libmyth-python: Issue with national characters in filename

[peterdv]

In MythTV/dataheap.py national characters in filename
causes an UnicodeEncodeError to be thrown in Video.open(),
in the statement:

        return ftopen('myth://Videos@{0.host}/{0.filename}'.format(self),
                    mode, False, nooverwrite, self._db)

For me the problem can be fixed by ensuring that
the filename format string is unicode:

        return ftopen(u'myth://Videos@{0.host}/{0.filename}'.format(self),
                    mode, False, nooverwrite, self._db)

The problem was actually experienced in mythvidexport.py,
from https://github.com/wagnerrp/mythtv-scripts,
as modified in commit 0c6ebb4b78dd398a9d9d329b7aeafa314f40c545 of
the fork https://github.com/spmorton/mythtv-scripts.
The system had MythTV/dataheap.py installed from the
Ubuntu repositories from the package
libmyth-python 2:0.26.0+fixes.2012

mythtv@media02:/var/log/mythtv$ /usr/share/mythtv/mythvidexport.py --chanid=1101 --starttime=1328386500
2013-05-22 17:50:59.805274 I [16429] mythvidexport.py Using recording -- Wallander: Lækken - Svensk krimi fra 2009.Brændende biler og afspærrede motorveje. Ystad rammes af en række spektakulære røverier mod pengetransport
2013-05-22 17:50:59.824353 I [16429] mythvidexport.py Attempting TV export.
2013-05-22 17:51:00.014421 I [16429] mythvidexport.py Falling back to generic export.
2013-05-22 17:51:00.026309 I [16429] mythvidexport.py Import complete
2013-05-22 17:51:00.028494 I [16429] mythvidexport.py Copying myth://Default@media02/1101_20120204211500.mpg to myth://Videos@media02/Videos/Wallander: Lækken.mpg
2013-05-22 17:51:00.051557 I [16429] mythvidexport.py Checking for duplication of -- Wallander: Lækken - Svensk krimi fra 2009.Brændende biler og afspærrede motorveje. Ystad rammes af en række spektakulære røverier mod pengetransport
Traceback (most recent call last):
Traceback (most recent call last):
  File "/usr/share/mythtv/mythvidexport.py", line 447, in <module>
    main()
  File "/usr/share/mythtv/mythvidexport.py", line 419, in main
    export = VIDEO(opts)
  File "/usr/share/mythtv/mythvidexport.py", line 75, in __init__
    self.copy()
  File "/usr/share/mythtv/mythvidexport.py", line 246, in copy
    dstfp = self.vid.open('w')
  File "/usr/lib/python2.7/dist-packages/MythTV/dataheap.py", line 931, in open
    return ftopen('myth://Videos@{0.host}/{0.filename}'.format(self),
UnicodeEncodeError: 'ascii' codec can't encode character u'\xe6' in position 19: ordinal not in range(128)
mythtv@media02:/var/log/mythtv$ printenv LANG
en_DK.UTF-8

Starting off in:
  "Software should only work with Unicode strings internally,
   converting to a particular encoding on output."

And noting that: "When it comes to unicode filenames in python,
on Unix systems, there will only be a filesystem encoding
if you’ve set the LANG or LC_CTYPE environment variables;
if you haven’t, the default encoding is ASCII.
When opening a file for reading or writing,
you can usually just provide the Unicode string as
the filename, and it will be automatically converted
to the right encoding for you."

I ended up believing that using an unicode format string
might be the right solution.