def crawl_DIV_type(self, year, stocktype): url = 'https://mops.twse.com.tw/server-java/t05st09sub' form_data = { 'encodeURIComponent': 1, 'step': 1, 'TYPEK': stocktype, 'YEAR': common.year_CE2RC(year), 'first': '', 'qryType': 2, } # 拆解內容 table_array = common.crawl_data2text(url, form_data, 'big5').split('<table') dfDIV = pd.DataFrame() for table in table_array: if '公司代號' in table: tr_array = table.split('<tr') for tr in tr_array: td_array = tr.split('<td') if len(td_array) > 15: # 公司代號 ticker = common.col_clear( td_array[1]).split('-')[0].strip() # 所屬年度 intyr, vaild = common.TryParse( 'int', common.col_clear(td_array[3]).split('年')[0]) yr = common.year_RC2CE(intyr) # 現金股利 CD, vaild = common.TryParse( 'float', common.col_clear(td_array[12])) # 股票股利 SD, vaild = common.TryParse( 'float', common.col_clear(td_array[15])) # 判斷是否有該公司當年度資料,更新/新增 index = (ticker, yr) if len(dfDIV.index) > 0 and index in dfDIV.index: data = dfDIV.loc[index] data[0] = data[0] + CD data[1] = data[1] + SD else: df = pd.DataFrame(data=[[CD, SD]], index=pd.MultiIndex.from_tuples( [index]), columns=['CD', 'SD']) df.index.set_names(['Ticker', 'yr'], inplace=True) dfDIV = dfDIV.append(df) return dfDIV
def crawl_Holiday(self, year): url = f'https://www.twse.com.tw/holidaySchedule/holidaySchedule?response=csv&queryYear={common.year_CE2RC(year)}' rText = common.crawl_data2text(url, '', 'big5', delay=3000).replace( '=', '').replace('\r', '').replace('"', '') # 整理資料,變成表格 df = pd.read_csv(StringIO(rText), header=['名稱' in l for l in rText.split("\n")].index(True), index_col=False, names=['name', 'date', 'weekday', 'comment'], dtype={'date': str}) df['Closed'] = df['name'].apply(lambda x: 'E' if '農曆春節後開始交易日' in x else ('S' if '農曆春節前最後交易日' in x else '')) index = [] lsDates = [] springdate = {'SDate': None, 'EDate': None} for i, row in df.iterrows(): if row['Closed'] != '': tmpdate = arrow.get( str(year) + '/' + row['date'].replace('月', '/').replace('日', '').strip()) if row['Closed'] == 'S' and springdate['SDate'] == None: springdate['SDate'] = tmpdate elif row['Closed'] == 'E': springdate['EDate'] = tmpdate else: for dd in str(row['date']).replace('月', '/').split('日'): if dd != '': try: lsDates.append( arrow.get(str(year) + '/' + dd).format('YYYYMMDD')) index.append(year) except: continue # spring ckdate = springdate['SDate'].shift(days=1) while ckdate < springdate['EDate']: strDate = ckdate.format('YYYYMMDD') if strDate not in lsDates: lsDates.append(strDate) index.append(year) ckdate = ckdate.shift(days=1) df = pd.DataFrame(data=lsDates, index=index, columns=['date']) df.index.set_names('yr', inplace=True) return df
def crawl_DQ(self, date): strDate = date.format('YYYYMMDD') # 下載股價 url = f'https://www.twse.com.tw/exchangeReport/MI_INDEX?response=csv&date={strDate}&type=ALL' rText = common.crawl_data2text(url, '', 'big5', delay=3000).replace( '=', '').replace('\r', '') def asFloat(x): val = x.replace(',', '') fval, vaild = common.TryParse('float', val) return fval if vaild else vaild # 整理資料,變成表格 df = pd.read_csv(StringIO(rText), header=['證券代號' in l for l in rText.split("\n")].index(True) - 1, index_col=False, converters={ '成交股數': asFloat, '成交筆數': asFloat, '成交金額': asFloat, '開盤價': asFloat, '最高價': asFloat, '最低價': asFloat, '收盤價': asFloat, '漲跌價差': asFloat, '最後揭示買價': asFloat, '最後揭示買量': asFloat, '最後揭示賣價': asFloat, '最後揭示賣量': asFloat, '本益比': asFloat, }, usecols=[ '證券代號', '成交股數', '成交筆數', '成交金額', '開盤價', '最高價', '最低價', '收盤價', '漲跌(+/-)', '漲跌價差', '最後揭示買價', '最後揭示買量', '最後揭示賣價', '最後揭示賣量', '本益比' ]) df.columns = [ 'Ticker', 'TVol', 'TXN', 'TV', 'OP', 'HP', 'LP', 'CP', 'Dir', 'CHG', 'LBBP', 'LBBV', 'LBSP', 'LBSV', 'PER' ] df['Date'] = strDate df = df.set_index(['Ticker', 'Date']) return df.sort_index()
def crawl_comInfo_type(self, stocktype): url = 'https://mops.twse.com.tw/mops/web/ajax_t51sb01' form_data = { 'encodeURIComponent': 1, 'step': 1, 'firstin': 1, 'TYPEK': stocktype, 'code': '', } # 拆解內容 table_array = common.crawl_data2text(url, form_data).split('<table') tr_array = table_array[2].split('<tr') # 拆解td data, index = [], [] for i in range(len(tr_array)): td_array = tr_array[i].split('<td') if (len(td_array) > 17): # 公司代號, 公司名稱, 公司簡稱, 產業類別 Ticker = common.col_clear(td_array[1]) ComName = common.col_clear(td_array[2]) Com = common.col_clear(td_array[3]) IC = common.col_clear(td_array[4]) ESTD = common.date_RC2CE(common.col_clear(td_array[14])) LISTD = common.date_RC2CE(common.col_clear(td_array[15])) AoC = common.col_clear(td_array[17]) index.append(Ticker) data.append([ComName, Com, IC, ESTD, LISTD, AoC]) dfComInfo = pd.DataFrame( data=data, index=index, columns=['ComName', 'Com', 'IC', 'ESTD', 'LISTD', 'AoC']) dfComInfo.index.set_names('Ticker', inplace=True) return dfComInfo
def crawl_FSA_type(self, year, stocktype): url = 'https://mops.twse.com.tw/mops/web/ajax_t51sb02' form_data = { 'encodeURIComponent': 1, 'run': 'Y', 'step': 1, 'TYPEK': stocktype, 'year': common.year_CE2RC(year), 'isnew': '', 'firstin': 1, 'off': 1, 'ifrs': 'Y', } # 拆解內容 table_array = common.crawl_data2text(url, form_data).split('<table') dfFSA = pd.DataFrame() if len(table_array) < 3: return dfFSA tr_array = table_array[3].split('<tr') for tr in tr_array: td_array = tr.split('<td') if len(td_array) > 15: # 公司代號 ticker = common.col_clear(td_array[1]).split('-')[0].strip() # 負債占資產比率 DR, vaild = common.TryParse('float', common.col_clear(td_array[3])) # 長期資金佔不動產廠房及設備比率 LER, vaild = common.TryParse('float', common.col_clear(td_array[4])) # 流動比率 CR, vaild = common.TryParse('float', common.col_clear(td_array[5])) # 速動比率 UR, vaild = common.TryParse('float', common.col_clear(td_array[6])) # 利息保障倍數 IPM, vaild = common.TryParse('float', common.col_clear(td_array[7])) # 應收款項周轉率 ARTR, vaild = common.TryParse('float', common.col_clear(td_array[8])) # 平均收現日數 ACCD, vaild = common.TryParse('float', common.col_clear(td_array[9])) # 存貨週轉率(次) ITR, vaild = common.TryParse('float', common.col_clear(td_array[10])) # 平均銷貨日數 ASD, vaild = common.TryParse('float', common.col_clear(td_array[11])) # 不動產廠房及設備週轉率(次) PETR, vaild = common.TryParse('float', common.col_clear(td_array[12])) # 總資產週轉率(次) TATR, vaild = common.TryParse('float', common.col_clear(td_array[13])) # 資產報酬率(%) ROA, vaild = common.TryParse('float', common.col_clear(td_array[14])) # 權益報酬率(%) ROE, vaild = common.TryParse('float', common.col_clear(td_array[15])) # 稅前純益佔實收資本比率(%) NPBT2PCR, vaild = common.TryParse( 'float', common.col_clear(td_array[16])) # 純益率(%) NPR, vaild = common.TryParse('float', common.col_clear(td_array[17])) # 每股盈餘(元) EPS, vaild = common.TryParse('float', common.col_clear(td_array[18])) # 現金流量比率(%) CFR, vaild = common.TryParse('float', common.col_clear(td_array[19])) # 現金流量允當比率(%) CFAR, vaild = common.TryParse('float', common.col_clear(td_array[20])) # 現金再投資比率(%) CRR, vaild = common.TryParse('float', common.col_clear(td_array[21])) # 判斷是否有該公司當年度資料,更新/新增 index = (ticker, common.year_RC2CE(year)) data = [ DR, LER, CR, UR, IPM, ARTR, ACCD, ITR, ASD, PETR, TATR, ROA, ROE, NPBT2PCR, NPR, EPS, CFR, CFAR, CRR ] df = pd.DataFrame(data=[data], index=pd.MultiIndex.from_tuples([index]), columns=[ 'DR', 'LER', 'CR', 'UR', 'IPM', 'ARTR', 'ACCD', 'ITR', 'ASD', 'PETR', 'TATR', 'ROA', 'ROE', 'NPBT2PCR', 'NPR', 'EPS', 'CFR', 'CFAR', 'CRR' ]) df.index.set_names(['Ticker', 'yr'], inplace=True) dfFSA = dfFSA.append(df) return dfFSA
def crawl_BS_type(self, year, season, stocktype): url = 'https://mops.twse.com.tw/mops/web/ajax_t163sb05' form_data = { 'encodeURIComponent': 1, 'step': 1, 'firstin': 1, 'off': 1, 'isQuery': 'Y', 'TYPEK': stocktype, 'year': common.year_CE2RC(year), 'season': season } # 拆解內容 table_array = common.crawl_data2text(url, form_data).split('<table') dfBS = pd.DataFrame() dtTitle = { 'TA': ['資產總額', '資產總計'], 'TL': ['負債總計', '負債總額'], 'TE': ['權益總計', '權益總額'], 'RNP': ['每股參考淨值'], 'CA': ['流動資產'], 'NCA': ['非流動資產'], 'CL': ['流動負債'], 'NCL': ['非流動負債'] } for table in table_array: if '代號</th>' in table: tr_array = table.split('<tr') dtIndex = { 'TA': -1, 'TL': -1, 'TE': -1, 'RNP': -1, 'CA': -1, 'NCA': -1, 'CL': -1, 'NCL': -1, } for tr in tr_array: if '<th' in tr: th_array = tr.split('<th') for thIndex in range(1, len(th_array)): title = common.col_clear(th_array[thIndex]).strip() for key in dtTitle.keys(): if title in dtTitle[key]: dtIndex[key] = thIndex continue td_array = tr.split('<td') if len(td_array) > 1: #公司代號, 年, 季 ticker = common.col_clear(td_array[1]) index = (ticker, common.year_RC2CE(year), season) dtData = { 'TA': 0, 'TL': 0, 'TE': 0, 'RNP': 0, 'CA': 0, 'NCA': 0, 'CL': 0, 'NCL': 0, } for key in dtIndex.keys(): if dtIndex[key] >= 0: val, vaild = common.TryParse( 'float', common.col_clear(td_array[dtIndex[key]])) dtData[key] = val data = [ dtData['TA'], dtData['TL'], dtData['TE'], dtData['RNP'], dtData['CA'], dtData['NCA'], dtData['CL'], dtData['NCL'] ] df = pd.DataFrame( data=[data], index=pd.MultiIndex.from_tuples([index]), columns=[ 'TA', 'TL', 'TE', 'RNper', 'CA', 'NCA', 'CL', 'NCL' ]) df.index.set_names(['Ticker', 'yr', 'qtr'], inplace=True) dfBS = dfBS.append(df) return dfBS
def crawl_SCI_type(self, year, season, stocktype): url = 'https://mops.twse.com.tw/mops/web/ajax_t163sb04' form_data = { 'encodeURIComponent': 1, 'step': 1, 'firstin': 1, 'TYPEK': stocktype, 'code': '', 'year': common.year_CE2RC(year), 'season': season } # 拆解內容 table_array = common.crawl_data2text(url, form_data).split('<table') dfcomprehensiveIncome = pd.DataFrame() dtTitle = { 'Rev1': ['利息淨收益', '營業收入', '淨收益', '收益', '收入'], 'Rev2': ['利息以外淨損益'], 'GP': ['營業毛利(毛損)'], 'OP': ['營業利益(損失)', '營業利益'], 'NPBT': ['繼續營業單位稅前淨利(淨損)', '稅前淨利(淨損)', '繼續營業單位稅前損益', '繼續營業單位稅前純益(純損)'], 'NPAT': ['本期稅後淨利(淨損)', '本期淨利(淨損)'], 'NPPC': ['淨利(損)歸屬於母公司業主', '淨利(淨損)歸屬於母公司業主'], 'EPS': ['基本每股盈餘(元)'] } for table in table_array: if '代號</th>' in table: tr_array = table.split('<tr') dtIndex = { 'Rev1': -1, 'Rev2': -1, 'GP': -1, 'OP': -1, 'NPBT': -1, 'NPAT': -1, 'NPPC': -1, 'EPS': -1 } for tr in tr_array: if '<th' in tr: th_array = tr.split('<th') for thIndex in range(1, len(th_array)): title = common.col_clear(th_array[thIndex]).strip() for key in dtTitle.keys(): if title in dtTitle[key]: dtIndex[key] = thIndex continue td_array = tr.split('<td') if len(td_array) > 1: #公司代號, 年, 季 ticker = common.col_clear(td_array[1]) index = (ticker, common.year_RC2CE(year), season) if index not in dfcomprehensiveIncome.index: dtData = { 'Rev1': 0, 'Rev2': 0, 'GP': 0, 'OP': 0, 'NPBT': 0, 'NPAT': 0, 'NPPC': 0, 'EPS': 0 } for key in dtIndex.keys(): if dtIndex[key] >= 0: val, vaild = common.TryParse( 'float', common.col_clear( td_array[dtIndex[key]])) dtData[key] = val data = [ dtData['Rev1'] + dtData['Rev2'], dtData['GP'], dtData['OP'] if dtData['OP'] > 0 else dtData['NPBT'], dtData['NPBT'], dtData['NPAT'], dtData['NPPC'], dtData['EPS'] ] df = pd.DataFrame( data=[data], index=pd.MultiIndex.from_tuples([index]), columns=[ 'Rev', 'GP', 'OP', 'NPBT', 'NPAT', 'NPPC', 'EPS' ]) df['GM'] = df['GP'] / df['Rev'] df.index.set_names(['Ticker', 'yr', 'qtr'], inplace=True) dfcomprehensiveIncome = dfcomprehensiveIncome.append( df) return dfcomprehensiveIncome