crawlab_feader/NoteWork/cesspider/magpces.py

# -*- coding: utf-8 -*-
"""
Created on 2018-09-06 14:21
---------
@summary: 工具
---------
@author: Boris
@email: boris_liu@foxmail.com
"""
import datetime
import json
import re
from pprint import pformat
print('sssssssss')
_regexs = {}
def get_info(html, regexs, allow_repeat=True, fetch_one=False, split=None):
    regexs = isinstance(regexs, str) and [regexs] or regexs

    infos = []
    for regex in regexs:
        if regex == "":
            continue

        if regex not in _regexs.keys():
            _regexs[regex] = re.compile(regex, re.S)

        if fetch_one:
            infos = _regexs[regex].search(html)
            if infos:
                infos = infos.groups()
            else:
                continue
        else:
            infos = _regexs[regex].findall(str(html))

        if len(infos) > 0:
            # print(regex)
            break

    if fetch_one:
        infos = infos if infos else ("",)
        return infos if len(infos) > 1 else infos[0]
    else:
        infos = allow_repeat and infos or sorted(set(infos), key=infos.index)
        infos = split.join(infos) if split else infos
        return infos
def get_json(json_str):
    """
    @summary: 取json对象
    ---------
    @param json_str: json格式的字符串
    ---------
    @result: 返回json对象
    """

    try:
        return json.loads(json_str) if json_str else {}
    except Exception as e1:
        try:
            json_str = json_str.strip()
            json_str = json_str.replace("'", '"')
            keys = get_info(json_str, "(\w+):")
            for key in keys:
                json_str = json_str.replace(key, '"%s"' % key)

            return json.loads(json_str) if json_str else {}

        except Exception as e2:
            print(
                """
                e1: %s
                format json_str: %s
                e2: %s
                """
                % (e1, json_str, e2)
            )

        return {}


def dumps_json(json_, indent=4, sort_keys=False):
    """
    @summary: 格式化json 用于打印
    ---------
    @param json_: json格式的字符串或json对象
    ---------
    @result: 格式化后的字符串
    """
    try:
        if isinstance(json_, str):
            json_ = get_json(json_)

        json_ = json.dumps(
            json_, ensure_ascii=False, indent=indent, skipkeys=True, sort_keys=sort_keys
        )

    except Exception as e:
        print(e)
        json_ = pformat(json_)

    return json_
def get_current_date(date_format="%Y-%m-%d %H:%M:%S"):
    return datetime.datetime.now().strftime(date_format)
    # return time.strftime(date_format, time.localtime(time.time()))
def key2hump(key):
    """
    下划线试变成首字母大写
    """
    return key.title().replace("_", "")