leetcode 爬取每日一题官方题解

leetcode_daily_solution.py

'''
Just for fun.
参考文章: https://wanakiki.github.io/2020/leetcode-spider/
'''

import requests, json

user_agent = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/81.0.4044.122 Safari/537.36"
url = "https://leetcode-cn.com/graphql"
session = requests.Session()
common_headers = {
    "User-Agent": user_agent,
    "Connection": "keep-alive",
    "Content-Type": "application/json",
}


def 找到每日一题标题() -> str:
    params = {
        "operationName": "questionOfToday",
        "variables": {},
        "query": """query questionOfToday {
          todayRecord {
            question {
              questionFrontendId
          questionTitleSlug
          __typename
        }
        date
      }
    }
    """,
    }
    json_data = json.dumps(params).encode("utf8")
    headers = {
        **common_headers,
        "Referer": "https://leetcode-cn.com/problems/",
    }
    resp = session.post(url, data=json_data, headers=headers, timeout=10)
    resp.encoding = "utf8"
    return resp.json()["data"]["todayRecord"][0]["question"][
        "questionTitleSlug"
    ]


def 找到官方题解文章(slug: str):
    params = {
        "operationName": "questionSolutionArticles",
        "variables": {
            "questionSlug": slug,
            "first": 1,
            "skip": 0,
            "orderBy": "DEFAULT",
        },
        "query": """query questionSolutionArticles($questionSlug: String!, $skip: Int, $first: Int, $orderBy: SolutionArticleOrderBy, $userInput: String, $tagSlugs: [String!]) {
              questionSolutionArticles(questionSlug: $questionSlug, skip: $skip, first: $first, orderBy: $orderBy, userInput: $userInput, tagSlugs: $tagSlugs) {
            edges {
                  node {
                    ...solutionArticle
              }
            }
          }
        }
        
        fragment solutionArticle on SolutionArticleNode {
          title
          slug
          tags {
                name
            nameTranslated
            slug
          }
        }
    """,
    }
    json_data = json.dumps(params).encode("utf8")
    headers = {
        **common_headers,
        "Referer": f"https://leetcode-cn.com/problems/{slug}/solution/",
    }
    resp = session.post(url, data=json_data, headers=headers, timeout=10)
    resp.encoding = "utf8"
    return resp.json()["data"]["questionSolutionArticles"]["edges"][0]["node"][
        "slug"
    ]


def 找到官方题解代码():
    daily_problem = 找到每日一题标题()
    题解文章 = 找到官方题解文章(daily_problem)
    params = {
        "operationName": "solutionDetailArticle",
        "variables": {
            "slug": 题解文章,
            "orderBy": "DEFAULT",
        },
        "query": """query solutionDetailArticle($slug: String!, $orderBy: SolutionArticleOrderBy!) {
              solutionArticle(slug: $slug, orderBy: $orderBy) {
                ...solutionArticle
            content
            question {
                  questionTitleSlug
            }
          }
        }
        
        fragment solutionArticle on SolutionArticleNode {
          title
          slug
          tags {
                name
            slug
          }
          summary
        }
        """,
    }
    json_data = json.dumps(params).encode("utf8")
    headers = {
        **common_headers,
        "Referer": f"https://leetcode-cn.com/problems/{daily_problem}/solution/{题解文章}/",
    }
    resp = session.post(url, data=json_data, headers=headers, timeout=10)
    resp.encoding = "utf8"
    title = resp.json()["data"]["solutionArticle"]["content"]
    print(title)
    return title


找到官方题解代码()