File I/O

A file lets a program keep data around after it ends — print() shows something on screen, but it's gone the moment the program stops. Saving to a file means that data is still there the next time it runs.

I/O stands for input/output — the general term for a program reading data in or sending data out, to somewhere other than just the screen. print() and input(), already covered on the Foundations page, are technically I/O too — output to the terminal, input from the keyboard. "File I/O" narrows that down to reading from and writing to files on disk specifically.

flowchart LR
    program@{ shape: procs, label: "your_program.py" }
    file@{ shape: paper-tape, label: "notes.txt" }

    file -->|"read()"| program
    program -->|"write()"| file

Fig. 11a — a program reading from and writing to a file

Opening a file

open() returns a file object to read from or write to. Wrap it in a with block so it's closed automatically once you're done, even if something goes wrong partway through.

with open("notes.txt", "w") as file:
    file.write("ball python, 4.5 ft")

print("saved")

The second argument is the mode — what you intend to do with the file:

Mode	Meaning
"r"	Read (default) — the file must already exist
"w"	Write — creates the file if it doesn't exist, erases its contents if it does
"a"	Append — creates the file if it doesn't exist, adds to the end if it does

Without with, you have to call file.close() yourself

with closes the file automatically once its block ends — even if the code inside raises an error. Without it, you have to call file.close() yourself, and forgetting to means your writes might never actually reach the file (still sitting in a buffer), or the file stays locked for anything else trying to open it.

file = open("notes.txt", "w")
file.write("ball python, 4.5 ft")
file.close()          # easy to forget

Run an opening a file example

All the examples above, combined into one script:

with open("notes.txt", "w") as file:
    file.write("ball python, 4.5 ft")

print("saved")

Reading a file

Say notes.txt already exists — written by an earlier run, or typed by hand in a text editor — and looks like this, one snake per line:

ball python, 4.5 ft
burmese python, 12 ft
boa, 8 ft

.read() returns the whole thing as one string, newlines and all.

with open("notes.txt", "w") as file:
    file.write("ball python, 4.5 ft\nburmese python, 12 ft\nboa, 8 ft\n")

with open("notes.txt", "r") as file:
    text = file.read()

print(text)

.readlines() instead returns a list, one string per line, each still ending in a trailing \n. Looping over the file object directly reads it the same way, one line at a time, without holding the whole list in memory at once.

with open("notes.txt", "r") as file:
    lines = file.readlines()
print(lines)    # ["ball python, 4.5 ft\n", "burmese python, 12 ft\n", "boa, 8 ft\n"]

with open("notes.txt", "r") as file:
    for line in file:
        print(line.strip())    # ball python, 4.5 ft / burmese python, 12 ft / boa, 8 ft

Run a reading a file example

All the examples above, combined into one script:

with open("notes.txt", "w") as file:
    file.write("ball python, 4.5 ft\nburmese python, 12 ft\nboa, 8 ft\n")

with open("notes.txt", "r") as file:
    text = file.read()

print(text)

with open("notes.txt", "r") as file:
    lines = file.readlines()
print(lines)

with open("notes.txt", "r") as file:
    for line in file:
        print(line.strip())

Writing multiple lines

.write() doesn't add a newline for you — add one yourself at the end of each line, usually by looping over a list.

species = ["ball python", "burmese python", "boa"]

with open("notes.txt", "w") as file:
    for s in species:
        file.write(s + "\n")

Run a writing multiple lines example

All the examples above, combined into one script:

species = ["ball python", "burmese python", "boa"]

with open("notes.txt", "w") as file:
    for s in species:
        file.write(s + "\n")

with open("notes.txt", "r") as file:
    print(file.read())

Appending vs. overwriting

"w" erases whatever was already in the file before writing anything new — opening a file you meant to add to with "w" is a common way to accidentally lose data. Use "a" instead to add to the end, keeping the existing contents in place.

flowchart TB

subgraph modeW["Overwrite"]
direction LR
w1@{ shape: paper-tape, label: "ball python, 4.5 ft" } -->|"w": erased, then written| w2@{ shape: paper-tape, label: "blood python, 3 ft" }
end

subgraph modeA["Append"]
direction LR
a1@{ shape: paper-tape, label: "ball python, 4.5 ft" } -->|&quot;a&quot;: kept, plus written| a2@{ shape: paper-tape, label: "ball python, 4.5 ft<br/>blood python, 3 ft" }
end

modeW ~~~ modeA

style modeW fill:none,stroke:none
style modeA fill:none,stroke:none

Fig. 11b — overwrite mode vs. append mode

with open("notes.txt", "a") as file:
    file.write("blood python, 3 ft\n")

with open("notes.txt", "r") as file:
    print(file.read())

Run an appending vs. overwriting example

All the examples above, combined into one script:

with open("notes.txt", "w") as file:
    for s in ["ball python", "burmese python", "boa"]:
        file.write(s + "\n")

with open("notes.txt", "a") as file:
    file.write("blood python, 3 ft\n")

with open("notes.txt", "r") as file:
    print(file.read())

For reading and writing rows of structured, comma-separated data specifically, see the csv page under Libraries — it's built on the same open() and file-mode basics covered here.