LongArticlesJob/coldStartTasks/crawler/wechat/article_association.py

"""
@author: luojunhui
"""

import time
import traceback
from datetime import datetime

import numpy as np

from pymysql.cursors import DictCursor
from tqdm import tqdm


from applications import WeixinSpider, log
from applications.api import similarity_between_title_list
from applications.const import ColdStartTaskConst
from applications.db import DatabaseConnector
from applications.functions import Functions
from applications.utils import get_inner_account_set
from applications.utils import whether_title_sensitive
from config import long_articles_config

spider = WeixinSpider()
functions = Functions()
const = ColdStartTaskConst()


class ArticleAssociationCrawler(object):
    """
    article association crawler task
    """

    def __init__(self):
        self.db_client = DatabaseConnector(db_config=long_articles_config)
        self.db_client.connect()
        self.inner_account_set = get_inner_account_set()

    def get_seed_url_list(self, biz_date):
        """
        获取种子url列表
        """
        sql = f"""
            select gh_id, title, link
            from datastat_sort_strategy
            where date_str > DATE_FORMAT(DATE_SUB('{biz_date}', INTERVAL 2 DAY), '%Y%m%d') 
                and view_count > {const.READ_COUNT_THRESHOLD} 
                and read_rate > {const.READ_AVG_THRESHOLD} 
                and type = {const.BULK_PUBLISH_TYPE}
            order by read_rate desc 
            limit {const.SEED_ARTICLE_LIMIT_NUM};
        """
        seed_article_list = self.db_client.fetch(query=sql, cursor_type=DictCursor)
        return seed_article_list

    def get_level_up_title_list(self):
        """
        获取晋级文章标题列表
        status: 1 表示文章已经溯源完成
        deleted: 0 表示文章正常
        level = 'autoArticlePoolLevel1' 表示头条
        """
        sql = f"""
            select distinct title 
            from article_pool_promotion_source 
            where level = 'autoArticlePoolLevel1' and status = 1 and deleted = 0;
        """
        mysql_response = self.db_client.fetch(query=sql)
        title_list = [i[0] for i in mysql_response]
        return title_list

    def get_recommend_url_list_with_depth(
        self, seed_url, source_title, source_account, base_title_list, depth=1
    ):
        """
        @param seed_url: good url from data_sort_strategy
        @param depth: association depth
        @param source_title: article title
        @param source_account: article account
        """
        if depth > const.ARTICLE_ASSOCIATION_MAX_DEPTH:
            return

        res = spider.get_recommend_articles(content_link=seed_url)
        related_articles = res["data"]["data"]["list"]
        if related_articles:
            title_list = [i["title"] for i in related_articles]
            similarity_array = similarity_between_title_list(
                title_list, base_title_list
            )

            recommend_articles = []
            for index, score_list in enumerate(similarity_array):
                sorted_score_list = sorted(score_list)
                percent_threshold_score = np.percentile(
                    sorted_score_list, const.PERCENT_THRESHOLD
                )
                if percent_threshold_score < const.CORRELATION_THRESHOLD:
                    continue

                else:
                    article_obj = related_articles[index]
                    article_obj["score"] = percent_threshold_score
                    recommend_articles.append(article_obj)

            recommend_process_bar = tqdm(
                recommend_articles, desc="save recommend articles"
            )
            for article in recommend_process_bar:
                obj = {
                    "title": article["title"],
                    "url": article["url"],
                    "gh_id": article["username"],
                    "index": article["idx"],
                    "send_time": article["send_time"],
                    "read_cnt": article["read_num"],
                    "depth": depth,
                    "source_article_title": source_title,
                    "source_account": source_account,
                }
                self.insert_recommend_article(obj)
                recommend_process_bar.set_postfix(
                    {"title": article["title"], "depth": depth}
                )
                self.get_recommend_url_list_with_depth(
                    seed_url=obj["url"],
                    source_title=obj["title"],
                    source_account=obj["gh_id"],
                    base_title_list=base_title_list,
                    depth=depth + 1,
                )
        else:
            return

    def insert_recommend_article(self, obj):
        """
        insert recommend article
        """
        # whether account inside
        if obj["gh_id"] in self.inner_account_set:
            return

        # whether article title exists
        title = obj["title"]
        select_sql = "select article_id from crawler_meta_article where title = %s;"
        res = self.db_client.fetch(query=select_sql, params=(title,))
        if res:
            return

        # whether title sensitive
        title_sensitivity = (
            const.TITLE_SENSITIVE
            if whether_title_sensitive(title)
            else const.TITLE_NOT_SENSITIVE
        )

        # insert this article
        insert_sql = f"""
            insert into crawler_meta_article 
            (platform, mode, category, out_account_id, article_index, title, link, read_cnt, publish_time, crawler_time, status, unique_index, source_article_title, source_account, title_sensitivity)
            values (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s);
        """
        self.db_client.save(
            query=insert_sql,
            params=(
                "weixin",
                "recommend",
                "article_association",
                obj["gh_id"],
                obj["index"],
                obj["title"],
                obj["url"],
                obj["read_cnt"],
                obj["send_time"],
                int(time.time()),
                const.DEFAULT_ARTICLE_STATUS,
                functions.generateGzhId(obj["url"]),
                obj["source_article_title"],
                obj["source_account"],
                title_sensitivity,
            ),
        )

    def deal(self, biz_date=None):
        """
        class entrance
        :param biz_date:
        """
        if biz_date is None:
            biz_date = datetime.today().strftime("%Y-%m-%d")

        seed_article_list = self.get_seed_url_list(biz_date)
        deal_bar = tqdm(seed_article_list, desc="article association crawler")
        base_title_list = self.get_level_up_title_list()
        for article in deal_bar:
            try:
                self.get_recommend_url_list_with_depth(
                    seed_url=article["link"],
                    source_title=article["title"],
                    source_account=article["gh_id"],
                    base_title_list=base_title_list,
                )
                deal_bar.set_postfix({"article_title": article["title"]})
            except Exception as e:
                log(
                    task="article_association_crawler",
                    function="deal",
                    message=f"article association crawler error, article title: {article['title']}, error: {e}",
                    data={"article": article, "traceback": traceback.format_exc()},
                )