writing mixed data from a .csv file to an .xls with xlwt - how to format numbers?

Question

I'm trying to read a bunch of mixed data (strings & numbers) from a csv & write everything to an excel file. unfortunately, everything is written as text since in csv text is the only format. how can i write the numerical data in the correct format? this is the code i have so far...

import xlwt, csv

wb = xlwt.Workbook()
ws = wb.add_sheet('testSheet')

sourceCSV = csv.reader(open('sourceCSVfile.csv', 'rb'), delimiter=",")

for rowi, row in enumerate(sourceCSV):
  for coli, value in enumerate(row):
      ws.write(rowi, coli, value)
wb.save('TEST.xls') 

Answer 1

Somehow, depending on your data, you need to be able to determine what data-type is in each field.

If your data has the same data-type in each column, you can do something like this:

# 5 columns: text, integer, float, float, date in YYYY-MM-DD format
import datetime
def date_conv(s):
    return datetime.datetime.strptime(s, "%Y-%m-%d")
converters = (str.strip, int, float, float, date_conv)
# use int if you want to check that it is an int.
# xlwt will convert the int to float anyway.
...
for rowi, row in enumerate(sourceCSV):
  for coli, value in enumerate(row):
      ws.write(rowi, coli, converters[coli](value))

Other possibilities:

(1) the suck-it-and-see approach:

def float_if_possible(strg):
    try:
        return float(strg)
    except ValueError:
        return strg
...
ws.write(rowi, coli, float_if_possible(value))

(2) the analytical approach:

You need to write carefully nitpicky regexes to analyse your text, and you need to apply them in the appropriate order.

For a float, consider this:

float_const_pattern = r"""
    [+-]? # optional sign
    (?:
        (?: \d* \. \d+ ) # .1 .12 .123 etc 9.1 etc 98.1 etc
        |
        (?: \d+ \. ) # 1. 12. 123. etc
        |
        (?: \d+ ) # 1 12 123 etc
    )
    # followed by optional exponent part
    (?: [Ee] [+-]? \d+ ) ?
    # followed by end of string
    \Z # don't use $
    """

along with the flag re.VERBOSE . Note in particular the "end of string" check. If you don't do that, given input 123qwerty , the regex will match 123 and the float("123qwerty") call will raise an exception.