Str.format() function fails on Norwegian characters

JosteinSvegården

I have a script with a SearchCursor that runs through a feature layer and writes the feature properties to a semicolon divided text file. When it comes to a feature which has a property that contains one of the Norwegian letters æ, ø or å (html: æ, ø or å), the script crashes on the str.format() function

Traceback info:
  File "C:\Infrastructure_calculation.py", line 97, in <module>
    propStr += ";{0}".format(IS[index])

Error Info:
'ascii' codec can't encode character u'\xc6' in position 1: ordinal not in range(128)

ISs = da.SearchCursor("ISpoint_lyr", "*")
for IS in ISs:
    for index in range(len(ISs.fields)):
        if index <> shapeFieldIndex:
             propStr += ";{0}".format(IS[index])

I guess this have something to do with unicode encoding, but I have not been able to find a solution. Any input is appreciated!

JosteinSvegården

Thanks, Jason, that did it!

It is easy to get lost in all this encoding stuff. So for readers that might bump into this thread later, I will add that to be able to write the resulting string to a file, I had to convert the unicode string to a string of bytes using the string.encode() function:

STR_result = STR_result.encode('utf8', 'replace')

See http://pythonhosted.org/kitchen/unicode-frustrations.html