Specifying formatting for csv.writer in Python

I am using csv.DictWriter to output csv files from a set of dictionaries. I use the following function:

def dictlist2file(dictrows, 开发者_运维问答filename, fieldnames, delimiter='\t',
          lineterminator='\n'):
    out_f = open(filename, 'w')

    # Write out header
    header = delimiter.join(fieldnames) + lineterminator
    out_f.write(header)

    # Write out dictionary
    data = csv.DictWriter(out_f, fieldnames,
              delimiter=delimiter,
              lineterminator=lineterminator)
    data.writerows(dictrows)
    out_f.close()

where dictrows is a list of dictionaries, and fieldnames provides the headers that should be serialized to file.

Some of the values in my dictionary list (dictrows) are numeric -- e.g. floats, and I'd like to specify the formatting of these. For example, I might want floats to be serialized with "%.2f" rather than full precision. Ideally, I'd like to specify some kind of mapping that says how to format each type, e.g.

{float: "%.2f"}

that says that if you see a float, format it with %.2f. Is there an easy way to do this? I don't want to subclass DictWriter or anything complicated like that -- this seems like very generic functionality.

How can this be done?

The only other solution I can think of is: instead of messing with the formatting of DictWriter, just use the decimal package to specify the decimal precision of floats to be %.2 which will cause to be serialized as such. Don't know if this is a better solution?

thanks very much for your help.

---

class TypedWriter:
    """
    A CSV writer which will write rows to CSV file "f",
    which uses "fieldformats" to format fields.
    """

    def __init__(self, f, fieldnames, fieldformats, **kwds):
        self.writer = csv.DictWriter(f, fieldnames, **kwds)
        self.formats = fieldformats

    def writerow(self, row):
        self.writer.writerow(dict((k, self.formats[k] % v) 
                                  for k, v in row.iteritems()))

    def writerows(self, rows):
        for row in rows:
            self.writerow(row)

Not tested.