietTest(unittest.TestCase):
 
   def testNonQuietWithoutErrors(self):
     # This will succeed. We filtered out all the known errors for that file.
-    (return_code, output) = self._runCppLint('--filter=' +
+    (return_code, output_bytes) = self._runCppLint('--filter=' +
                                                    '-legal/copyright,' +
                                                    '-build/header_guard')
+    output = output_bytes.decode('utf-8')
     self.assertEquals(0, return_code, output)
     # No cpplint errors are printed since there were no errors.
     self.assertNotIn("[legal/copyright]", output)
@@ -5758,10 +5761,11 @@ class QuietTest(unittest.TestCase):
 
   def testQuietWithoutErrors(self):
     # This will succeed. We filtered out all the known errors for that file.
-    (return_code, output) = self._runCppLint('--quiet',
+    (return_code, output_bytes) = self._runCppLint('--quiet',
                                                    '--filter=' +
                                                    '-legal/copyright,' +
                                                    '-build/header_guard')
+    output = output_bytes.decode('utf-8')
     self.assertEquals(0, return_code, output)
     # No cpplint errors are printed since there were no errors.
     self.assertNotIn("[legal/copyright]", output)