hint: unicode to ascii with minimal loss of information

[Alexander Belchenko]

Recent "Daily Python-URL" has link to small article "Unicode strings to 
ASCII ...nicely" (http://www.peterbe.com/plog/unicode-to-ascii). This 
article show the example how to convert unicode string to almost 
equivalent ascii-string with some loss of data. I think it's may be 
interested for creating file-id for non-ascii filenames.

Unfortunately it's works only for latin-based alphabets (latin-1 
encoding) but anyway it's good approach, IMO.

--
Alexander