Computer Science · Class 12 · CSV Files

CSV fileswriting⏱️ 13 min read

writer() and writerow()

Two lines of new code, and one of them only runs once. csv.writer(f) wraps the open file; then writerow() takes a list and puts it in the file as one line — commas and all, which is three jobs you no longer do by hand.

1csv.writer() — wrapping the file

w = csv.writer(file)

One argument — the file open() gave you, in 'w' or 'a'. What comes back is the writer.

The name w is just a variable. Call it writer, csvwriter, or anything else — the board's own papers use several. What matters is that the object it holds is the thing with the methods on it.

Note

Nothing is written to the file by this line. Wrapping is all it does — a file opened in 'w' is empty before it and empty after it.

2writerow() — one list, one line

w.writerow(a list)

Each item of the list becomes one value in the row. Call it again for the next row.

create_marks.py

import csv

f = open('marks.csv', 'w', newline='')
w = csv.writer(f)

w.writerow(['Roll', 'Name', 'Marks'])
w.writerow([1, 'Ravi', 78])
w.writerow([2, 'Meera', 91])
w.writerow([3, 'Amit', 65])

f.close()
print('marks.csv written')

Output

marks.csv written

And here is the file it produced, opened in Notepad:

marks.csv

Roll,Name,Marks
1,Ravi,78
2,Meera,91
3,Amit,65

w.writerow(['Roll', 'Name', 'Marks'])

The header row is nothing special — just the first writerow(). The csv module has no idea it is a header, and neither will the reader.

w.writerow([1, 'Ravi', 78])

1 and 78 are numbers and went straight in. write() would have refused them; writerow() converts them for you.

1,Ravi,78

Two commas for three values, and a line ending you never typed. That is the work the module took over.

3Watch the list become a line

✍️ One list in, one line out

Pick a row and watch what writerow() writes — then what reader() gives back when the file is read again.

the list you pass

['Roll''Name''Marks']

w.writerow(['Roll', 'Name', 'Marks'])

the line in marks.csv

Roll,Name,Marks

highlighted comma = a separator the writer added

read the file again — what the row comes back as

['Roll', 'Name', 'Marks']

Two commas for three values — one between each pair, none at the end.

4The same file, written by hand

Worth comparing directly, because the exam sometimes asks for the difference:

Let's Recap!

	f.write()	w.writerow()
What it takes	one string	a list of values
Numbers	refused — TypeError, you need str()	accepted as they are
Separators	you type every comma yourself	put in for you
Line ending	you add '\n' yourself	added for you
A comma inside a value	breaks the file silently	the value is quoted automatically
Gives back	the number of characters written	the number of characters written

both_ways.py

import csv

# by hand — three things to remember on every row
f = open('a.csv', 'w')
f.write('Ravi' + ',' + str(78) + '\n')
f.close()

# with the module — none of them
f = open('b.csv', 'w', newline='')
w = csv.writer(f)
w.writerow(['Ravi', 78])
f.close()

print(open('a.csv', 'rb').read())
print(open('b.csv', 'rb').read())

Output

b'Ravi,78\n'
b'Ravi,78\r\n'

Same values, same commas — and one difference at the very end. The line you wrote by hand ends in \n, because that is what you typed. The writer's line ends in \r\n, because that is the ending the CSV standard asks for and the writer always uses it.

Remember this one — it comes back

\r\n is two characters, not one. Nothing goes wrong here — every program on earth reads that file correctly. But it is the whole reason newline='' exists, and the lesson three pages from now is about what happens when you leave it out.

5Rows come from a loop, not from ten typed lines

Real programs do not know their rows in advance. The data sits in a list, and writerow() goes inside a for:

write_loop.py

import csv

students = [[1, 'Ravi', 78], [2, 'Meera', 91], [3, 'Amit', 65]]

f = open('marks.csv', 'w', newline='')
w = csv.writer(f)

w.writerow(['Roll', 'Name', 'Marks'])
for s in students:
    w.writerow(s)

f.close()

# read it back to prove it worked
f = open('marks.csv', 'r', newline='')
print(f.read())
f.close()

import csv

students = [[1, 'Ravi', 78], [2, 'Meera', 91], [3, 'Amit', 65]]

f = open('marks.csv', 'w', newline='')
w = csv.writer(f)

w.writerow(['Roll', 'Name', 'Marks'])
for s in students:
    w.writerow(s)

f.close()

# read it back to prove it worked
f = open('marks.csv', 'r', newline='')
print(f.read())
f.close()

The header goes outside the loop

Put w.writerow(['Roll', 'Name', 'Marks']) inside the loop by mistake and you get a header line before every single record. It is written once, so it sits above the loop.

6What writerow() gives back

It returns the number of characters it wrote — the same thing write() returns, and just as ignorable:

returned.py

import csv

f = open('t.csv', 'w', newline='')
w = csv.writer(f)
print(w.writerow(['Ravi', 78]))
f.close()

Output

Nine, because the line is Ravi,78 followed by the two characters of the line ending. Nobody uses this value. The point of the call is the change it makes to the file.

7Four mistakes worth knowing in advance

w.writerow('Ravi')

A string, not a list. Python takes each letter as a value and writes R,a,v,i. No error — just a wrong file.

csv.writerow(['Ravi', 78], f)

writerow belongs to the writer object, not to the module. AttributeError: module 'csv' has no attribute 'writerow'.

Forgetting f.close()

The rows may still be in the buffer. Open the file and it looks empty or half written.

Opening in 'w' to add a record

'w' empties the file first. Every record already in it is gone before writerow() runs once.

string_not_list.py

import csv

f = open('trap.csv', 'w', newline='')
w = csv.writer(f)
w.writerow('Ravi')
f.close()

print(open('trap.csv').read())

Output

R,a,v,i

Key Takeaway

writerow() takes a list of values. One value still needs its square brackets: w.writerow(['Ravi']). Without them Python looks inside the string and finds four characters, which are four values as far as it is concerned.

8Recap

w = csv.writer(f)

Wraps the open file. Runs once, writes nothing by itself.

w.writerow(list)

Writes one row. Each item of the list becomes one value.

Commas and newline are added

The three jobs write() left to you — the separator, the line ending and str() — are all gone.

Numbers need no str()

78 goes in as 78. It comes back as '78', but that is the reader's lesson.

The header is just a row

One writerow() above the loop. The module gives it no special status.

One value still needs brackets

writerow('Ravi') writes R,a,v,i. writerow(['Ravi']) writes Ravi.

✍️ Now write these yourself

1
Write a CSV of five of your subjects and their marks, with a header row.
Hint · One writerow() for the header, then a loop for the rest.
2
Run w.writerow('Ravi') and open the file.
Hint · R,a,v,i — and no error at all, which is the danger.
3
Write a row with a name that contains a comma, and open the file in Notepad.
Hint · Double quotes you never typed. The writer put them there.
4
Write the header inside the loop on purpose and look at the result.
Hint · Four headers for three records.

Quick Check

What does w.writerow(['Ravi', 78]) put in the file?

Quick Check

Do numbers need str() before writerow()?

Quick Check

w.writerow('Meera') writes what into the file?

Quick Check

Where should the header's writerow() go when the records are written in a loop?

← Previous

📂 Opening & Closing a CSV File

writerows() 📚