CODESARRAY

Lesson 11, by Codesarray, , 4 min read

Python read and write files and JSON: open, pathlib, json.dump and json.load

Read and write files in Python with open and a with block, build paths with pathlib, and save a dictionary as JSON with json.dump and get it back with json.load.

Everything your program knows disappears when it ends, unless you save it. By the end of this page you will read and write files safely with open and a with block, build paths that work on every system with pathlib, and save your data as JSON so it outlives your program.

open and with

To work with a file, call open with its name and a mode. "r" reads, "w" writes and replaces the whole file, and "a" adds to the end. Always open inside a with block. When the block ends, Python closes the file for you, even if an error happens in the middle.

with open("notes.txt", "w") as f:
    f.write("hi")

pathlib

Build paths with pathlib, not by gluing strings together. Path("data") makes a path object. The / operator joins pieces and uses the right separator on every system. read_text() and write_text() handle a whole file in one call, and exists() tells you whether it is there.

from pathlib import Path

path = Path("data") / "notes.txt"
path.exists()

JSON

JSON is the standard text format for data. json.dump turns a dictionary or a list into JSON text in a file. json.load reads it back into real Python objects. Strings, numbers, lists and dictionaries all survive the trip, and True becomes true in the file. For spreadsheet style rows, the csv module works the same way.

The whole program

from pathlib import Path
import json

path = Path("scores.json")
scores = {"ada": 91, "linus": 78}

with path.open("w") as f:
    json.dump(scores, f)

with path.open() as f:
    loaded = json.load(f)

loaded["grace"] = 88
print(loaded)
print(path.read_text())
{'ada': 91, 'linus': 78, 'grace': 88}
{"ada": 91, "linus": 78}

Line 1 imports Path, and line 2 imports json. Line 4 makes a path to scores.json. Line 5 is a dictionary of scores. Line 7 opens the file for writing inside a with block, and line 8 dumps the dictionary into it. Line 10 opens it again, this time to read, and line 11 loads it back into loaded. Line 13 adds a new score for grace. Line 14 prints the dictionary, and line 15 prints the raw text of the file. The first output line has grace, because it is in memory. The second has only ada and linus, because we never saved grace.

What you learned

  • open takes a mode: "r" reads, "w" replaces, "a" appends, and a with block closes the file for you.
  • Path joins pieces with / and reads or writes a whole file with read_text and write_text.
  • json.dump saves data to a file, and json.load brings it back.
  • A change in memory is not in the file until you save it.

Common questions

Why use a with block? It closes the file when the block ends, even if an error is raised inside it. Without it, you have to remember to call close() yourself.

What does the "w" mode do to an existing file? It empties the file first, then writes. Use "a" if you want to add to the end instead.

Why use pathlib instead of joining strings? Separators differ between systems, and Path picks the right one. It also gives you exists(), read_text() and write_text() on the same object.

Which Python values can JSON store? Strings, numbers, booleans, None, lists and dictionaries with string keys. Other types, such as sets, need converting first.

Why is grace missing from the file? Adding grace changed the dictionary in memory only. Call json.dump again to save it.

This lesson follows on from Python errors and exceptions. Next in the series: modules and uv.

This article goes with the video Files and JSON. Watch it on YouTube.