Unicode support

[pelson]

Follows up from #134 (so that needs merging first) to allow unicode units, such as π m², as is supported by udunits2.

In addition to this, I've extended the coding standard test to handle an encoding preamble.

Closes #133.

[coveralls]

Coverage decreased (-1.4%) to 88.511% when pulling 5df834a on pelson:unicode_support into 18b72bc on SciTools:master.

[pelson]

Interestingly the failing test doesn't fail for me locally on py27... Will push a commit shortly.

[pelson]

Tempting to put a six.PY2 test here.

[pelson]

Have done that @bjlittle

[bjlittle]

Nice...

[bjlittle]

@pelson Looks good to me... I'll let travis do it's thing, then merge 👍

[pelson]

@bjlittle - thanks for merging #134. Have rebased (sorry if it has moved under your feet 😉)

[bjlittle]

@pelson No worries... I was expecting the rebase 😉

[pelson]

OK, so something isn't quite right yet for py2k...

>>> import cf_units
>>> cf_units.Unit("m²")
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
UnicodeEncodeError: 'ascii' codec can't encode character u'\xb2' in position 7: ordinal not in range(128)
>>> cf_units.Unit(u"m²")
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "cf_units/__init__.py", line 810, in __init__
    unit = str(unit).strip()
UnicodeEncodeError: 'ascii' codec can't encode character u'\xb2' in position 1: ordinal not in range(128)

[bjlittle]

@pelson You might want to account for white space after the trailing -*- ...?

[pelson]

Meh. Happy if you want me to, but why be generous on this? Do it right, or fix it is my opinion 😉

[bjlittle]

@pelson I'll let you chase down the py2k unicode barf... ping me when you want me to re-review 😉

[bjlittle]

Nice, thanks @pelson 👍

[pelson]

What a right old shambles py2's unicode handling is. I think I've now got this right. From the start, this was a trivial change for py3k - all of the work (and there has now been several hours worth) has been dealing with the py2 fallout.

With the current setup, the following will occur for py2 users:

>>> import cf_units
>>> cf_units.Unit(u'ø')
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "cf_units/__init__.py", line 851, in __init__
    self._propogate_error('Failed to parse unit "%s"' % str_unit, e)
  File "cf_units/__init__.py", line 886, in _propogate_error
    raise ValueError('[%s] %s%s' % (ud_err.status_msg(), msg, error_msg))
ValueError: [UT_UNKNOWN] Failed to parse unit "?"

>>> u = cf_units.Unit(u'π')
>>> u
Unit('?')
>>> u.symbol
'3.14159265358979 1'
>>> u.origin
u'\u03c0'

I think I can do better than that though, so there is another PR incoming.

[pelson]

As promised #137.