# -*- coding: utf-8 -*-
from datetime import datetime
from pymongo import MongoClient
import psycopg2


def mongo():
    # mongodb
    client = MongoClient('mongodb://root:aaaAAA111!!!@192.168.31.178:38000/')

    # 指定数据库名称
    db_name = 'NEWS'  # 替换为你的数据库名称

    # 选择数据库
    db = client[db_name]

    # 列出数据库中的所有集合
    collections = db.list_collection_names()

    all_data = []

    for collection_name in collections:
        # 选择集合
        collection = db[collection_name]

        # 读取集合中的所有数据
        for document in collection.find({}, {'_id': 0}):
            all_data.append(document)

    sorted_data = []

    if all_data:
        sorted_data = sorted(all_data, key=lambda x: x['create_time'], reverse=True)

    return sorted_data


def pg(sorted_data):
    # 数据库连接参数
    db_params = {
        'dbname': 'auto',
        'user': 'odoo',
        'password': 'odoo',
        'host': '192.168.31.178',
        'port': '5432'
    }
    conn = psycopg2.connect(**db_params)
    for doc in sorted_data:
        # 如果有title，就不插入
        try:
            cur = conn.cursor()

            create_time_dt = None
            if doc.get('create_time'):
                create_time_dt = datetime.utcfromtimestamp(doc['create_time'])

            values = {
                'name': doc.get('title'),
                'context': doc.get('context') or '',
                'source_url': doc.get('source_url') or '',
                'link': doc.get('line') or '',
                'article_type': doc.get('article_type') or '',
                'article_source': doc.get('article_source') or '',
                'img_url': doc.get('img_url') or '',
                'keyword': doc.get('keyword') or '',
                'posted_date': doc.get('posted_date') or '',
                'create_time_ts': doc.get('create_time') or '',
                'create_time': create_time_dt,
                'create_datetime': datetime.strptime(doc['create_datetime'], '%Y-%m-%d %H:%M:%S') if doc.get('create_datetime') else None
            }

            # 将create_time转换为适合数据库的时间戳格式
            create_time_dt = datetime.utcfromtimestamp(values['create_time_ts']) if values['create_time_ts'] else None
            values['create_time'] = create_time_dt

            # 将create_datetime转换为适合数据库的时间戳格式
            create_datetime_str = doc.get('create_datetime')
            values['create_datetime'] = datetime.strptime(create_datetime_str, '%Y-%m-%d %H:%M:%S') if create_datetime_str else None

            # 检查数据库中是否已存在相同title的记录
            check_query = "SELECT id FROM news_info WHERE name = %s;"
            cur.execute(check_query, (values['name'],))

            # 如果没有找到记录，则插入新记录
            if not cur.fetchone():
                insert_query = """
                    INSERT INTO news_info (name, context, source_url, link, article_type, article_source, img_url, keyword, posted_date, create_time, create_datetime)
                    VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s)
                    """
                cur.execute(insert_query, (
                    values['name'],
                    values['context'],
                    values['source_url'],
                    values['link'],
                    values['article_type'],
                    values['article_source'],
                    values['img_url'],
                    values['keyword'],
                    values['posted_date'],
                    values['create_time'],
                    values['create_datetime']
                ))
                conn.commit()
                print(f'已保存{values}')
        except Exception as e:
            print("Error during search: ", e)
        finally:
            # 关闭游标和连接
            if 'cur' in locals():
                cur.close()
    conn.close()


sorted_data = mongo()
if sorted_data:
    pg(sorted_data)

print("Done!")