File Handling

Always use with

open() gives you a file object. with closes it for you, even if something goes wrong in the middle:

with open("notes.txt", "w") as f:
    f.write("first line\n")

with open("notes.txt") as f:
    print(f.read())

first line

Without with you have to remember f.close(), and an exception before it leaves the file open. There is no reason to write that version.

The modes

with open("log.txt", "w") as f:
    f.write("one\n")

with open("log.txt", "a") as f:
    f.write("two\n")

with open("log.txt") as f:
    print(f.read())

one
two

"x" is the safe way to say "only if it is not already there":

with open("once.txt", "x") as f:
    f.write("created")

try:
    with open("once.txt", "x") as f:
        f.write("again")
except FileExistsError:
    print("already exists, refused to overwrite")

already exists, refused to overwrite

Text and binary

By default a file is text: you read and write str, and Python decodes the bytes for you. Add "b" for raw bytes:

with open("data.bin", "wb") as f:
    f.write(b"\x00\x01\x02")

with open("data.bin", "rb") as f:
    print(f.read())

b'\x00\x01\x02'

Use binary mode for images, archives and anything that is not text.

Always name the encoding

with open("city.txt", "w", encoding="utf-8") as f:
    f.write("München")

with open("city.txt", encoding="utf-8") as f:
    print(f.read())

München

Without encoding=, Python uses the platform default, which is UTF-8 on Linux and macOS and historically something else on Windows. The same code then produces different results on different machines. Passing encoding="utf-8" costs nothing and removes a whole category of bug.

Missing files

try:
    with open("not-there.txt") as f:
        print(f.read())
except FileNotFoundError:
    print("no such file")

no such file

Catching the error beats checking first: between a check and the open, the file could disappear.

pathlib

For paths, pathlib is clearer than string joining and works on every platform:

from pathlib import Path

p = Path("notes.txt")
p.write_text("written with pathlib\n", encoding="utf-8")

print(p.read_text(encoding="utf-8").strip())
print(p.name, p.suffix, p.stem)
print(p.exists())

written with pathlib
notes.txt .txt notes
True

Path("folder") / "file.txt" builds a path with the right separator, so you never have to think about backslashes.

from pathlib import Path

path = Path("data") / "reports" / "march.csv"
print(path.parts)
print(path.name, path.parent.name)

('data', 'reports', 'march.csv')
march.csv reports

Printing the path itself would show forward slashes here and backslashes on Windows, which is the whole point: you write it once and pathlib renders it for whichever machine it runs on.