Add E265 for space before block comment; issue #190

Author: florentx

CHANGES.txt

@@ -22,6 +22,8 @@ Changelog
 
 * Fix false positive E261/E262 when the file contains a BOM. (Issue #193)
 
+* Add E265 for space before block comment. (Issue #190)
+
 
 1.4.5 (2013-03-06)
 ------------------

docs/intro.rst

@@ -252,6 +252,8 @@ This is the current list of error and warning codes:
 +----------+----------------------------------------------------------------------+
 | E262     | inline comment should start with '# '                                |
 +----------+----------------------------------------------------------------------+
+| E265     | block comment should start with '# '                                 |
++----------+----------------------------------------------------------------------+
 +----------+----------------------------------------------------------------------+
 | E271     | multiple spaces after keyword                                        |
 +----------+----------------------------------------------------------------------+

pep8.py

@@ -764,31 +764,41 @@ def whitespace_around_named_parameter_equals(logical_line, tokens):
         prev_end = end
 
 
-def whitespace_before_inline_comment(logical_line, tokens):
+def whitespace_before_comment(logical_line, tokens):
     """
     Separate inline comments by at least two spaces.
 
     An inline comment is a comment on the same line as a statement.  Inline
     comments should be separated by at least two spaces from the statement.
     They should start with a # and a single space.
 
+    Each line of a block comment starts with a # and a single space
+    (unless it is indented text inside the comment).
+
     Okay: x = x + 1  # Increment x
     Okay: x = x + 1    # Increment x
+    Okay: # Block comment
     E261: x = x + 1 # Increment x
     E262: x = x + 1  #Increment x
     E262: x = x + 1  #  Increment x
+    E265: #Block comment
     """
     prev_end = (0, 0)
     for token_type, text, start, end, line in tokens:
         if token_type == tokenize.COMMENT:
-            if not line[:start[1]].strip():
-                continue
-            if prev_end[0] == start[0] and start[1] < prev_end[1] + 2:
-                yield (prev_end,
-                       "E261 at least two spaces before inline comment")
+            inline_comment = line[:start[1]].strip()
+            if inline_comment:
+                if prev_end[0] == start[0] and start[1] < prev_end[1] + 2:
+                    yield (prev_end,
+                           "E261 at least two spaces before inline comment")
             symbol, sp, comment = text.partition(' ')
-            if symbol not in ('#', '#:') or comment[:1].isspace():
-                yield start, "E262 inline comment should start with '# '"
+            bad_prefix = symbol not in ('#', '#:')
+            if inline_comment:
+                if bad_prefix or comment[:1].isspace():
+                    yield start, "E262 inline comment should start with '# '"
+            elif bad_prefix:
+                if text.rstrip('#') and (start[0] > 1 or symbol[1] != '!'):
+                    yield start, "E265 block comment should start with '# '"
         elif token_type != tokenize.NL:
             prev_end = end
 
@@ -1307,9 +1317,9 @@ def check_logical(self):
         """
         self.build_tokens_line()
         self.report.increment_logical_line()
-        first_line = self.lines[self.mapping[0][1][2][0] - 1]
-        indent = first_line[:self.mapping[0][1][2][1]]
-        self.previous_indent_level = self.indent_level
+        token0 = self.mapping[0][1] if self.mapping else self.tokens[0]
+        first_line = self.lines[token0[2][0] - 1]
+        indent = first_line[:token0[2][1]]
         self.indent_level = expand_indent(indent)
         if self.verbose >= 2:
             print(self.logical_line[:80].rstrip())
@@ -1321,12 +1331,17 @@ def check_logical(self):
                 if isinstance(offset, tuple):
                     orig_number, orig_offset = offset
                 else:
+                    orig_number = token0[2][0]
+                    orig_offset = token0[2][1] + offset
                     for token_offset, token in self.mapping:
                         if offset >= token_offset:
                             orig_number = token[2][0]
                             orig_offset = (token[2][1] + offset - token_offset)
                 self.report_error(orig_number, orig_offset, text, check)
-        self.previous_logical = self.logical_line
+        if self.logical_line:
+            self.previous_indent_level = self.indent_level
+            self.previous_logical = self.logical_line
+        self.tokens = []
 
     def check_ast(self):
         try:
@@ -1359,6 +1374,7 @@ def check_all(self, expected=None, line_offset=0):
         self.line_number = 0
         self.indent_char = None
         self.indent_level = 0
+        self.previous_indent_level = 0
         self.previous_logical = ''
         self.tokens = []
         self.blank_lines = blank_lines_before_comment = 0
@@ -1383,20 +1399,21 @@ def check_all(self, expected=None, line_offset=0):
                     if self.blank_lines < blank_lines_before_comment:
                         self.blank_lines = blank_lines_before_comment
                     self.check_logical()
-                    self.tokens = []
                     self.blank_lines = blank_lines_before_comment = 0
                 elif token_type == tokenize.NL:
                     if len(self.tokens) == 1:
                         # The physical line contains only this token.
                         self.blank_lines += 1
-                    self.tokens = []
+                        del self.tokens[0]
+                    else:
+                        self.check_logical()
                 elif token_type == tokenize.COMMENT and len(self.tokens) == 1:
                     if blank_lines_before_comment < self.blank_lines:
                         blank_lines_before_comment = self.blank_lines
                     self.blank_lines = 0
                     if COMMENT_WITH_NL:
                         # The comment also ends a physical line
-                        self.tokens = []
+                        self.check_logical()
         return self.report.get_file_results()
 
 

testsuite/E11.py

@@ -10,3 +10,7 @@
 #: E113
 print
     print
+#: E111 E113
+mimetype = 'application/x-directory'
+     # 'httpd/unix-directory'
+create_date = False

testsuite/test_api.py

@@ -124,15 +124,15 @@ def test_styleguide(self):
         report = pep8.StyleGuide().check_files([E11])
         stdout = sys.stdout.getvalue().splitlines()
         self.assertEqual(len(stdout), report.total_errors)
-        self.assertEqual(report.total_errors, 4)
+        self.assertEqual(report.total_errors, 6)
         self.assertFalse(sys.stderr)
         self.reset()
 
         # Passing the paths in the constructor gives same result
         report = pep8.StyleGuide(paths=[E11]).check_files()
         stdout = sys.stdout.getvalue().splitlines()
         self.assertEqual(len(stdout), report.total_errors)
-        self.assertEqual(report.total_errors, 4)
+        self.assertEqual(report.total_errors, 6)
         self.assertFalse(sys.stderr)
         self.reset()
 

testsuite/test_shell.py

@@ -75,7 +75,7 @@ def test_check_simple(self):
         stdout = stdout.splitlines()
         self.assertEqual(errcode, 1)
         self.assertFalse(stderr)
-        self.assertEqual(len(stdout), 4)
+        self.assertEqual(len(stdout), 6)
         for line, num, col in zip(stdout, (3, 6, 9, 12), (3, 6, 1, 5)):
             path, x, y, msg = line.split(':')
             self.assertTrue(path.endswith(E11))