Cannot handle non-ascii strings in source code

[Henning Eggers]

Pocket-lint trips over source files with non-ascii characters in it. The example contained "português". It probably needs to honor PEP 0263.

Traceback (most recent call last):
  File "/usr/bin/pocketlint", line 6, in <module>
    sys.exit(main(sys.argv))
  File "/usr/lib/pymodules/python2.7/pocketlint/formatcheck.py", line 557, in main
    check_sources(sources, reporter)
  File "/usr/lib/pymodules/python2.7/pocketlint/formatcheck.py", line 542, in check_sources
    checker.check()
  File "/usr/lib/pymodules/python2.7/pocketlint/formatcheck.py", line 224, in check
    PythonChecker(self.file_path, self.text, self._reporter).check()
  File "/usr/lib/pymodules/python2.7/pocketlint/formatcheck.py", line 416, in check
    self.check_text()
  File "/usr/lib/pymodules/python2.7/pocketlint/formatcheck.py", line 465, in check_text
    self.check_ascii(line_no, line)
  File "/usr/lib/pymodules/python2.7/pocketlint/formatcheck.py", line 479, in check_ascii
    ascii_line = line.encode('ascii')
UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 29: ordinal not in range(128)

[Curtis Hovey (sinzui)]

which kind of source file? The Python checker does do an encoding check...ascii, or the encoding specified at the start of the file? This rule also applies to doctests.

[Curtis Hovey (sinzui)]

If this is a python or doctest file, can you paste the encoding line, eg:
    # -*-

[Henning Eggers (henninge)]

It was python file and its first line was:
# coding=utf-8

That is what PEP 0263 requires. Alternatively it allows "formats recognized by popular editors", one of which is the -*- you suggested. So I guess the bug is that not all formats allowed by PEP 0263 are supported.

I'll have a look at the source code.