How to rename key/header in csv DictReader

Question

For instance given mycsv.csv file

h1,h2
a,b
c,d

how to rename h2 onto HTwo with reader

reader = csv.DictReader(open('mycsv.csv'))

(Additionally how to write the csv file back with the updated header.)

Note approaches with awk are also valued.

Answer 1

For a simple CSV file (one that contains no quoted strings), awk is a fine solution, and Brian's answer will work fine. If your CSV file contains quoted strings that themselves contain commas, like this:

h1,h2
"this, is, value, 1",value2

...then awk will fall over.

If you want to do this in Python, there's no reason to use DictReader. Something like this will work:

import csv

with open('mycsv.csv') as infd, open('mycsv.out', 'w') as outfd:
  reader = csv.reader(infd)
  writer = csv.writer(outfd)

  # read the header
  header = next(reader)

  # modify the column title
  header[1] = 'hTwo'

  # write the new header out
  writer.writerow(header)

  # copy all other rows unmodified
  for row in reader:
    writer.writerow(row)

# and then rename mycsv.out to mycsv.csv if you wish.

Answer 2

One way to solve it is to use pop()

reader = csv.DictReader(open('mycsv.csv'))

for d in reader:
   d['HTwo'] = d.pop('h2')

Answer 3

awk oneliner:

awk -F, -v OFS=, 'NR==1 && $2=="h2" {$2="hTwo"};1' file

The above is very specific: it says "if the first line contains 'h2' in the second field, change it to 'hTwo'"

Answer 4

inplace replacement with sed

$ sed -i '1s/h2/HTwo/' mycsv.csv

Answer 5

Using CSV DictReader:

Change headers of CSV file and copy the contents to another file:

import csv
with open('old_file.csv','r') as file:
    reader = csv.DictReader(file)
    # read the orignal headers ["head1", "head2"]
    print("old headers list", reader.fieldnames)

    # set new_headers_list []
    new_headers_list = ["new_head1", "new_head2"]
    reader.fieldnames = new_headers_list

    rows = list(reader)
    with open("new_file", "w") as new_file:
        writer = csv.DictWriter(new_file, fieldnames=new_headers_list)
        # write the new header out
        writer.writeheader()
        # copy all other rows unmodified
        writer.writerows(rows)