match_spider/ybw/utils/tools.py

import socket
import hashlib
import time
import re


def clean_title(title):
    if title:
        rule_list = [
            '\(\d{1,20}\)',
            '\[[\u4e00-\u9fa5]{1,9}\]',
            '【[\u4e00-\u9fa5]{1,9}】',
        ]
        for rule in rule_list:
            title = re.sub(rule, '', title)

    return title


def get_host_ip():
    s = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
    try:
        s.connect(('8.8.8.8', 80))
        ip = s.getsockname()[0]
    finally:
        s.close()
    return ip


def sha1(text: str):
    """
    十六进制数字字符串形式摘要值

    @param text: 字符串文本
    @return: 摘要值
    """
    _sha1 = hashlib.sha1()
    _sha1.update(text.encode("utf-8"))
    return _sha1.hexdigest()


def wait(interval=None):
    _interval = (interval or 600)
    time.sleep(_interval)