CSV Files

Do not split on commas

A real CSV can hold commas inside quoted fields, and splitting by hand cuts them in half:

Python

line = 'Ada,"maths, and computing",1815'

print(line.split(","))

Output

['Ada', '"maths', ' and computing"', '1815']

Three fields became four. The csv module knows the rules:

Python

import csv
import io

line = 'Ada,"maths, and computing",1815'
print(next(csv.reader(io.StringIO(line))))

Output

['Ada', 'maths, and computing', '1815']

Writing

Python

import csv

rows = [["name", "born"], ["Ada", 1815], ["Grace", 1906]]

with open("people.csv", "w", encoding="utf-8", newline="") as f:
    writer = csv.writer(f)
    writer.writerows(rows)

with open("people.csv", encoding="utf-8") as f:
    print(f.read().strip())

Output

name,born
Ada,1815
Grace,1906

Reading rows

Python

import csv

with open("people.csv", "w", encoding="utf-8", newline="") as f:
    csv.writer(f).writerows([["name", "born"], ["Ada", 1815]])

with open("people.csv", encoding="utf-8", newline="") as f:
    for row in csv.reader(f):
        print(row)

Output

['name', 'born']
['Ada', '1815']

Note '1815' is a string. CSV has no types — everything comes back as text, and converting is your job.

DictReader is usually what you want

It uses the header row, so you read by column name rather than by position:

Python

import csv

with open("people.csv", "w", encoding="utf-8", newline="") as f:
    csv.writer(f).writerows([["name", "born"], ["Ada", 1815], ["Grace", 1906]])

with open("people.csv", encoding="utf-8", newline="") as f:
    for row in csv.DictReader(f):
        print(row["name"], int(row["born"]) + 100)

Output

Ada 1915
Grace 2006

Now inserting a column in the file cannot break your code.

DictWriter

Python

import csv

people = [
    {"name": "Ada", "born": 1815},
    {"name": "Grace", "born": 1906},
]

with open("people.csv", "w", encoding="utf-8", newline="") as f:
    writer = csv.DictWriter(f, fieldnames=["name", "born"])
    writer.writeheader()
    writer.writerows(people)

with open("people.csv", encoding="utf-8") as f:
    print(f.read().strip())

Output

name,born
Ada,1815
Grace,1906

fieldnames fixes the column order and, by default, a key not in that list raises rather than being silently dropped.

Other separators

Python

import csv

with open("data.tsv", "w", encoding="utf-8", newline="") as f:
    csv.writer(f, delimiter="\t").writerow(["a", "b"])

with open("data.tsv", encoding="utf-8", newline="") as f:
    print(next(csv.reader(f, delimiter="\t")))

Output

['a', 'b']

Semicolons are common in countries where the comma is the decimal separator, so delimiter=";" comes up often with spreadsheet exports.

Large files

A reader is an iterator, so it never loads the whole file:

Python

import csv

with open("big.csv", "w", encoding="utf-8", newline="") as f:
    writer = csv.writer(f)
    writer.writerow(["n"])
    writer.writerows([[i] for i in range(1000)])

total = 0
with open("big.csv", encoding="utf-8", newline="") as f:
    for row in csv.DictReader(f):
        total += int(row["n"])

print(total)

Output

That works the same on a file of ten million rows.

Do not split on commas

Writing

Reading rows

DictReader is usually what you want

DictWriter

Other separators

Large files

Test yourself

Why not split a CSV line on commas yourself?

Why pass newline="" when opening a CSV?

Context Managers