搭建自己的金融数据源和量化分析平台(六):下载并存储沪深两市上市公司财报

基于不依赖wind、某花顺等第三方平台数据的考虑,尝试直接从财报中解析三大报表进而计算ROE等财务指标,因此需要下载沪深两市的上市公司财报数据,便于后续从pdf中解析三大报表。

深市爬虫好做,先放深市爬虫:

注:同一个IP频繁向两市服务器发起请求会导致HTTPSConnectionPool(host='xxxx', port=443): Max retries exceeded with url:报错

解决方法有两个:一是使用代理池,二是爬虫休眠。

python 复制代码
'''
根据时间段下载深交所上市公司财报
path str 指定财报存储路径
time str 财报年度 如[2023,2024]
stock_list list 下载财报的股票代码列表 例如['000001','000002']
financial_statements_type list 财报的类别 例如['annual','semi-annual','quarterly_1','quarterly_3'] 分别为年报、半年报、一季报、三季报
'''
def get_financial_statements(path, time, stock_list, financial_statements_type):
    url = "https://www.szse.cn/api/disc/announcement/annList"
    headers = {
        'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.3',
        'Content-Type': 'application/json',
        'Connection': 'close'
    }
    download_url = "https://disc.static.szse.cn/download"
    # 逐只股票读取相应pdf报表
    for stock in stock_list:
        # 逐年循环
        for year in time:
            # 根据财报类型逐个读取pdf
            for fs_type in financial_statements_type:
                if fs_type == 'annual':
                    title = "年报"
                    bigCategoryId = '010301'   # 年报查询代码
                    timestart = str(year)+"-12-31"
                    timeend = str(year+1)+"-09-01"  # 防止出现财报更正之后时间节点覆盖不到,统一往后推三个月
                elif fs_type == 'semi-annual':
                    title = "中报"
                    bigCategoryId = '010303'   # 中报查询代码
                    timestart = str(year) + "-07-01"
                    timeend = str(year) + "-12-31"
                elif fs_type == 'quarterly_1':
                    title = "一季报"
                    bigCategoryId = '010305'   # 一季报查询代码
                    timestart = str(year) + "-04-01"
                    timeend = str(year) + "-07-31"
                else:
                    title = "三季报"
                    bigCategoryId = '010307'   # 三季报查询代码
                    timestart = str(year) + "-10-01"
                    timeend = str(year) + "-12-31"
                data = {
                    "seDate": [timestart, timeend],
                    "stock": [stock],
                    "channelCode": ["listedNotice_disc"],
                    "bigCategoryId": [bigCategoryId],
                    "pageSize": 50,
                    "pageNum": 1
                }
                response = requests.post(url=url, data=json.dumps(data), headers=headers)
                data = json.loads(response.text)["data"]
                if len(data) == 0 or data is None:
                    print("警告:股票代码:"+stock+" "+str(year)+title+"不存在!")
                else:
                    for entry in data:
                        # 对摘要栏目做特殊处理
                        if entry['title'].find("报告摘要") < 0:
                            # 检查path路径下stock代码文件夹、年份文件夹是否存在,不存在则创建
                            file_path = path+stock+"/"+str(year)
                            if Tools.check_folder_exists(path+stock) == False:
                                os.mkdir(path+stock)
                            if Tools.check_folder_exists(file_path) == False:
                                os.mkdir(file_path)
                            file = file_path + "/" + str(year) + title + "##" + entry['title'].replace("*", "") + ".pdf"
                            # 检查文件是否已存在,不存在再下载
                            if os.path.exists(file):
                                print("警告:股票代码:" + stock + " " + str(year) + title + "已存在!")
                            else:
                                filecontent = requests.get(download_url + entry["attachPath"])
                                with open(file, "wb") as pdf:
                                    pdf.write(filecontent.content)
                                print("股票代码:" + stock + " " + str(year) + title + "写入成功。")

# 爬虫调用实例:
# timestart = [2023,2024]
# stock_list = ['000001','000002']
# financial_statements_type = ['annual', 'semi-annual', 'quarterly_1', 'quarterly_3']
# SZ_financial_statement_path = "F:/data/SZ/"
# get_financial_statements(SZ_financial_statement_path, timestart,stock_list,financial_statements_type)

沪市爬虫:

python 复制代码
'''
根据时间段下载上交所上市公司财报
time str 财报年度 如2024、2023
stock_list list 下载财报的股票代码列表 例如['000001','000002']
financial_statements_type list 财报的类别 例如['annual','semi-annual','quarterly_1','quarterly_3'] 分别为年报、半年报、一季报、三季报
'''
def get_financial_statements(path, time, stock_list, financial_statements_type):
    headers = {
        'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.3',
        'Referer': 'https://www.sse.com.cn/',
        'Connection': 'close'
    }
    download_url = "https://www.sse.com.cn"
    # 逐只股票读取相应pdf报表
    for stock in stock_list:
        # 逐年循环
        for year in time:
            # 根据财报类型逐个读取pdf
            for fs_type in financial_statements_type:
                if fs_type == 'annual':
                    title = "年报"
                    bigCategoryId = 'YEARLY'  # 年报查询代码
                    timestart = str(year) + "-12-31"
                    timeend = str(year + 1) + "-09-01"  # 防止出现财报更正之后时间节点覆盖不到,统一往后推三个月
                elif fs_type == 'semi-annual':
                    title = "中报"
                    bigCategoryId = 'QUATER2'  # 中报查询代码
                    timestart = str(year) + "-07-01"
                    timeend = str(year) + "-12-31"
                elif fs_type == 'quarterly_1':
                    title = "一季报"
                    bigCategoryId = 'QUATER1'  # 一季报查询代码
                    timestart = str(year) + "-04-01"
                    timeend = str(year) + "-07-31"
                else:
                    title = "三季报"
                    bigCategoryId = 'QUATER3'  # 三季报查询代码
                    timestart = str(year) + "-10-01"
                    timeend = str(year) + "-12-31"
                url = "https://query.sse.com.cn/security/stock/queryCompanyBulletin.do?jsonCallBack=jsonpCallback"+str(random.randint(10000, 999999))+"&isPagination=true&pageHelp.pageSize=50&pageHelp.pageNo=1&pageHelp.beginPage=1&pageHelp.cacheSize=1&pageHelp.endPage=1&productId="+stock+"&securityType=0101%2C120100%2C020100%2C020200%2C120200&reportType2=DQBG&reportType="+bigCategoryId+"&beginDate="+timestart+"&endDate="+timeend
                response = requests.get(url=url, headers=headers)
                datas = json.loads(response.text.split('"keyWord":null,"pageHelp":')[1].split(',"productId":')[0])['data']
                if len(datas) == 0 or datas is None:
                    print("警告:股票代码:" + stock + " " + str(year) + title + "不存在!")
                else:
                    for entry in datas:
                        # 对摘要栏目做特殊处理,去除摘要
                        if entry['TITLE'].find("摘要") < 0:
                            # 检查path路径下stock代码文件夹、年份文件夹是否存在,不存在则创建
                            file_path = path + stock + "/" + str(year)
                            if Tools.check_folder_exists(path + stock) == False:
                                os.mkdir(path + stock)
                            if Tools.check_folder_exists(file_path) == False:
                                os.mkdir(file_path)
                            file = file_path + "/" + str(year) + title + "##" + entry['TITLE'].replace("*", "") + ".pdf"
                            # 检查文件是否已存在,不存在再下载
                            if os.path.exists(file):
                                print("警告:股票代码:" + stock + " " + str(year) + title + "已存在!")
                            else:
                                filecontent = requests.get(download_url + entry["URL"])
                                with open(file, "wb") as pdf:
                                    pdf.write(filecontent.content)
                                print("股票代码:" + stock + " " + str(year) + title + "写入成功。")
# timestart = [2023]
# stock_list = ['600011']
# financial_statements_type = ['annual', 'semi-annual', 'quarterly_1', 'quarterly_3']
# SZ_financial_statement_path = "F:/data/SH/"
# get_financial_statements(SZ_financial_statement_path, timestart,stock_list,financial_statements_type)

控制模块代码:

python 复制代码
#更新A股股票财报数据
def update_A_financial_data(SZ=False,SH=False,BJ=False,time=[],financial_statements_type=[]):
    database = "stock_a"
    if SZ == True:
        # 读取深交所上市处于上市状态的公司股票代码
        select_sql = "select stock_code from stock_list where exchange = '"+SZSE+"' and list_status = '"+LIST+"'"
        select_result = ExecSelect(database, select_sql)  # 读取查询结果
        stocks = []
        for stock in select_result:
            stocks.append(stock[0])
        A_SZ_basic.get_financial_statements(SZ_financial_statement_path, time, stocks, financial_statements_type)
    if SH == True:
        # 读取上交所上市处于上市状态的公司股票代码
        select_sql = "select stock_code from stock_list where exchange = '" + SSE + "' and list_status = '" + LIST + "'"
        select_result = ExecSelect(database, select_sql)  # 读取查询结果
        stocks = []
        for stock in select_result:
            stocks.append(stock[0])
        A_SH_basic.get_financial_statements(SH_financial_statement_path, time, stocks, financial_statements_type)
    if BJ == True:
        pass
相关推荐
Hotchip华芯邦科技1 天前
MEMS硅麦克风应用电子烟雾化产业稳步爬升,耐高温、 防油、防酸、防腐蚀等性能优势和可实现自动化贴片及极高的一致性等特性使其必将成为主流
科技·单片机·金融·生活·社交电子·健康医疗·制造
gooxi_hui2 天前
Gooxi受邀参加海通证券AI+应用生态大会,助力数智金融高质量发展
人工智能·金融
君臣Andy2 天前
金融量化框架选择
金融
AIGC方案2 天前
免费下载 | 2024年金融数字化转型白皮书
金融
中国云报2 天前
金融数据中心容灾“大咖说” | Veritas的“高举高打”之道
金融
lyw_YTU_Sussex2 天前
人工智能在金融领域的创新与应用
人工智能·金融
NiNg_1_2342 天前
Python中Tushare(金融数据库)入门详解
数据库·python·金融
中国云报2 天前
金融数据中心容灾“大咖说” | 英方软件的“分而治之”之道
金融
菩提祖师_3 天前
人工智能在金融领域的应用与风险防范研究
人工智能·百度·金融
fgl1004 天前
富格林:安全指正规防欺诈套路
金融