Question

I am using the csv python package to read a csv file like this:

r = csv.DictReader(open(r'd:\aaa.csv', 'rb'))
for row in r:
  print row

The problem is that all the values in the resulting dictionary are strings and I need them to be converted to the respective types. I have a mapping between column names and the types.

Are there any shortcuts here or do I have to do it all myself?

EDIT

I have marked https://stackoverflow.com/a/9720733/80002 as the answer. Only, I have slightly modified it. There is no point to use DictReader if I am going to postprocess it anyway, so here is my modification of the Sven’s code:

r = csv.reader(open(r'd:\aaa.csv', 'rb'))
header = r.next()
converters = [converters_map[c] for c in header]
for row in r:
  row = {title:converter(value) for title, converter, value in zip(header, converters, row)}
  print row

Answer 1

If you have a dictionary of converter functions converting strings to the respective types, then all you need is

with open("d:/aaa.csv", "rb") as input_file:
    reader = csv.DictReader(input_file)
    for row in reader:
        row = {k: converters[k](v) for k, v in row.items()}
        print row

(Python 2.7. For earlier versions use dict() instead of the dictionary comprehension.)

You can wrap this code in a generator function (simply use yield row instead of print row).