Python pathlib — Modern File Paths

pathlib provides an object-oriented way to work with filesystem paths. Instead of juggling strings and os.path.join, you use a Path object with clean attributes and methods that work the same on Windows, macOS, and Linux.

quickstart.py

from pathlib import Path
 
p = Path("data") / "reports" / "2025.csv"   # join with the / operator
print(p)            # data/reports/2025.csv  (uses the OS separator)
print(p.name)       # 2025.csv
print(p.suffix)     # .csv
print(p.parent)     # data/reports

Why pathlib over os.path?

os.path (old)	pathlib (modern)
os.path.join(a, b)	Path(a) / b
os.path.basename(p)	p.name
os.path.splitext(p)[1]	p.suffix
os.path.exists(p)	p.exists()
Strings everywhere.	One tidy object.

Building paths

Use the / operator to join parts — it’s readable and cross-platform.

building.py

from pathlib import Path
 
base = Path("/home/user")
config = base / "app" / "config.json"
print(config)                 # /home/user/app/config.json
 
# Current working directory and home directory
print(Path.cwd())             # where the script runs
print(Path.home())            # the user's home folder

Inspecting a path

Attribute / method	Returns
p.name	Final component (2025.csv).
p.stem	Name without suffix (2025).
p.suffix	The extension (.csv).
p.parent	The containing directory.
p.parts	A tuple of all components.
p.with_suffix(".txt")	A new path with a changed extension.
p.with_name("new.csv")	A new path with a changed final name.

inspecting.py

from pathlib import Path
 
p = Path("reports/q1/sales.csv")
print(p.name)               # sales.csv
print(p.stem)               # sales
print(p.suffix)             # .csv
print(p.parent)             # reports/q1
print(p.parts)              # ('reports', 'q1', 'sales.csv')
print(p.with_suffix(".json"))  # reports/q1/sales.json

sketch pathlib knows about extensions; os.path.splitext does not p5.js Pause Restart

A path is not a string, and the difference shows up on the parts people actually want. For file.tar.gz, pathlib gives the suffix, the stem and the full list of suffixes, while os.path.splitext strips exactly one extension and calls the rest the name. pathlib also costs more -- about three times os.path.join per operation -- because it builds an object rather than concatenating text. At two microseconds per call that only matters in a loop over very many paths.

Checking existence and type

checks.py

from pathlib import Path
 
p = Path("notes.txt")
print(p.exists())     # True if the path exists
print(p.is_file())    # True if it's a file
print(p.is_dir())     # True if it's a directory

Reading and writing files

pathlib can read and write whole files in one line — no open() boilerplate for simple cases.

read_write.py

from pathlib import Path
 
p = Path("greeting.txt")
p.write_text("Hello, file!", encoding="utf-8")   # write (and create)
print(p.read_text(encoding="utf-8"))              # Hello, file!
 
# Binary variants
p.write_bytes(b"\x00\x01")
print(p.read_bytes())                             # b'\x00\x01'

Listing and globbing directories

listing.py

from pathlib import Path
 
folder = Path(".")
 
# Everything directly inside
for item in folder.iterdir():
    print(item)
 
# Only .py files in this folder
for py in folder.glob("*.py"):
    print(py)
 
# Recursively, every .py file under here
for py in folder.rglob("*.py"):
    print(py)

Method	Finds
p.iterdir()	Every item directly inside p.
p.glob(pattern)	Items matching a pattern in p.
p.rglob(pattern)	Matches recursively in all subfolders.

Creating and removing

create_remove.py

from pathlib import Path
 
d = Path("output/logs")
d.mkdir(parents=True, exist_ok=True)   # create folders; don't error if present
 
f = d / "run.log"
f.touch()                              # create an empty file
f.unlink(missing_ok=True)             # delete a file (no error if missing)

Common pitfalls

• Path("a") / "b" only works with a Path on the left — "a" / Path("b") fails; start from a Path.
• p.suffix includes the dot — .csv, not csv.
• mkdir errors if the parent is missing — pass parents=True.
• read_text raises if the file doesn’t exist — check p.exists() first or catch the error.
• Convert to str(p) only when a library insists on a string.

Practice Exercises

Exercise 1 – Extract the file extension

pch.exerciseTag pch.exerciseTitle

Exercise 2 – Join paths with the / operator

pch.exerciseTag pch.exerciseTitle

Exercise 3 – Name without the extension

pch.exerciseTag pch.exerciseTitle

Summary

• pathlib.Path is the modern, object-oriented replacement for os.path.
• Join paths with /; inspect them with name, stem, suffix, parent, parts.
• Read/write files directly with read_text/write_text (and _bytes variants).
• List and search folders with iterdir, glob, and rglob.
• Create and remove with mkdir(parents=True, exist_ok=True), touch, and unlink.