gh-tokens-loader

GitHub Tokens Loader

A simple utility library to load GitHub tokens from a structured file which helps track who the token is linked to and when it expires.

Usage

GitHub Tokens File

To use this library, create a file with the following structure:

tokens:
  jane:
    token: "github_pat_56..."
    expiration_date: null
  john:
    token: "github_pat_41..."
    expiration_date: "2026-02-13"
  bob:
    token: "ghp_VeaD3..."
    expiration_date: null
  sally:
    token: "ghp_VVv5E..."
    expiration_date: "2023-02-13"

Each token has a short name (generally who or what it is linked to), the token itself, and an optional expiration date. The expiration date should be in the format YYYY-MM-DD.

⚠️⚠️ Important: Do not share this file with anyone else. Keep it secure and private. It is recommended to add the filename to your .gitignore as well. ⚠️⚠️

Loading Tokens

To load the tokens from the file, use the following code:

from gh_tokens_loader import load_github_tokens

tokens = load_github_tokens("path/to/tokens.yaml")

This will return only the tokens that are valid (i.e., not expired).

Why

I am a researcher who spends a lot of time mining the GitHub API for data about scientific software. Due to the rate limits on the GitHub API, my research collaborators and I will commonly pool our tokens. I found myself copy pasting some form of this code around to different projects and decided to make it a library.

In my own usage, I generally use the GitHubTokensCycler that is also builtin to the library:

import time
from concurrent.futures import ThreadPoolExecutor

from gh_tokens_loader import GitHubTokensCycler

# Load tokens
gh_tokens_cycler = GitHubTokensCycler("path/to/tokens.yaml")

# Imagine some function that uses the GitHub API
def get_repo_data(repo: str, gh_token: str) -> dict:
    # Important to sleep to avoid rate limits
    time.sleep(1)

    # Use the token to get data from the GitHub API
    # ...
    # Return the data
    return {"some": "data"}

# Imagine some list of repos
repos = ["repo1", "repo2", "repo3", "..."]

# Thread with cycling tokens
with ThreadPoolExecutor(max_workers=len(gh_tokens_cycler)) as exe:
    results = list(exe.map(
        get_repo_data,
        repos,
        [next(gh_tokens_cycler) for _ in range(len(repos))],
    ))

# Do something with the results
# ...

The GitHubTokensCycler will occasionally also refresh it's internal set of tokens to only include valid tokens.

License

This project is licensed under the Mozilla Public License 2.0 - see the LICENSE file for details.