def getItrInfos(edp):
logging.info('Getting ITR Infos for %s', edp)
# Getting newer version of documents
itrs = dict([(i['date'], i) for i in sorted(getITRs(edp), cmp=mycmp) if i['version'] > 0.0])
for dt, itr in itrs.iteritems():
if dt.year < 2015:
logging.info('Ignoring ITR of %s', dt.year)
continue
tm = '{:%Y%m}'.format(dt)
logging.info('Getting ITR of %s for %s', edp, tm)
# Get NSD and NSR
url_fdf = RE_LOC.search(openUrl(itr['url'], True)).group(1)
params = urlparse.parse_qs(urlparse.urlparse(url_fdf).query)
nsd, nsr = params['NumeroSequencialDocumento'][0], params['NumeroSequencialRegistroCvm'][0]
infos = dict()
infos['id'] = edp
infos['period'] = tm
infos['ptype'] = 'ITR'
# Get Infos From DFPs Consolidadas - Balanco Patrimonial Ativo
getDfpConBPA(nsd, nsr, infos)
# Get Infos From DFPs Consolidadas - Balanco Patrimonial Passivo
getDfpConBPP(nsd, nsr, infos)
# Get Infos From DFPs Consolidadas - Demonstracao do Resultado
getDftDemRes(nsd, nsr, infos, PARMS_GERAL_ITR)
# Get Infos From Dados da Empresa - Composicao do Capital
getDaeCc(nsd, nsr, infos)
yield infos
def getDetails(enterprise):
logging.info('Getting Details of Company %s', enterprise[0])
edp,name = enterprise[0],enterprise[1]
d = dict()
d['id'] = edp
d['name'] = name
d['register_date'] = datetime.now().isoformat()
page = openUrl(buildUrl(URL_EDETAIL, 'CodCVM=' + edp))
# Company Data
att = getTables(page, 'panel1a')
for table in att[:3]:
for row in getRows(table): rowToDict(row,d)
#print tostring(page)
att = HtmlElement(page).get_element_by_id('panel1a')
d['trading_codes'] = [ l.text for l in HtmlElement(att).find_class('LinkCodNeg') ]
att = d['industry_classification'].split('/')
d['sector'] = att[0].strip()
d['subsector'] = att[1].strip()
d['segment'] = att[2].strip()
d['detailwebsite'] = buildUrl(URL_RESUMO, 'codigoCvm=' + edp, 'idioma=pt-BR')
if 'website' in d: d['website'] = urlparse.urlsplit(d['website'], 'http').geturl()
return d
def getDfpInfos(edp):
logging.info('Getting DFP Infos for %s', edp)
# Getting newer version of documents
dfps = dict([(i['date'], i) for i in sorted(getDFPs(edp), cmp=mycmp) if i['version'] > 0.0])
for year, dfp in dfps.iteritems():
logging.info('Getting DFP of %s for %s', edp, year.year)
# Get NSD and NSR
url_fdf = RE_LOC.search(openUrl(dfp['url'], True)).group(1)
params = urlparse.parse_qs(urlparse.urlparse(url_fdf).query)
nsd, nsr = params['NumeroSequencialDocumento'][0], params['NumeroSequencialRegistroCvm'][0]
infos = dict()
infos['id'] = edp
infos['period'] = str(year.year)
infos['ptype'] = 'DFP'
# Get Infos From DFPs Consolidadas - Balanco Patrimonial Ativo
getDfpConBPA(nsd, nsr, infos)
# Get Infos From DFPs Consolidadas - Balanco Patrimonial Passivo
getDfpConBPP(nsd, nsr, infos)
# Get Infos From DFPs Consolidadas - Demonstracao do Resultado
getDftDemRes(nsd, nsr, infos)
# Get Infos From Dados da Empresa - Composicao do Capital
getDaeCc(nsd, nsr, infos)
yield infos
def updateEntries(cls, feed):
xml = utils.openUrl(feed.url)
if not xml:
return
dom = utils.parseXmlString(xml)
parser = parsers.FeedParserFactory.create(dom)
pagingKey = 0
for entryDict in parser.entries():
key = entryDict['key']
entry = EntryModel.get_by_key_name(key, parent=feed)
if not entry:
entry = EntryModel(parent=feed, key_name=key)
entry.feed = feed
feed.total += 1
if entry.fromDict(entryDict):
entry.read = False
entry.setPagingKey(pagingKey)
entry.put()
pagingKey += 1
feed.unread += 1
if pagingKey > 0:
feed.put()
def downloadBdiFile(dt):
fileurl = '{0}bdi{1}.zip'.format(URL_BDI, dt.strftime('%m%d'))
logging.info('Downloading file %s', fileurl)
# save file
fp = StringIO(openUrl(fileurl, True))
z = zipfile.ZipFile(fp)
for name in z.namelist():
yield z.open(name)
def msnMoneyTenYearSummary(symbol, local=False):
import bs4
import utils
if not len(symbol):
return None
url = "http://investing.money.msn.com/investments/financial-statements?symbol=" + symbol
url, page = utils.openUrl(url, local)
print url
soup = bs4.BeautifulSoup(page, "lxml")
# Extract tables
tables = soup.find_all("table", {"class": "mnytbl"})
# Parse income statement table
income = {}
for row in tables[0].find_all("tr"):
cols = row.find_all("td")
if len(cols) == 0:
continue
tuple = ()
for icol, col in enumerate(cols):
entry = col.find_all(text=True)[1].strip()
if icol == 0:
try:
secs = utils.makeEpochTime(str(entry), "%m/%y")
tuple = tuple + (secs,)
except ValueError:
tuple = tuple + (str(entry),)
else:
tuple = utils.extractData(entry, tuple)
income[tuple[0]] = tuple[1:]
# Parse balance sheet table
balance = {}
for row in tables[1].find_all("tr"):
cols = row.find_all("td")
if len(cols) == 0:
continue
tuple = ()
for icol, col in enumerate(cols):
entry = col.find_all(text=True)[1].strip()
if icol == 0:
try:
secs = utils.makeEpochTime(str(entry), "%m/%y")
tuple = tuple + (secs,)
except ValueError:
tuple = tuple + (str(entry),)
else:
tuple = utils.extractData(entry, tuple)
balance[tuple[0]] = tuple[1:]
return income, balance
def getDaeCc(nsd, nsr, infos):
logging.info('Getting Dados da Empresa - Composicao do Capital - %s %s', nsd, nsr)
page = HtmlElement(openUrl(buildUrl(URL_DCC, PARMS_GERAL, PARMS_BPP, PARMS_BPP,
'NumeroSequencialDocumento=' + nsd, 'NumeroSequencialRegistroCvm=' + nsr)))
# Multiplicador
em = page.xpath('.//div[@id="UltimaTabela"]/table/tr/td/b/text()')
m = 1000 if len(em) > 0 and em[0].find('(Mil)') != -1 else 1
for i in ['QtdAordCapiItgz', 'QtdAprfCapiItgz', 'QtdTotAcaoCapiItgz', 'QtdAordTeso', 'QtdAprfTeso', 'QtdTotAcaoTeso']:
qnt = page.get_element_by_id('ctl00_cphPopUp_{0}_1'.format(i))
infos[i] = 0 if qnt is None else toInt(qnt.text) * m
def getDftDemRes(url, infos):
logging.info('Getting DFs Consolidadas - Demonstracao do Resultado')
t = dict()
page = openUrl(url)
table = HtmlElement(page).get_element_by_id(TABLE_BPP)
for row in getRows(table): rowToDict(row, t)
# Multiplicador
m = getMultiplicador(page)
# Receita Liquida
infos['RL'] = toInt(t['3.01'][1])*m if '3.01' in t else 0
# Lucro Liquito
infos['LL'] = next((toInt(v[1])*m for k,v in t.iteritems() if re.match('^Lucro.+odo$', v[0])), 0)
def getDfpConBPA(url, infos):
logging.info('Getting DFs Consolidadas - Balanco Patrimonial Ativo')
t = dict()
page = HtmlElement(openUrl(url))
table = page.get_element_by_id(TABLE_BPP)
for row in getRows(table): rowToDict(row, t)
# Multiplicador
m = getMultiplicador(page)
# Caixa
cxa = toInt(t['1.01.01'][1])*m if '1.01.01' in t else 0
apf = toInt(t['1.01.02'][1])*m if '1.01.02' in t else 0
infos['CAIXA'] = cxa + apf
def getDfpConBPA(nsd, nsr, infos):
logging.info('Getting DFs Consolidadas - Balanco Patrimonial Ativo - %s %s', nsd, nsr)
t = dict()
page = HtmlElement(openUrl(buildUrl(URL_FDF, PARMS_GERAL, PARMS_BPP,
'NumeroSequencialDocumento=' + nsd, 'NumeroSequencialRegistroCvm=' + nsr,
'Informacao=2', 'Demonstracao=2')))
table = page.get_element_by_id(TABLE_BPP)
for row in getRows(table): rowToDict(row, t)
# Multiplicador
m = getMultiplicador(page)
# Caixa
cxa = toInt(t['1.01.01'][1])*m if '1.01.01' in t else 0
apf = toInt(t['1.01.02'][1])*m if '1.01.02' in t else 0
infos['CAIXA'] = cxa + apf
def getDftDemRes(nsd, nsr, infos, params_geral=PARMS_GERAL):
logging.info('Getting DFs Consolidadas - Demonstracao do Resultado - %s %s', nsd, nsr)
t = dict()
page = openUrl(buildUrl(URL_FDF, params_geral, PARMS_BPP,
'NumeroSequencialDocumento=' + nsd, 'NumeroSequencialRegistroCvm=' + nsr,
'Informacao=2', 'Demonstracao=4'))
table = HtmlElement(page).get_element_by_id(TABLE_BPP)
for row in getRows(table): rowToDict(row, t)
# Multiplicador
m = getMultiplicador(page)
# Receita Liquida
infos['RL'] = toInt(t['3.01'][1])*m if '3.01' in t else 0
# Lucro Liquito
infos['LL'] = toInt(t['3.11'][1])*m if '3.11' in t else 0
def msnMoneyQuote( symbol, local = False ) :
import bs4
import utils
if not len(symbol) : return None
url = 'http://investing.money.msn.com/investments/stock-price?symbol=' + symbol
url,page = utils.openUrl(url,local)
print url
soup = bs4.BeautifulSoup(page,"lxml")
# Extract date stamp from below "details" table
footers = soup.find_all("span",{"class":"foot"})
string = footers[0].find_all(text=True)[0].strip().split(' ')[2]
date = utils.makeEpochTime(string,'%m/%d/%Y')
# Extract tables
tables = soup.find_all("table",{"class":"mnytbl"})
# Parse "details" table
details = {}
tuple = ()
cntr = 0
for row in tables[0].find_all("tr") :
cells = row.find_all("td")
if len(cells) == 0 : continue
data = cells[1].find_all(text=True)[1].strip()
tuple = utils.extractData(data,tuple)
cntr = cntr + 1
details[date] = tuple
# Parse "financial highlights" table
highlights = {}
tuple = ()
cntr = 0
for row in tables[1].find_all("tr") :
cells = row.find_all("td")
if len(cells) == 0 : continue
index = 2 if ( cntr == 2 or cntr == 3 ) else 1
data = cells[1].find_all(text=True)[index].strip()
tuple = utils.extractData(data,tuple)
cntr = cntr + 1
highlights[date] = tuple
return details,highlights
def getDfpConBPP(nsd, nsr, infos):
logging.info('Getting DFs Consolidadas - Balanco Patrimonial Passivo - %s %s', nsd, nsr)
t = dict()
page = openUrl(buildUrl(URL_FDF, PARMS_GERAL, PARMS_BPP,
'NumeroSequencialDocumento=' + nsd, 'NumeroSequencialRegistroCvm=' + nsr,
'Informacao=2', 'Demonstracao=3'))
table = HtmlElement(page).get_element_by_id(TABLE_BPP)
for row in getRows(table): rowToDict(row, t)
# Multiplicador
m = getMultiplicador(page)
# Patrimonio Liquido
infos['PL'] = toInt(t['2.03'][1])*m if '2.03' in t else 0
# Divida Bruta
CP = toInt(t['2.01.04'][1])*m if '2.01.04' in t else 0
LP = toInt(t['2.02.01'][1])*m if '2.02.01' in t else 0
infos['DB'] = CP + LP
def getDfpConBPP(url, infos):
logging.info('Getting DFs Consolidadas - Balanco Patrimonial Passivo')
t = dict()
page = openUrl(url)
table = HtmlElement(page).get_element_by_id(TABLE_BPP)
for row in getRows(table): rowToDict(row, t)
# Multiplicador
m = getMultiplicador(page)
# Patrimonio Liquido
infos['PL'] = toInt(t['2.03'][1])*m if '2.03' in t else 0
# Divida Bruta
CP = toInt(t['2.01.04'][1])*m if '2.01.04' in t else 0
LP = toInt(t['2.02.01'][1])*m if '2.02.01' in t else 0
infos['DB'] = CP + LP
# Dividendos
Div = toInt(t['2.01.05.02.01'][1])*m if '2.01.05.02.01' in t else 0
infos['DIV'] = Div
def msnMoneyBalanceSheet( symbol, local = False ) :
import bs4
import utils
if not len(symbol) : return None
url = 'http://investing.money.msn.com/investments/stock-balance-sheet/?symbol=' + symbol + '&stmtView=Ann'
url,page = utils.openUrl(url,local)
print url
soup = bs4.BeautifulSoup(page,"lxml")
rows = soup.find_all("tr")
ncols = len(rows[-1].find_all("td"))-1
titles = []
tuples = [() for x in range(ncols)]
for irow,row in enumerate( rows ) :
for icol,col in enumerate( row.find_all("td") ) :
entries = col.find_all(text=True)
index = None
if len(entries) == 1 : index = 0
elif len(entries) == 3 : index = 1
elif len(entries) == 7 : index = 4
else : continue
entry = entries[index].strip().encode("utf-8")
if irow == 7 or irow == 30 : continue
if len(entry) :
if icol == 0 :
titles.append(str(entry))
else :
dates = {1:'%Y',2:'%m/%d/%Y',4:'%m/%d/%Y'}
try :
date = dates[irow]
secs = utils.makeEpochTime(str(entry),date)
tuples[icol-1] = tuples[icol-1] + (secs,)
except KeyError :
tuples[icol-1] = utils.extractData(entry,tuples[icol-1])
dict = {}
for col in range(len(tuples)) :
dict[tuples[col][0]] = tuples[col][1:]
return titles,dict
def getLinks(url):
resp = dict()
trp = openUrl(url, True)
url_fdf = RE_LOC.search(trp).group(1)
params = urlparse.parse_qs(urlparse.urlparse(url_fdf).query)
resp['ctd'] = params['CodTipoDocumento'][0]
resp['nsd'] = params['NumeroSequencialDocumento'][0]
resp['nsr'] = params['NumeroSequencialRegistroCvm'][0]
resp['cti'] = params['CodigoTipoInstituicao'][0]
fparams = '&CodTipoDocumento={0}&NumeroSequencialDocumento={1}&'\
'NumeroSequencialRegistroCvm={2}&CodigoTipoInstituicao={3}'.\
format(resp['ctd'], resp['nsd'], resp['nsr'], resp['cti'])
for i in re.findall('"Text":"([^"]+)","Value":"([^"]+)"', trp):
if re.match('^Balan.+Ativo$', i[0]): resp['dfp_bpa'] = escapeUrl(URL_RAD + i[1] + fparams)
if re.match('^Balan.+Passivo$', i[0]): resp['dfp_bpp'] = escapeUrl(URL_RAD + i[1] + fparams)
if re.match('^Demonstra.+Resultado$', i[0]): resp['dfp_dr'] = escapeUrl(URL_RAD + i[1] + fparams)
return resp
def msnMoneyCompanyProfile( symbol, local = False ) :
import bs4
import utils
import datetime
if not len(symbol) : return None
url = 'http://investing.money.msn.com/investments/company-report?symbol=' + symbol
url,page = utils.openUrl(url,local)
print url
soup = bs4.BeautifulSoup(page,"lxml")
tables = soup.find_all("table",{"class":"mnytbl"})
text = tables[1].find_all("span")[0].find_all(text=True)[0].strip().encode("utf-8")
d = datetime.date(1900,1,1)
length = {0:len(text)}
profile = {0:text}
return length,profile
def getFdFiles(edp, dftype):
logging.info('Getting {0} Infos for {1}'.format(dftype, edp))
# Getting newer version of documents
fps = dict([(i['date'], i) for i in sorted(getFPs(edp, dftype), cmp=mycmp) if i['version'] > 0.0])
for year, fp in fps.iteritems():
logging.info('Getting {0} of {1} for {2}'.format(dftype, edp, year.year))
params = urlparse.parse_qs(urlparse.urlparse(fp['url']).query)
query = 'NumeroSequencialDocumento={0}&CodigoInstituicao={1}'.format(params['NumeroSequencialDocumento'][0], params['CodigoTipoInstituicao'][0])
filepath = path.join(getPath(dftype), str(edp), str(year.year))
filename = '{0}/{1}_{2}-{3}.zip'.format(filepath, params['NumeroSequencialDocumento'][0], params['CodigoTipoInstituicao'][0], fp['version'])
if path.isfile(filename):
logging.info('File {0} exists, skiping...'.format(filename))
yield filename
continue
if not path.isdir(filepath):
makedirs(filepath)
logging.info('Downloading file {0}'.format(filename))
with open(filename, 'wb+') as f:
f.write(openUrl(buildUrl(URL_DD, query), True))
yield filename
def getFPs(edp, dftype):
page = openUrl(buildUrl(URL_DF, 'codigoCVM=' + edp, 'idioma=pt-br', 'tipo='+dftype.lower()))
return [parseFd(dfp) for dfp in page.findall('.//div[@id="' + DIV_DFPS + '"]/div/div/div/div/p/a')]
def msnMoneyHistoricalPrices(symbol, local=False):
import bs4
import utils
if not len(symbol):
return None
url = (
"http://investing.money.msn.com/investments/equity-historical-price/?PT=7&D4=1&DD=1&D5=0&DCS=2&MA0=0&MA1=0&CF=0&nocookie=1&SZ=0&symbol="
+ symbol
)
# url = 'http://investing.money.msn.com/investments/equity-historical-price/?symbol=us%3a' + symbol + '&CA=0&CB=0&CC=0&CD=0&D4=1&DD=1&D5=0&DCS=2&MA0=0&MA1=0&C5=0&C5D=0&C6=0&C7=0&C7D=0&C8=0&C9=0&CF=4&D8=0&DB=1&DC=1&D9=0&DA=0&D1=0&SZ=0&PT=11'
url, page = utils.openUrl(url, local)
print url
soup = bs4.BeautifulSoup(page, "lxml")
rows = soup.find_all("tr")
titles = []
prices = {}
dividends = {}
for irow, row in enumerate(rows):
cols = row.find_all("td")
# Extract titles from table header
headers = row.find_all("th")
for header in headers:
entries = header.find_all(text=True)
entry = entries[1].strip()
if not len(entry):
continue
titles.append(str(entry))
# Extract ex-dividend dates, dividends paid, and share price
if len(cols) == 3:
date = 0
div = 0.0
price = 0.0
try:
entries = cols[0].find_all(text=True)
entry = entries[1].strip()
if len(entry):
date = utils.makeEpochTime(str(entry), "%m/%d/%Y")
except ValueError:
date = 0
try:
entries = cols[1].find_all(text=True)
entry = entries[1].strip().split(" ")[0]
if len(entry):
div = float(entry)
except ValueError:
div = 0.0
try:
if irow < len(rows):
entries = rows[irow + 1].find_all("td")[4].find_all(text=True)
entry = entries[1].strip()
price = float(entry)
else:
price = 0.0
except ValueError:
price = 0.0
if date != 0:
dividends[date] = (div, price)
# Loop through rows and extract share prices
else:
tuple = ()
if len(cols) != 6:
continue
for icol, col in enumerate(cols):
entries = col.find_all(text=True)
entry = entries[1].strip()
if not len(entry):
continue
try:
secs = utils.makeEpochTime(str(entry), "%m/%d/%Y")
tuple = tuple + (secs,)
except ValueError:
tuple = utils.extractData(entry, tuple)
prices[tuple[0]] = tuple[1:]
return titles, prices, dividends
def getITRs(edp):
page = openUrl(buildUrl(URL_DF, 'codigoCVM=' + edp, 'idioma=pt-br', 'tipo=itr'))
return [parseFd(itr) for itr in page.findall('.//div[@id="' + DIV_DFPS + '"]/div/div/div/div/p/a')]
