HarvardMIT/download_script/download.py

from pathlib import Path
import requests
from bs4 import BeautifulSoup
from requests.adapters import HTTPAdapter
from urllib3.util.retry import Retry


def build_session(
    max_retries: int = 3,
    backoff_factor: int = 2,
    session: requests.Session = None
) -> requests.Session:
    """
    Build a requests session with retries

    Args:
        max_retries (int, optional): Number of retries. Defaults to 3.
        backoff_factor (int, optional): Backoff factor. Defaults to 2.
        session (requests.Session, optional): Session object. Defaults to None.
    """
    session = session or requests.Session()
    adapter = HTTPAdapter(max_retries=Retry(total=max_retries, backoff_factor=backoff_factor))
    session.mount("http://", adapter)
    session.mount("https://", adapter)
    session.headers.update({
        "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.3"
    })

    return session


def main():
    output_dir = Path(__file__).parent.parent / "raw"
    output_dir.mkdir(parents=True, exist_ok=True)

    base = 'https://www.hmmt.org'
    req_session = build_session()

    response = req_session.get(f'{base}/www/archive/problems')
    if response.status_code == 200:
        soup = BeautifulSoup(response.text, 'html.parser')
        tags = soup.find_all('a', href=lambda t: '/archive/' in t if t else False)

        urls = []
        for tag in tags:
            url = tag.get('href')
            try:
                num = int(url.split('/')[-1])
                urls.append(base + url)
            except:
                pass

    for url in urls:
        id = url.split('/')[-1]
        response = req_session.get(url)

        if response.status_code == 200:
            soup = BeautifulSoup(response.text, 'html.parser')
            tags = soup.find_all('a', href=lambda t: '.pdf' in t if t else False)
            links = ['https:' + tag.get('href').lstrip('https:') for tag in tags if 'solution' in tag.text.lower()]

            for link in links:
                year, month, topic, name = link.split('/')[-4:]
                response = req_session.get(link)
                (output_dir / f'en-{id}-{year}-{month}-{topic}-{name}').write_bytes(response.content)
        else:
            print('Failed to reach', url)


if __name__ == '__main__':
    main()