Resources

Set up

from pathlib import Path

Basics

Working directory

my_dir = Path('.')

Path to current script

my_dir = Path(__file__)

Parent path

Directory of current script:

my_dir = Path(__file__).parent

Navigate up the path tree for any number of levels.

my_dir = Path(file_path).resolve().parent.parent.parent
# Or simply:
my_dir = Path(file_path).resolve().parents[3]

Build a path

my_path = my_dir / 'foo' / 'bar.txt'

List subdirectories

p = Path('.')

dir_paths = [x for x in p.iterdir() if x.is_dir()]

Glob

p.glob(PATH)

# e.g.

list(p.glob('*/*.py'))
list(p.glob('*/abc/*.py'))

Use recursive globe with .rglob.

list(p.rglob('*.py'))
# equivalent to:
list(p.glob('**/*.py'))

Opening a file

with q.open() as f_in:
    text = f_in.readline()

Path relative to current script

Get the path to the current script, the directory of the module, then add the path to that base.

CSV_PATH = Path(__file__).parent / "var" / "my-data.csv"

Home

h = Path.home()
# PosixPath('/Users/mcurrin')

Path.home() / 'abc'
# PosixPath('/Users/mcurrin/abc')

Expand user directory.

p = Path("~/abc")
p
# PosixPath('~/abc')
p.expanduser()
# PosixPath('/Users/my-user/abc')

Path properties

p = Path("abc/def.txt")

p
# PosixPath('abc/def.txt')

p.name
# 'def'

p.parts
('abc', 'def')

p.stem
# 'def'

p.suffix
# '.txt'

Check path properties

p.exists()
p.is_file()
p.is_dir()
p.is_symlink()

Resolve

Resolve to an absolute path - it will resolve symlinks and eliminate “..” components.

p.resolve()

This adds the current working directory to the path object. e.g.

p = Path("abc/def")

p.resolve()
# PosixPath('/Users/.../repos/testing/abc/def')

Path joining

Combine Path objects or Path and str.

Path("my_directory") / "another_dir" / "file.txt"

p = Path('/etc')

q = p / 'init.d' / 'reboot'
# => PosixPath('/etc/init.d/reboot')

q.resolve()
# => PosixPath('/etc/rc.d/init.d/halt')

Relative to

Computes the relative path from one path to another (i.e., “subtracts” a base path from a full path).

Useful for printing, to show only the end part for a few levels, without the base.

path.relative_to(base_path)

Printing a path object will just be the file or foldername. And using resolve above will make it very long. So you can use this to specify a path that the object should be relative to.

Directories

Create with mkdir

e.g.

folder_path.mkdir(parents=True, exist_ok=True)

Practical examples

From my projects.

Absolute path to app directory

It is dangerous to just use Path('.') as that is the current working directory, not the directory of the script.

So use Path(__file__) rather.

Then .resolve() to make it absolute. This is useful when running the script from anywhere in or outside the repo and keeping paths relative to that app directory i.e. the top-level module inside the repo.

my_repo/
    my_app/          # App directory.
        my_app.py    # Current script
        constants.py

After research, I found this to be effective. It resolves relative and symlinks (I think?).

APP_DIR = Path(__file__).resolve().parent

This should be set in the config or constants module, so it can be imported by other scripts.

from constants import APP_DIR

Path to var directory

For writing CSV or log files, for example.

VAR_DIR = APP_DIR / "var"
APP_LOG_PATH = VAR_DIR / "app.log"