topic_spider/find_source/common/databases.py

import bson
import pymongo
import redis
import requests
from elasticsearch import Elasticsearch

from config.load import mongo_conf, redis_conf, es_conf, analyze_url

# ---------------------------------- mongo ----------------------------------
MONGO_URI_CLIENTS = {}    # a dictionary hold all client with uri as key


def mongo_client(cfg=None, host=None, port=None, fork=False, **kwargs):
    if host is not None and port is not None:
        uri = f'mongodb://{host}:{port}'
    else:
        _cfg = (cfg or mongo_conf)
        uri = f'mongodb://{_cfg["host"]}:{_cfg["port"]}'

    if fork:
        return pymongo.MongoClient(uri, **kwargs)
    global MONGO_URI_CLIENTS
    matched_client = MONGO_URI_CLIENTS.get(uri)
    if matched_client is None:
        new_client = pymongo.MongoClient(uri, **kwargs)
        if new_client is not None:
            MONGO_URI_CLIENTS[uri] = new_client
        return new_client
    return matched_client


def mongo_database(name: str, **kw):
    client = mongo_client(**kw)
    return client.get_database(name)


def mongo_table(db: str, name: str, **kw):
    database = mongo_database(db, **kw)
    return database.get_collection(name)


def int2long(param: int):
    """int 转换成 long """
    return bson.int64.Int64(param)


def object_id(_id: str):
    return bson.objectid.ObjectId(_id)


# ---------------------------------- es ----------------------------------
def es_client(cfg=None):
    if cfg is None:
        cfg = es_conf
    return Elasticsearch([{"host": cfg['host'], "port": cfg['port']}])


def es_participles_service(text: str):
    """
    获取文本的分词列表

    :param text: 需要分词的文本
    :return: 分词列表
    """
    result = []
    params = {"text": text, "analyzer": "ik_smart"}
    res = requests.get(analyze_url, params=params, timeout=60)
    if res.status_code == 200:
        tokens = res.json().get('tokens', [])
        for x in tokens:
            if x["token"].encode('utf-8').isalpha():
                continue
            result.append(x["token"])
    return result


def es_query(title: str, publish_time: int):
    """
    查询es

    :param title: 标题
    :param publish_time: 发布时间
    :return:
    """
    client = es_client()
    stime = publish_time - 432000  # 往前推5天
    etime = publish_time + 432000
    conditions = []
    participles = es_participles_service(title)
    for word in participles:
        conditions.append({
            "multi_match": {
                "query": word,
                "type": "phrase",
                "fields": ["title"]
            }
        })
    conditions.append({
        "range": {"publishtime": {"from": stime, "to": etime}}
    })
    query = {
        "query": {
            "bool": {
                "must": conditions,
                "minimum_should_match": 1
            }
        }
    }
    result = client.search(index='bidding', body=query, request_timeout=100)
    count = len(result['hits']['hits'])
    return count


# ---------------------------------- redis ----------------------------------
def redis_client(cfg=None, host=None, port=None, db=None, password=None):
    if all([host is not None, port is not None, db is not None]):
        host, port, pwd, db = host, port, password, db
    else:
        _cfg = (cfg or redis_conf)
        host, port, pwd, db = _cfg['host'], _cfg['port'], _cfg['pwd'], _cfg['db']
    pool = redis.ConnectionPool(
        host=host,
        port=port,
        password=pwd,
        db=db
    )
    return redis.Redis(connection_pool=pool, decode_responses=True)