def get_xbrl_from_file(file_path: str) -> DartXbrl:
""" XBRL 파일 로드 함수
XBRL 파일을 로드하기 위한 함수로 로딩완료 후 DartXbrl 클래스를 반환한다
Parameters
----------
file_path: str
XBRL 파일 경로
Returns
-------
DartXbrl
DartXbrl 클래스
"""
# PyPI를 통해 설치된 Arelle 라이브러리 사용시 Linux에서 발생하는 오류 수정을 위한코드
# Github 버전에서는 수정된 상태로 이후 삭제예정
if sys.platform == 'win32':
pass
elif sys.platform == 'darwin':
pass
else:
arelle_app_dir = os.path.join(os.path.expanduser("~/.config"),
"arelle")
if not os.path.exists(arelle_app_dir):
os.makedirs(arelle_app_dir)
model_xbrl = Cntlr.Cntlr().modelManager.load(file_path)
filename = file_path.split('\\')[-1]
return DartXbrl(filename, model_xbrl)
def main():
if IS_EXTRACTED:
pass
else:
edinet_zip_dir = os.path.join(EDINET_ROOT_DIR, "zip")
extract_files_from_zip(edinet_zip_dir,
dest_dir_root=EDINET_ROOT_DIR,
unzip_members_regrep="|".join([
f"XBRL/PublicDoc/.*\.{extension}"
for extension in ["xbrl", "xsd", "xml"]
]))
# XBRLから情報取得
xbrl_file_regrex = os.path.join(EDINET_ROOT_DIR, EDINET_XBRL_REGREX)
xbrl_files = glob.glob(xbrl_file_regrex)
list_dict_facts = []
ctrl = Cntlr.Cntlr()
model_manager = ModelManager.initialize(ctrl)
for index, xbrl_file in enumerate(xbrl_files):
print(xbrl_file, ":", index + 1, "/", len(xbrl_files))
list_dict_facts_per_file = get_facts(model_manager, xbrl_file)
if list_dict_facts_per_file is not None:
list_dict_facts = list_dict_facts + list_dict_facts_per_file
if list_dict_facts:
df_yuho = pd.DataFrame(list_dict_facts)
# Edinetコードリストの情報をマージ
df_edinetcd_info = get_edinetcd_info(EDINETCDDLINFO_COLS)
df_yuho = df_yuho.merge(df_edinetcd_info, on=EDINETCD_COL, how="left")
df_yuho.to_csv(os.path.join(EDINET_ROOT_DIR, OUTPUT_FILE_NAME),
index=False,
encoding="cp932")
print(f"{'-'*10} 情報抽出 完了 {'-'*10}")
else:
print("処理対象のデータはありませんでした。")
def get_current_df(xbrl_path):
arelle = Cntlr.Cntlr()
arelle.webCache.cacheDir = '/mnt/disks/disk-1/arelle/cache/'
arelle_xbrl = arelle.modelManager.load(xbrl_path)
arelle_df = pd.DataFrame(
data=[(fact, fact.value, fact.concept.qname, fact.context.hasSegment,
fact.context.period.viewText(), fact.contextID,
fact.context.isStartEndPeriod, fact.context.isInstantPeriod,
fact.context.isForeverPeriod, fact.context.startDatetime,
fact.context.endDatetime) for fact in arelle_xbrl.facts],
columns=(
'Fact', #Fact; We leave the fact in the table in case we want more out of it
'Value', #Value, like 5,200,000 in "5.2m"
'Account', #Account, like "Cash" in "Cash of 5.2m"
'Category', #Category, like "for the parent company" in "Cash of 5.2m for the parent company"b
'Period',
'ContextID',
'isStartEndPeriod',
'isInstantPeriod',
'isForeverPeriod',
'startDateTime',
'endDateTime'))
arelle_df = arelle_df[(
arelle_df.Fact.map(lambda f: f.isNumeric and not f.isNil)
)] # & #Fact is Numeric (i.e. can be converted to a number)
# arelle_df.Account.map(lambda a: a=='us-gaap:EarningsPerShareBasic') & #Just Cash; Account mentions "Cash", anywhere would give us too many values
# arelle_df.Category.map(lambda c:not c))] # Could also use "~", which does boolean "not" on the entire column
# file_path = arelle.webCache.getfilename(xbrl_path)
# if os.path.exists(file_path):
# shutil.move(file_path,os.path.join('/disk-1/arelle/cache',os.path.basename(file_path)))
return arelle_df #current_df
def main():
# ArelleCntlrSample().run()
# xbrl = Cntlr.Cntlr().modelManager.load('https://www.sec.gov/Archives/edgar/data/101984/000010198416000062/ueic-20151231.xml')
xbrl = Cntlr.Cntlr().modelManager.load('./nke-20190531.xml')
# ViewFileFactTable.viewFacts(xbrl, './ueic-20151231.csv')
# modelDoc = ModelDocument.load(xbrl,'./ueic-20151231.xml')
# mf = ModelInstanceObject.ModelFact()
for fact in xbrl.undefinedFacts:
print(fact)
def main():
if IS_EXTRACTED:
pass
else:
edinet_zip_dir = os.path.join(EDINET_ROOT_DIR, "zip")
extract_files_from_zip(edinet_zip_dir,
dest_dir_root=EDINET_ROOT_DIR,
unzip_members_regrep="|".join([
f"XBRL/PublicDoc/.*\.{extension}"
for extension in ["xbrl", "xsd", "xml"]
]))
# XBRLから情報取得
xbrl_file_regrex = os.path.join(EDINET_ROOT_DIR, EDINET_XBRL_REGREX)
xbrl_files = glob.glob(xbrl_file_regrex)
ctrl = Cntlr.Cntlr()
model_manager = ModelManager.initialize(ctrl)
for index, xbrl_file in enumerate(xbrl_files):
print(xbrl_file, ":", index + 1, "/", len(xbrl_files))
export_facts(model_manager, xbrl_file)
print(f"{'-'*10} XBRL出力 完了 {'-'*10}")
def get_xbrl_from_file(file_path: str) -> DartXbrl:
""" XBRL 파일 로드 함수
XBRL 파일을 로드하기 위한 함수로 로딩완료 후 DartXbrl 클래스를 반환한다
Parameters
----------
file_path: str
XBRL 파일 경로
Returns
-------
DartXbrl
DartXbrl 클래스
"""
# PyPI를 통해 설치된 Arelle 라이브러리 사용시 발생하는 오류 수정을 위한코드
from .spinner import Spinner
spinner = Spinner('XBRL Loading')
spinner.start()
if sys.platform == 'win32':
pass
elif sys.platform == 'darwin':
arelle_app_dir = os.path.join(
os.path.expanduser('~/Library/Application Support'), 'Arelle')
if not os.path.exists(arelle_app_dir):
os.makedirs(arelle_app_dir)
else:
arelle_app_dir = os.path.join(os.path.expanduser("~/.config"),
"arelle")
if not os.path.exists(arelle_app_dir):
os.makedirs(arelle_app_dir)
model_xbrl = Cntlr.Cntlr().modelManager.load(file_path)
filename = file_path.split('\\')[-1]
xbrl = DartXbrl(filename, model_xbrl)
spinner.stop()
return xbrl
def load_edinet():
edinet_list = []
# xbrl 読み込み
xbrl_files = glob.glob(const.FOLDER_XBRL + '*.xbrl')
# EdinetcodeDlInfo.csv 読み込み
edinetcodedata = pd.read_csv(const.FILE_EDINETCODE,
skiprows=1,
encoding='cp932')
for index, xbrl_file in enumerate(xbrl_files):
ctrl = Cntlr.Cntlr()
model_manager = ModelManager.initialize(ctrl)
model_xbrl = model_manager.load(xbrl_file)
print(xbrl_file, ":", index + 1, "/", len(xbrl_files))
# XBRLをDataFrameにセット
factData = pd.DataFrame(data=[
(fact.concept.qname.localName, fact.value, fact.isNumeric,
fact.contextID, fact.decimals) for fact in model_xbrl.facts
],
columns=[
'element_id', 'value', 'isNumeric',
'contextID', 'decimals'
])
# 取得したい要素
datalist = []
datalist_ren = []
flg_ren = False
for key, value in dict_cols.items():
contextRef = value['contextRef']
element_ids = value['element_id']
if len(element_ids) > 0:
for element_id in element_ids:
# data1 = factData[factData['element_id'].str.contains(element_id)]
data1 = []
data2 = []
data3 = []
data1 = factData[factData['element_id'] == element_id]
if len(data1) == 1: break
data2 = data1[data1['contextID'] == (
contextRef + '_NonConsolidatedMember')]
data3 = data1[data1['contextID'] == contextRef]
if len(data2) >= 1 or len(data3) >= 1: break
data, data_ren, unitRef = '', '', ''
if len(data1) == 1:
data = data1['value'].values[0]
data_ren = data
print(key + ' : ' + data)
if len(data2) >= 1 and len(data3) == 0:
data = data2['value'].values[0]
data_ren = data
print(key + ' : ' + data)
if len(data3) >= 1:
data = ''
data_ren = data3['value'].values[0]
print(key + '(連結) : ' + data_ren)
if len(data1) == 0:
print(key + ' : ' + 'NoData ')
data = ''
datalist.append(data)
if flg_ren:
datalist_ren.append(data_ren)
if key == 'EDINETコード':
company = edinetcodedata[edinetcodedata['ＥＤＩＮＥＴコード'] ==
data]
if len(company) == 1:
print('提出者業種 : ' + company['提出者業種'].values[0])
data = company['提出者業種'].values[0]
flg_ren = False
if company['連結の有無'].values[0] == '有':
print('連結の有無 : ' + company['連結の有無'].values[0])
flg_ren = True
else:
print('提出者業種 : ' + 'NoData ')
data = ''
datalist.append(data)
datalist.append('連結' if flg_ren else '単独')
if flg_ren:
# 連結用にデータをコピー
datalist_ren = datalist[:]
if flg_ren:
edinet_list.append(datalist_ren)
else:
edinet_list.append(datalist)
return edinet_list