Python Openpyxl convert CSV to XLSX & removing “$ ,” from cells containing numbers

Question

I have to read a csv file that's generated by a third party and contains a mixture of strings, integers and prices (sometimes with $ signs) into a XLSX file. This is the sample data that's stored in the csv file, a_test_f.csv, that I've got:

ColA,ColB
1,$11.00
2,22
3,"$1,000.56"
4,44

and here is the code that I've written. My question is, is this the most efficient way of performing this conversion. Is there an alternative method that would use less processing power / memory? This is especially important given that the real csv file will contain thousands of records and hundreds of columns and the conversion operation will have to be performed tens of thousands of times per day.

import csv
import openpyxl

#
# Convert the data in csv file format that contains a mix of
# strings, integers and dollar amounts into xlsx file format
#

csvfile  = 'a_test_f.csv'
xlsxfile = 'new_xlsx_f.xlsx'

wb = openpyxl.Workbook()
ws = wb.active

# remove $ and , from numbers
class Clean:
    def __init__(self, data=''):
        self.__obj = data
    def __repr__(self):
        return f"{self.__obj}"
    def getData(self):
        return self.__obj

    def dollar(self):
        try:
            return Clean(data=self.__obj.replace('$',''))
        except TypeError as err:
            print(err)

    def comma(self):
        try:
            return Clean(data=self.__obj.replace(',',''))
        except TypeError as err:
            print(err)

    def digit(self):
        try:
            float(self.__obj)
            return True
        except ValueError:
            return False            

with open(csvfile) as f:
    reader = csv.reader(f, delimiter=',', quotechar='"')
    row_count=1
    for row in reader:
        for i in range(len(row)):
            if Clean(data=row[i]).dollar().comma().digit():
                content = float(repr(Clean(data=row[i]).dollar().comma()))
            else:
                content = row[i]                
            ws.cell(row=row_count,column=i+1).value = content
        row_count +=1

wb.save(xlsxfile)

print('Finished!')

Answer 1

Following Charlie's suggestion, I rewrote the conversion using Functions instead of a Class and then tried processing a million items in a csv file using the Class and Functions methods. Results:

• The Function and the Class methods used equivalent amount of CPU and memory
• The Class method was 9.4% slower than using Functions

Functions win. Thank you Charlie!

The Function method is below:

import csv
import openpyxl

#
# Convert the data in csv file format that contains a mix of
# strings, integers and dollar amounts into xlsx file format
#

csvfile  = 'large_test_export.csv'
xlsxfile = 'new_xlsx_f.xlsx'

wb = openpyxl.Workbook()
ws = wb.active

# remove $ and , from numbers

def strip_stuff(a_string):
    try:
        temp = a_string.replace(',','')
    except TypeError as err:
        print(err)      
    
    try:
        temp2 = temp.replace('$','')
    except TypeError as err:
        print(err)      

    try:
        temp3 = float(temp2)
        return temp3
    except ValueError as err:
        return temp2


def is_number(b_string):
    temp = strip_stuff(b_string)
    try:
        float (temp)
        return True
    except ValueError:
        return False

with open(csvfile) as f:
    reader = csv.reader(f, delimiter=',', quotechar='"')
    row_count=1
    for row in reader:
        for i in range(len(row)):
            if is_number(row[i]):
                content = strip_stuff(row[i])
            else:
                content = row[i]                
            ws.cell(row=row_count,column=i+1).value = content
        row_count +=1

wb.save(xlsxfile)

print('Finished!')