The with statement

Files, network connections and locks are resources: things you must give back when you're done. Forgetting to close a file can lose data that hasn't been written yet, and leaves the file locked on some systems.

The problem

f = open("data.txt", "w")
f.write("important")
# if an error happened here, f.close() below would never run
f.close()
print("closed:", f.closed)

You could protect it with try/finally:

f = open("data.txt", "w")
try:
    f.write("important")
finally:
    f.close()
print("closed:", f.closed)

That's correct, but wordy, and easy to forget.

The with statement

with does the same job in one line. The file is closed when the block ends, even if an exception happens inside it:

with open("data.txt", "w") as f:
    f.write("important")
print("closed:", f.closed)

try:
    with open("data.txt", "w") as f:
        f.write("first part")
        raise RuntimeError("something went wrong mid-write")
except RuntimeError as e:
    print("Error:", e)
print("Still closed properly:", f.closed)

Several resources at once

with open("source.txt", "w") as f:
    f.write("line one\nline two\n")

with open("source.txt") as src, open("copy.txt", "w") as dst:
    for line in src:
        dst.write(line.upper())

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

Not just files

Anything that needs setting up and tearing down can work with with. Such objects are called context managers. Examples you'll meet:

• open() for files,
• locks in threaded programs (with lock:),
• database connections and transactions,
• decimal.localcontext() for temporary maths settings.

You can also write your own; that's the "Writing context managers" lesson in the last module.