python how to trim trailing spaces in csv DictReader keys

Question

I am using python (2.6) csv DictReader. My input file has a header line where the column names have trailing spaces:

colname1,      colname2     ,col3, etc.
XX, YY, ZZ

The returned dict object has key() = ['colname1', 'colname2 ', 'col3']

Is there an option to trim leading and trailing spaces from the keys?

--edit

The problem arises in processing by key names:

with open(fname) as f:
   r = csv.DictReader(f)
   for row in r:
      print "processing", r["column1"], r["column2"]

The files are database dumps. And the dump program is way too smart - it adjust the output column width depending on data -- which means different sets of selects are going to have different column width and different key lengths. Sometimes I must use r['column2 '] and sometimes pad or reduce spaces. ouch!

Answer 1

Just read the first line manually and pass it along to the DictReader.

with open('file.csv') as fh:
    header = [h.strip() for h in fh.next().split(',')]
    reader = csv.DictReader(fh, fieldnames=header)

Answer 2

You need to register a custom dialect in the csv module

csv.register_dialect('MyDialect', quotechar='"', skipinitialspace=True, quoting=csv.QUOTE_NONE, lineterminator='\n', strict=True)

then use the dialect when creating the DictReader:

my_reader = csv.DictReader(trip_file, dialect='MyDialect')

Here's all the Dialect Options

Answer 3

Python3 version

with open('file.csv') as fh:
    header = [h.strip() for h in fh.readline().split(',')]
    reader = csv.DictReader(fh, fieldnames=header)

Answer 4

Following in the vein of other answers, but why not use the CSV reader for that header row?

header = [h.strip() for h in next(csv.reader(f))]
reader = csv.DictReader(f, fieldnames=header)