Downloading and extracting a whl file without using pip

To download and extract a .whl file without using pip, you can download it with curl or wget and extract it with unzip or Python’s built-in zipfile library.

Since .whl files are essentially ZIP archives, pip is not required.

1. Download the .whl file

If obtaining it from PyPI, first access:

https://pypi.org/project/requests/#files

Obtain the URL of the desired .whl file and download it.

curl -L -o package.whl "https://files.pythonhosted.org/.../package.whl"

Or:

wget -O package.whl "https://files.pythonhosted.org/.../package.whl"

2. Extract the .whl file

Method A: Using unzip

mkdir extracted
unzip package.whl -d extracted/

Method B: Using only the Python standard library

import zipfile
with zipfile.ZipFile("package.whl") as z:
    z.extractall("extracted")

This will extract the contents of the .whl file.

3. Automate the download and extraction process using the PyPI API

This can be achieved using only the Python standard library, without using pip or external commands.

import json
import urllib.request
import zipfile
from pathlib import Path

package = "requests"
version = "2.32.3"

# Obtain package information from the PyPI API
url = f"https://pypi.org/pypi/{package}/{version}/json"
with urllib.request.urlopen(url) as response:
    data = json.load(response)

# Obtain the wheel file
wheel = next(
    f for f in data["urls"]
    if f["filename"].endswith(".whl")
)
wheel_path = Path(wheel["filename"])

# Download
urllib.request.urlretrieve(wheel["url"], wheel_path)

# Extract
output_dir = Path("extracted")
output_dir.mkdir(exist_ok=True)
with zipfile.ZipFile(wheel_path) as z:
    z.extractall(output_dir)

print(f"Downloaded: {wheel_path}")
print(f"Extracted: {output_dir}")

4. Directory structure after extraction

For example, for requests:

extracted/
├── requests/
│   ├── __init__.py
│   ├── api.py
│   ├── sessions.py
│   └── ...
└── requests-2.32.3.dist-info/
    ├── METADATA
    ├── WHEEL
    └── RECORD

If you want to import the extracted files, you can do so as follows:

import sys
sys.path.insert(0, "extracted")
import requests

However, you will also need to install the libraries that requests depends on separately. Simply extracting the files will not automatically resolve dependencies.

Also, for wheel files that include C extensions, you need to choose a wheel that is compatible with your Python version, OS, and CPU architecture.