handle unicode encoding errors

[bmschwa]

On line 368, if we fail to convert a unicode string - replace the unicode chars

[foutaise]

Hi Bryan,

Could you please provide an example in order to reproduce
the problem, and details about the version of python you're using please.

I'll have to improve this part of the code obviously, cause it does not handle
unicode literals under python2. I already have a fix in mind, but I'd like to have
details about your specific case.

As a reminder for myself, here goes an example that does not work under python2 right now:

$ cat test.py
# coding=utf-8

import texttable

table = texttable.Texttable()
table.add_rows([
    ["Name", "age"],
    [u'Éric', 12]])
print(table.draw())

Gives the following error under python2:

$ python2 test.py
Traceback (most recent call last):
  File "test.py", line 8, in <module>
    [u'Éric', 12]])
  File "/usr/home/jef/git/src/github.com/foutaise/texttable/texttable.py", line 331, in add_rows
    self.add_row(row)
  File "/usr/home/jef/git/src/github.com/foutaise/texttable/texttable.py", line 310, in add_row
    cells.append(self._str(i, x))
  File "/usr/home/jef/git/src/github.com/foutaise/texttable/texttable.py", line 369, in _str
    return str(x)
UnicodeEncodeError: 'ascii' codec can't encode character u'\xc9' in position 0: ordinal not in range(128)

Thanks.

[foutaise]

Released version 0.8.5 to fix those issues.