Computer Science · Class 12 · Text Files

Text filesWriting⏱️ 14 min read

write() and writelines()

Two methods put data into a text file, and both of them are far less helpful than print() has taught you to expect. They add no newline, no spaces and no commas — and they refuse a number outright.

1write() puts a string in the file

f.write(s) writes the string s into the file, starting wherever the pointer is. The file must be open in a mode that allows writing: 'w', 'a', or one of the + modes.

first_write.py

f = open('marks.txt', 'w')
f.write('Ravi')
f.write('Meera')
f.write('Amit')
f.close()

f = open('marks.txt', 'r')
print(f.read())
f.close()

Output

RaviMeeraAmit

Three calls, one line

print() ends every call with a newline, which is why two print()s give you two lines. write() ends with nothing. It writes precisely the characters you gave it, and the next write() carries straight on from there.

So the newline is your job. Put \n at the end of every line you write:

with_newlines.py

f = open('marks.txt', 'w')
n = f.write('Ravi\n')
print(n)
f.write('Meera\n')
f.close()

f = open('marks.txt', 'r')
print(f.read())
f.close()

Output

5
Ravi
Meera

What write() hands back

The 5 is the answer from write(): the number of characters it wrote. 'Ravi\n' is five characters — four letters and the newline, which is one character, not two. You can ignore this answer, and almost every program does.

2Strings only — numbers are refused

A text file holds characters, so write() takes a string and nothing else. Hand it a number and the program stops:

not_a_string.py

f = open('marks.txt', 'w')
f.write(95)
f.close()

Output

Traceback (most recent call last):
  File "not_a_string.py", line 2, in <module>
    f.write(95)
TypeError: write() argument must be str, not int

The fix is str(), or an f-string, both of which you have been using since Class 11:

numbers_in.py

marks = 95

f = open('marks.txt', 'w')
f.write(str(marks) + '\n')
f.write(f'Total: {marks}\n')
f.close()

f = open('marks.txt', 'r')
print(f.read())
f.close()

Output

95
Total: 95

3writelines() writes a whole list

f.writelines(lines) takes a list of strings and writes every one of them. The name is misleading: it does not put each item on its own line. It is exactly write() called once per item.

writelines_ok.py

names = ['Ravi\n', 'Meera\n', 'Amit\n']

f = open('names.txt', 'w')
f.writelines(names)
f.close()

f = open('names.txt', 'r')
print(f.read())
f.close()

Output

Ravi
Meera
Amit

writelines_glued.py

plain = ['Ravi', 'Meera', 'Amit']

f = open('glued.txt', 'w')
f.writelines(plain)
f.close()

f = open('glued.txt', 'r')
print(f.read())
f.close()

Output

RaviMeeraAmit

The newline has to be inside the strings

Same list of names, one difference: the first one carried \n on the end of each item. writelines() adds no separator of any kind — not a newline, not a space, not a comma. If you want lines, the list must already contain them.

And like write(), every item must be a string. f.writelines([1, 2, 3]) raises TypeError: write() argument must be str, not int — the same message, because that is what it is doing underneath.

4write() or writelines()?

Let's Recap!

	write()	writelines()
Takes	one string	a list of strings
Adds a newline?	no	no
Gives back	the number of characters written	None
Non-string items	TypeError	TypeError
Use it when	you build the line yourself	the lines are already in a list

These two programs write the same file. Neither is better; use whichever matches the data you are holding.

either_way.py

names = ['Ravi', 'Meera', 'Amit']

# one write() per name
with open('names.txt', 'w') as f:
    for name in names:
        f.write(name + '\n')

# or build the list first and write it in one call
lines = []
for name in names:
    lines.append(name + '\n')

with open('names.txt', 'w') as f:
    f.writelines(lines)

print(open('names.txt').read())

Output

Ravi
Meera
Amit

5Writing records: a name and a mark on each line

A text file has no columns and no fields — only lines. So a record is a line you build yourself, with a separator you choose. A comma is the usual one, and that is all a CSV file is.

records.py

names = ['Ravi', 'Meera', 'Amit']
marks = [78, 91, 65]

with open('marks.txt', 'w') as f:
    for i in range(len(names)):
        f.write(names[i] + ',' + str(marks[i]) + '\n')

print(open('marks.txt').read())

names = ['Ravi', 'Meera', 'Amit']
marks = [78, 91, 65]

with open('marks.txt', 'w') as f:
    for i in range(len(names)):
        f.write(names[i] + ',' + str(marks[i]) + '\n')

print(open('marks.txt').read())

Run it and the file holds three records:

marks.txt

Ravi,78
Meera,91
Amit,65

Choose a separator that is not in the data

A comma is fine for names and marks. It is a bad choice for addresses, which are full of commas — the line would split into pieces that no longer mean anything. Programs that store real text use a character the data cannot contain, such as | or a tab.

6Adding, rather than replacing

Everything above opened the file in 'w', so every run started with an empty file. Change one letter and the same program adds to what is there:

adding.py

📄 marks.txt

📁 marks.txt (virtual file)

Ravi,78
Meera,91
Amit,65

Accessible via standard python read calls.

# run this twice and look at the file both times
with open('marks.txt', 'a') as f:
    f.write('Neha,88\n')

print(open('marks.txt').read())

# run this twice and look at the file both times
with open('marks.txt', 'a') as f:
    f.write('Neha,88\n')

print(open('marks.txt').read())

Watch Out

Change the 'a' to 'w' above and run it again. Ravi, Meera and Amit are gone — one letter, three records.

7Recap

f.write(string)

Writes exactly that string at the pointer, and gives back how many characters it wrote.

No newline is added

Unlike print(). Three writes make one long line unless you put \n in yourself.

Strings only

A number raises TypeError: write() argument must be str, not int. Wrap it in str() or use an f-string.

f.writelines(list)

Writes every string in the list, one after another, adding nothing between them. Returns None.

A record is a line you build

Join the fields with a separator you choose, and end the line with \n.

The mode decides the damage

'w' empties the file first; 'a' adds to the end. The writing code is identical.

✍️ Now write these yourself

1
Write the names of five subjects into subjects.txt, one per line.
Hint · A list, a for loop, and \n on the end of every write().
2
Do the same with one writelines() call instead of a loop of write().
Hint · Build the list with the newlines already in it.
3
Ask the user for three names with input() and append each one to names.txt.
Hint · Mode 'a', so a second run does not wipe the first.
4
Write a name and a mark on each line, then print how many characters write() reported for the longest one.
Hint · write() gives the count back — put it in a variable.

Quick Check

f.write('Ravi'), f.write('Meera'), f.write('Amit'). What is in the file?

Quick Check

What does f.write('Hello\\n') give back?

Quick Check

marks = 95. Which line writes it into a text file?

Quick Check

f.writelines(['a', 'b', 'c']) — what does the file hold?

← Previous

🧷 Opening a File with with

read(), readline() and readlines() 📖