Pythonで著作権・MIT・Apache2.0ライセンス情報を自動収集・分析する

# GitHub APIとPypi APIを活用した簡易スクリプト例
import requests
import pandas as pd

# PyPIで人気のパッケージリストを取得（例:上位100件）
def fetch_top_packages(n=100):
    url = f"https://hugovk.github.io/top-pypi-packages/top-pypi-packages-30-days.min.json"
    res = requests.get(url).json()
    return [pkg['project'] for pkg in res['rows'][:n]]

# 各パッケージのPyPIメタデータからライセンスを取得
def fetch_license(pkg_name):
    url = f"https://pypi.org/pypi/{pkg_name}/json"
    res = requests.get(url)
    if res.status_code != 200:
        return None, None
    data = res.json()
    info = data['info']
    return info.get('license', ''), info.get('author', '')

# 100件分のライセンス＆著者情報を自動収集
pkgs = fetch_top_packages(100)
results = []
for pkg in pkgs:
    lic, auth = fetch_license(pkg)
    results.append({'package': pkg, 'license': lic, 'author': auth})

df = pd.DataFrame(results)
df.to_csv("oss_license_list.csv", index=False)
print(df.head())

import matplotlib.pyplot as plt

# MIT, Apache, その他の比率円グラフ
lic_counts = df['license'].value_counts()
lic_counts.plot.pie(autopct='%1.1f%%', startangle=90)
plt.title("Python OSSライセンス分布")
plt.ylabel("")
plt.show()

import requests
import pandas as pd
import time

def get_license_info(pkg_name):
    url = f"https://pypi.org/pypi/{pkg_name}/json"
    try:
        res = requests.get(url)
        data = res.json()
        info = data['info']
        return info.get('license', ''), info.get('author', '')
    except:
        return None, None

# 例: 人気OSS 50件
pkgs = ["numpy", "pandas", "matplotlib", "scikit-learn", "flask", "django", "requests", "scipy", "tensorflow", "torch"] # ...and so on
results = []
for pkg in pkgs:
    lic, author = get_license_info(pkg)
    print(f"{pkg}: {lic} by {author}")
    results.append({"package": pkg, "license": lic, "author": author})
    time.sleep(0.5)  # API負荷回避

df = pd.DataFrame(results)
df.to_csv("major_oss_license.csv", index=False)
print(df)