naver cafe search

search.py

import sys

import requests

from pyquery import PyQuery


def get_content(link):
    resp = requests.get(
        link,
        headers={
            'Referer': ('https://m.cafe.naver.com/' +
                        'SectionArticleSearch.nhn?query')
        }
    )

    html = PyQuery(resp.content)
    post = html('#postContent')[0]

    return '\n'.join(post.itertext()).strip().replace('\xa0', '')


def search(term):
    resp = requests.get(
        'https://m.cafe.naver.com/SectionArticleSearch.nhn',
        params={
            'query': f'"{term}"',
        }
    )

    html = PyQuery(resp.content)

    links = [a.get('href') for a in html('ul#searchList a')]

    return {
        link: get_content(link)
        for link in links
    }


if __name__ == '__main__':
    for url, content in search(sys.argv[1]).items():
        print(url)
        print(content)