Merge lp:~adiroiban/pocket-lint/bug-889648 into lp:pocket-lint

[Adi Roiban]

Hi,

This is a stupid code that I have added just to start the discussion about how this should be solved.

The problem is that when calling pocket-lint using 'check_sources', PythonChecker will always receive an ascii text.
In PythonChecker tests, it is passed both ascii and unicode texts.

This inconsistency does not allow using sometihg line 'line = line.decode(self.encoding)' and the line lenght 'smart' check is hardcoded to unicode.

Cheers,
Adi

[Curtis Hovey (sinzui)]

Line 52 should read
    full_text = encoding_line + line_79_chars

I have fixed this in my merge.

review: Approve (code)

=== modified file 'pocketlint/formatcheck.py'
--- pocketlint/formatcheck.py	2011-09-10 14:53:27 +0000
+++ pocketlint/formatcheck.py	2011-11-13 00:00:29 +0000
@@ -2,6 +2,7 @@
 # Copyright (C) 2009-2011 - Curtis Hovey <sinzui.is at verizon.net>
 # This software is licensed under the MIT license (see the file COPYING).
 """Check for syntax and style problems."""
+from __future__ import with_statement
 
 __metaclass__ = type
 
@@ -527,6 +528,15 @@
             self.message(
                 line_no, 'Line contains a call to pdb.', icon='error')
 
+    def check_length(self, line_no, line):
+        # isinstance is checked since 'test_code_utf8' is passing an unicode
+        # text to PythonChecker, but check_sources always opens the file in
+        # ascii mode, so PythonChecker alwasy received the text as ascii
+        # encoded.
+        if self.encoding == 'utf-8' and not isinstance(line, unicode):
+            line = line.decode('utf-8')
+        super(PythonChecker, self).check_length(line_no, line)
+
     def check_ascii(self, line_no, line):
         """Check that the line is ascii."""
         if self.encoding != 'ascii':

=== modified file 'pocketlint/tests/test_python.py'
--- pocketlint/tests/test_python.py	2011-05-04 15:39:13 +0000
+++ pocketlint/tests/test_python.py	2011-11-13 00:00:29 +0000
@@ -256,3 +256,26 @@
         self.assertEqual(
             [(1, 'Non-ascii characer at position 21.')],
             self.reporter.messages)
+
+    def test_length_unicode_ok(self):
+        encoding_line = '# -*- coding: utf-8 -*-\n'
+        unicode_4_chars = 'mi\xc8\x9bi'
+        line_78_chars = 'pa' + unicode_4_chars * 19 + '\n'
+        full_text =  encoding_line + line_78_chars
+
+        checker = PythonChecker('bogus',full_text, self.reporter)
+        checker.check_text()
+
+        self.assertEqual([], self.reporter.messages)
+
+    def test_length_unicode_bad(self):
+        encoding_line = '# -*- coding: utf-8 -*-\n'
+        unicode_4_chars = 'mi\xc8\x9bi'
+        line_79_chars = 'pap' + unicode_4_chars * 19 + '\n'
+        full_text =  encoding_line + line_78_chars
+
+        checker = PythonChecker('bogus',full_text, self.reporter)
+        checker.check_text()
+
+        self.assertEqual(
+            [(2, 'Line exceeds 78 characters.')], self.reporter.messages)
\ No newline at end of file