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_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_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_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