Bio and Publications

Thursday, January 19, 2012

Python 2.7 CSV files with Unicode Characters

The csv module in Python 2.7 is more-or-less hard-wired to work with ASCII and only ASCII.

Sadly, we're often confronted with CSV files that include Unicode characters.  There are numerous Stack Overflow questions on this topic.  http://stackoverflow.com/search?q=python+csv+unicode

What to do?  Since csv is married to seeing ASCII/bytes, we must explicitly decode the column values.

One solution is to wrap csv.DictReader, something like the following.  We need to decode each individual column before attempting to do anything with value.

class UnicodeDictReader( object ):
    def __init__( self, *args, **kw ):
        self.encoding= kw.pop('encoding', 'mac_roman')
        self.reader= csv.DictReader( *args, **kw )
    def __iter__( self ):
        decode= codecs.getdecoder( self.encoding )
        for row in self.reader:
            t= dict( (k,decode(row[k])[0]) for k in row )
            yield t

This new object is an iterable which contains a DictReader. We could subclass DictReader, also.

The use case, then, becomes something simple like this.
with open("some.csv","rU") as source:
    rdr= UnicodeDictReader( source )
    for row in rdr:
        # process the row

We can now get Unicode characters from a CSV file.

2 comments:

  1. I like that you think 'mac_roman' is a sensible default encoding.

    ReplyDelete
  2. It's one of those things... The files came to me in that encoding. UTF-8 seems so much more sensible.

    ReplyDelete

Note: Only a member of this blog may post a comment.