python爬蟲爬去東方財富財務數據

python爬蟲爬去東方財富財務數據

import requests
import re
from multiprocessing import Pool
import json
import csv
import pandas as pd
import os
import time

# 設置文件保存在D盤eastmoney文件夾下
file_path = r'C:\Users\admir\Desktop\銀行競爭\報表數據'
if not os.path.exists(file_path):
    os.mkdir(file_path)
os.chdir(file_path)

# 1 設置表格爬取時期
def set_table():
    # 1 設置財務報表獲取時期
    year = int(float(input('請輸入要查詢的年份(四位數2007-2020)：\n')))
    # int表示取整，里面加float是因為輸入的是str，直接int會報錯，float則不會
    quarter = int(float(input('請輸入小寫數字季度(1:1季報，2-年中報，3：3季報，4-年報)：\n')))
    while (quarter < 1 or quarter > 4):
        quarter = int(float(input('季度數值輸入錯誤，請重新輸入：\n')))
    quarter = '{:02d}'.format(quarter * 3) 
    # 確定季度所對應的最后一天是30還是31號
    if (quarter == '06') or (quarter == '09'):
        day = 30
    else:
        day = 31
    date = '{}-{}-{}' .format(year, quarter, day)

    # 2 設置財務報表種類
    tables = int(
        input('請輸入查詢的報表種類對應的數字(1-業績報表；2-業績快報表：3-業績預告表；4-預約披露時間表；5-資產負債表；6-利潤表；7-現金流量表): \n'))

    dict_tables = {1: '業績報表', 2: '業績快報表', 3: '業績預告表',
                   4: '預約披露時間表', 5: '資產負債表', 6: '利潤表', 7: '現金流量表'}

    dict = {1: 'YJBB', 2: 'YJKB', 3: 'YJYG',
            4: 'YYPL', 5: 'ZCFZB', 6: 'LRB', 7: 'XJLLB'}
    category = dict[tables]

    # js請求參數里的type，第1-4個表的前綴是'YJBB20_'，后3個表是'CWBB_'
    # 設置set_table()中的type、st、sr、filter參數
    if tables == 1:
        category_type = 'YJBB20_'
        st = 'latestnoticedate'
        sr = -1
        filter =  "(securitytypecode in ('058001001','058001002'))(reportdate=^%s^)" %(date)
    elif tables == 2:
        category_type = 'YJBB20_'
        st = 'ldate'
        sr = -1
        filter = "(securitytypecode in ('058001001','058001002'))(rdate=^%s^)" %(date)
    elif tables == 3:
        category_type = 'YJBB20_'
        st = 'ndate'
        sr = -1
        filter=" (IsLatest='T')(enddate=^2018-06-30^)"
    elif tables == 4:
        category_type = 'YJBB20_'
        st = 'frdate'
        sr = 1
        filter =  "(securitytypecode ='058001001')(reportdate=^%s^)" %(date)
    else:
        category_type = 'CWBB_'
        st = 'noticedate'
        sr = -1
        filter = '(reportdate=^%s^)' % (date)

    category_type = category_type + category
    # print(category_type)
    # 設置set_table()中的filter參數

    yield{
    'date':date,
    'category':dict_tables[tables],
    'category_type':category_type,
    'st':st,
    'sr':sr,
    'filter':filter
    }

# 2 設置表格爬取起始頁數
def page_choose(page_all):

    # 選擇爬取頁數范圍    
    start_page = int(input('請輸入下載起始頁數：\n'))
    nums = input('請輸入要下載的頁數，（若需下載全部則按回車）：\n')
    print('*' * 80)

    # 判斷輸入的是數值還是回車空格
    if nums.isdigit():
        end_page = start_page + int(nums)
    elif nums == '':
        end_page = int(page_all.group(1))
    else:
        print('頁數輸入錯誤')

    # 返回所需的起始頁數，供后續程序調用
    yield{
        'start_page': start_page,
        'end_page': end_page
    }

# 3 表格正式爬取
def get_table(date, category_type,st,sr,filter,page):
    # 參數設置
    params = {
        # 'type': 'CWBB_LRB',
        'type': category_type,  # 表格類型
        'token': '70f12f2f4f091e459a279469fe49eca5',
        'st': st,
        'sr': sr,
        'p': page,
        'ps': 50,  # 每頁顯示多少條信息
        'js': 'var LFtlXDqn={pages:(tp),data: (x)}',
        'filter': filter,
        # 'rt': 51294261  可不用
    }
    url = 'http://dcfm.eastmoney.com/em_mutisvcexpandinterface/api/js/get?'

    # print(url)
    response = requests.get(url, params=params).text
    # print(response)
    # 確定頁數
    pat = re.compile('var.*?{pages:(\d+),data:.*?')
    page_all = re.search(pat, response)
    print(page_all.group(1))  # ok

    # 提取{},json.loads出錯
    # pattern = re.compile('var.*?data: \[(.*)]}', re.S)

    # 提取出list，可以使用json.dumps和json.loads
    pattern = re.compile('var.*?data: (.*)}', re.S)
    items = re.search(pattern, response)
    # 等價於
    # items = re.findall(pattern,response)
    # print(items[0])
    data = items.group(1)
    data = json.loads(data)
    # data = json.dumps(data,ensure_ascii=False)

    return page_all, data,page

# 寫入表頭
# 方法1 借助csv包，最常用
def write_header(data,category):
    with open('{}.csv' .format(category), 'a', encoding='utf_8_sig', newline='') as f:
        headers = list(data[0].keys())
        # print(headers)  # 測試 ok
        writer = csv.writer(f)
        writer.writerow(headers)

def write_table(data,page,category):
    print('\n正在下載第 %s 頁表格' % page)
    # 寫入文件方法1
    for d in data:
        with open('{}.csv' .format(category), 'a', encoding='utf_8_sig', newline='') as f:
            w = csv.writer(f)
            w.writerow(d.values())

def main(date, category_type,st,sr,filter,page):
    func = get_table(date, category_type,st,sr,filter,page)
    data = func[1]
    page = func[2]
    write_table(data,page,category)


if __name__ == '__main__':
    # 獲取總頁數，確定起始爬取頁數
    for i in set_table():
        date = i.get('date')
        category = i.get('category')
        category_type = i.get('category_type')
        st = i.get('st')
        sr = i.get('sr')
        filter = i.get('filter')

    constant = get_table(date,category_type,st,sr,filter, 1)
    page_all = constant[0]

    for i in page_choose(page_all):
        start_page = i.get('start_page')
        end_page = i.get('end_page')

    # 寫入表頭
    write_header(constant[1],category)
    start_time = time.time()  # 下載開始時間
    # 爬取表格主程序
    for page in range(start_page, end_page):
        main(date,category_type,st,sr,filter, page)

    end_time = time.time() - start_time  # 結束時間
    print('下載完成')
    print('下載用時: {:.1f} s' .format(end_time))

https://github.com/makcyun/eastmoney_spider

• 發表於: 2018-10-13
• 原文鏈接：https://kuaibao.qq.com/s/20181013G1EQ5V00?refer=cp_1026
• 騰訊「雲+社區」是騰訊內容開放平台帳號（企鵝號）傳播渠道之一，根據《騰訊內容開放平台服務協議》轉載發布內容。
• 如有侵權，請聯系 [e-mail] 刪除。