def parse_record(self, line):
if line[0] == "Kontonummer":
# it's the table header
return None
if len(line) < 3:
"""e.g.: ['# 1 vorgemerkte Umsätze nicht angezeigt']"""
return None
if not line[2]:
return None
if self.statement.account_id is None:
self.statement.account_id = line[0]
sl = StatementLine()
sl.id = line[1]
sl.date = self.parse_datetime(line[2])
sl.date_avail = self.parse_datetime(line[3])
sl.amount = self.parse_float(line[4])
sl.trntype = TMAPPINGS.get(line[5],
'DEBIT' if sl.amount < 0 else 'CREDIT')
sl.payee = line[7][:32]
sl.memo = "%s: %s" % (line[6],
" ".join(x for x in line[13:31] if len(x) > 0))
if len(line[8]) > 0 and len(line[9]) > 0:
# additional bank information is present
splitted = self.parse_iban(line[8])
if splitted:
sl.bank_account_to = BankAccount(**splitted)
else:
sl.bank_account_to = BankAccount(line[9], line[8])
return sl
def parse_record(self, line):
# FIXME: add header validation
if self.cur_record < 2:
return None
if len(line) < 3:
"""e.g.: ['# 1 vorgemerkte Umsätze nicht angezeigt']"""
return None
if not line[2]:
return None
sl = StatementLine()
sl.id = line[1]
sl.date = self.parse_datetime(line[2])
sl.amount = self.parse_float(line[4])
sl.trntype = 'DEBIT' if sl.amount < 0 else 'CREDIT'
sl.payee = line[7]
# check for special transactions
if line[6] == "Entgeltabschluss":
sl.memo = "%s: %s %s" % (line[6], line[13], line[14])
elif line[6] == "Wertpapiere" or line[7] == "KREDITKARTENABRECHNUNG":
sl.memo = "(%s/%s): %s" % (line[8], line[9], " ".join(line[15:]).strip())
elif not line[8] and not line[9]:
# empty transaction
print("empty", line)
return None
else:
sl.memo = "(%s/%s): %s" % (line[8], line[9], " ".join(e for e in line[13:] if e).strip())
return sl
def parse_record(self, rec):
MultilineRecRE = re.compile(r"\n\s+")
SplitItemRE = re.compile(r":\s+")
# join together multiline items
rec = MultilineRecRE.sub(" ", rec)
res_iter = (SplitItemRE.split(line) for line in rec.split("\n"))
res_dict = dict((item[0].strip(), item[1].strip())
for item in res_iter)
res_key_set = set(res_dict.keys())
# Insurance against unknown keys in the statement
assert res_key_set <= self._keys, \
"Unknown keys in the transaction: %s" % (res_key_set - self._keys)
# no idea, how to make this unique and monotonically increasing
trans_ID = res_dict['datum zaúčtování'] + res_dict['částka'] + \
res_dict['variabilní symbol'] + res_dict['protiúčet']
date_str = res_dict['datum zaúčtování'].split('.')
stat_line = StatementLine(id=trans_ID,
date=datetime.date(int(date_str[2]), int(date_str[1]),
int(date_str[0])),
memo="%s\nozn. operace: %s" % (res_dict['poznámka'],
res_dict['označení operace']),
amount=float(res_dict['částka']))
stat_line.payee = res_dict['název protiúčtu']
# According to OFX spec this could be "check (or other reference) no."
# I don't see any requirement on monotonicity or uniqueness, but
# I wonder how GnuCash understands that
stat_line.check_no = res_dict['variabilní symbol']
return stat_line
def testAllFilled(self):
"""All of the fields used in the hash have a value"""
sline = StatementLine()
sline.date = date(2015, 10, 2)
sline.memo = "Test statement line"
sline.amount = 123.45
self.assertEqual(self.all_hash, generate_stable_transaction_id(sline))
def parse_record(self, line):
san_line = {}
for lkey, lval in line.items():
san_line[lkey] = re.sub("\$", "", lval)
stmt_line = StatementLine()
for field, col in self.mappings.items():
rawvalue = san_line[col]
try:
value = self.parse_value(rawvalue, field)
except ValueError as e:
# Invalid data line; skip it
warning("Error parsing value '{}' on line '{}': {}".format(rawvalue, san_line, e))
return None
setattr(stmt_line, field, value)
if self.statement.filter_zeros and is_zero(stmt_line.amount):
return None
try:
stmt_line.end_balance = float(san_line['Ending Balance'])
except ValueError:
# Usually indicates a pending transaction
return None
# generate transaction id out of available data
stmt_line.id = generate_stable_transaction_id(stmt_line)
return stmt_line
def parse_record(self, row):
"""Parse given table row and return StatementLine object
"""
# if we're still in header, no transactions yet, just collect account data
if self.in_header:
m = match(
r"""(Prometi\ za\ razdoblje\ od\ (?P<start>[0-9.]+)\ do\ (?P<end>[0-9.]+))
|(Račun:\ (?P<acct>\w+)) # account ^- start&end dates
|(Valuta:\ (?P<curr>\w+)) # currency
|(?P<eoh>Datum) # end of headers""",
row[0].value, VERBOSE)
if m:
# start/end
if m['start']:
self.statement.start_date = datetime.strptime(
m['start'], self.header_date_format)
self.statement.end_date = datetime.strptime(
m['end'], self.header_date_format)
# account number
if m['acct']:
self.statement.account_id = m['acct']
# currency
elif m['curr']:
self.statement.currency = m['curr']
# end of headers
elif m['eoh']:
self.in_header = False
self.row_nr += 1
return None
# main body of transactions after headers
stmt_line = StatementLine(amount=0)
for field, col in self.mappings.items():
if col >= len(row):
raise ValueError(
"Cannot find column %s in a row of %s items " %
(col, len(row)))
cell = row[col]
value = self.parse_value(cell)
# calculate debits and credits to amount
if field == 'debit' and value != 0:
stmt_line.amount += value
stmt_line.trntype = 'DEBIT'
elif field == 'credit' and value != 0:
stmt_line.amount -= value
stmt_line.trntype = 'CREDIT'
else:
setattr(stmt_line, field, value)
# apply generated transaction id
stmt_line.id = self.gen_id(stmt_line)
# overwrite end balance with current balance
self.statement.end_balance = stmt_line.balance
# next row
self.row_nr += 1
return stmt_line
def parse_record(self, record):
P = self.P
sl = StatementLine()
sl.date = self.parse_datetime(record.findall(P('ValDt', 'Dt'))[0].text)
sl.amount = float(record.findall(P('Amt'))[0].text)
sl.trntype = 'DEBIT' if (record.findall(P('CdtDbtInd'))[0].text
== "DBIT") else 'CREDIT'
if sl.trntype == 'DEBIT' and sl.amount > 0:
sl.amount = -sl.amount
sl.payee = record.findall(
P('NtryDtls', 'TxDtls', 'RltdPties', 'Cdtr', 'Nm'))[0].text
sl.memo = ' '.join([
e.text
for e in record.findall(P('NtryDtls', 'TxDtls', 'RmtInf', 'Ustrd'))
])
# generate unique id
h = hashlib.sha256()
h.update(str(sl.date).encode('utf-8'))
h.update(str(sl.amount).encode('utf-8'))
h.update(str(sl.trntype).encode('utf-8'))
h.update(str(sl.payee).encode('utf-8'))
h.update(str(sl.memo).encode('utf-8'))
sl.id = h.hexdigest()
return sl
def parse_record(self, df_row):
"""Parse given transaction line and return StatementLine object
"""
stmt_line = StatementLine()
# date field
stmt_line.date = self.xls_date(
self.df['Data Contabile'][self.df_row_idx])
# amount field
stmt_line.amount = self.df['Importo'][self.df_row_idx]
# transaction type field
if(stmt_line.amount < 0):
stmt_line.trntype = "DEBIT"
else:
stmt_line.trntype = "CREDIT"
# memo field
stmt_line.memo = self.df['Causale / Descrizione'][self.df_row_idx]
if(pd.isnull(stmt_line.memo)):
stmt_line.memo = ''
# id field
stmt_line.id = generate_transaction_id(stmt_line)
#print(str(stmt_line))
return stmt_line
def to_statement_line(self, raw_records):
ret =[]
for record in raw_records:
date = dateparser.parse(record['Data'])
memo = record['Detalii tranzactie']
line = StatementLine(date=date, memo=memo)
if 'Credit' in record:
line.amount = D(record['Credit'].replace('.','').replace(',','.'))
line.trntype = 'CREDIT'
elif 'Debit' in record:
line.amount = D(record['Debit'].replace('.','').replace(',','.'))
line.trntype='DEBIT'
else:
raise ArgumentError
if line.trntype=='CREDIT':
r = re.compile('ordonator:?(?P<payee>.*)$', re.MULTILINE| re.IGNORECASE)
m = r.search( memo)
if m:
d = m.groupdict()
line.payee = d['payee']
line.trntype='XFER'
#r = re.compile('din contul:?(?P<payee>.*)$', re.MULTILINE| re.IGNORECASE)
#m = r.search( memo)
#if m:
# d = m.groupdict()
# line.payee = d['payee']
r = re.compile('referinta:?(?P<refnum>.*)$', re.MULTILINE| re.IGNORECASE)
m = r.search( memo)
if m:
d = m.groupdict()
line.refnum = line.check_no = d['refnum']
line.id= generate_transaction_id(line)
ret.append(line)
return ret
def parse_record(self, mov):
stat_line = StatementLine(None,
mov.data_contabile,
mov.descrizione_estesa,
Decimal(mov.accrediti) if mov.accrediti else
Decimal(mov.addebiti))
stat_line.id = generate_transaction_id(stat_line)
stat_line.date_user = mov.data_valuta
stat_line.trntype = IntesaSanPaoloXlsxParser._get_transaction_type(mov)
logging.debug(stat_line)
return stat_line
def parse_record(self, line):
if len(line) < 5:
return None
elif len(line) < 12:
# possibly meta information about the account
if "BLZ" in line[0]:
self.statement.bank_id = line[1]
elif "Konto" in line[0]:
self.statement.account_id = line[1]
return None
if line[9] == "Anfangssaldo":
self.statement.start_date = self.parse_datetime(line[0])
self.statement.start_balance = self.parse_float(line[11])
return None
elif line[9] == "Endsaldo":
self.statement.end_date = self.parse_datetime(line[0])
self.statement.end_balance = self.parse_float(line[11])
return None
elif line[0] == "Buchungstag":
# it's the table header
return None
sl = StatementLine()
sl.date = self.parse_datetime(line[0])
sl.date_avail = self.parse_datetime(line[1])
# Note: amount has no sign. We need to guess it later...
sl.amount = self.parse_float(line[11])
info = self.parse_transaction_info(line)
sl.amount *= info["sign"]
sl.trntype = info["ttype"]
if "iban" in info:
# additional bank information if present
sl.bank_account_to = BankAccount(**self.parse_iban(info["iban"]))
if line[10] != self.statement.currency:
# different currency is used
sl.currency = line[10]
# remove additional spaces in the payee
sl.payee = re.sub(" +", " ", line[3].replace("\n", " ").strip())[:32]
# remove additional spaces in the memo
sl.memo = re.sub(" +", " ", info["memo"].strip())
# we need to generate an ID because nothing is given
sl.id = generate_stable_transaction_id(sl)
return sl
def parse_record(self, line):
"""
Parse given transaction line and return StatementLine object
"""
sl = StatementLine()
sl.amount = Decimal(line['Betrag'].replace(',', '.'))
# TODO trntype could be improved using 'Buchungstext'
if sl.amount.is_signed():
sl.trntype = 'DEBIT'
else:
sl.trntype = 'CREDIT'
# .date: It is debatable whether to use 'Buchungstag' or 'Valutadatum'
sl.date = self.parse_datetime(line['Valutadatum'])
# .date_user is not contained in the original CSV
# .payee becomes OFX.NAME which becomes "Description" in gnuCash
# .memo becomes OFX.MEMO which becomes "Notes" in gnuCash
# When .payee is empty, GnuCash imports .memo to "Description" and
# keeps "Notes" empty
#
# OFX's <NAME> and <PAYEE> are distinct fields. But ofxstatement's
# .payee is translated to OFX's <NAME>
#
# According to the OFX spec (version 2.1.1):
# <NAME> Name of payee or description of transaction, A-32
# Note: Provide NAME or PAYEE, not both
# <MEMO> Extra information (not in <NAME>)
#
# I prefer to have a description in .payee because that's what it ends
# up being in gnuCash.
recipient = _truncate_str(line['Beguenstigter/Zahlungspflichtiger'])
if not recipient:
recipient = 'UNBEKANNT'
sl.payee = "{}; {}".format(
_truncate_str(line['Verwendungszweck']),
recipient,
)
sl.memo = "{}; IBAN: {}; BIC: {}".format(
line['Buchungstext'].strip(),
line['Kontonummer/IBAN'].strip(),
line['BIC (SWIFT-Code)'].strip(),
)
m = sha256()
m.update(str(sl.date).encode('utf-8'))
m.update(sl.payee.encode('utf-8'))
m.update(sl.memo.encode('utf-8'))
m.update(str(sl.amount).encode('utf-8'))
# Shorten the hash to the first 16 digits just to make it more
# manageable. It should still be enough.
sl.id = str(abs(int(m.hexdigest(), 16)))[:16]
return sl
def parse_record(self, row):
row = take(5, row)
stmt_line = StatementLine()
stmt_line.date = self.parse_datetime(row[0])
_ = self.parse_datetime(row[1]) # TODO: ???
stmt_line.refnum = row[2]
stmt_line.memo = row[3]
stmt_line.amount = row[4]
#
# Looks like SEB formats description for card transactions so it includes the actual purchase date
# within e.g. 'WIRSTRÖMS PU/14-12-31' and it means that description is 'WIRSTRÖMS PU' while the actual
# card operation is 2014-12-31.
#
# P.S. Wirströms Irish Pub is our favorite pub in Stockholm: http://www.wirstromspub.se
#
m = re.match('(.*)/([0-9]{2}-[0-9]{2}-[0-9]{2})$', stmt_line.memo)
if m:
card_memo, card_date = m.groups()
if self.brief:
stmt_line.memo = card_memo
stmt_line.date_user = datetime.strptime(card_date, '%y-%m-%d')
stmt_line.id = generate_transaction_id(stmt_line)
return stmt_line
def parse_record(self, row):
id_idx = self.valid_header.index("Transaction ID")
date_idx = self.valid_header.index("Date")
memo_idx = self.valid_header.index("Name")
refnum_idx = self.valid_header.index("Reference Txn ID")
amount_idx = self.valid_header.index("Gross")
payee_idx = self.valid_header.index("To Email Address")
title_idx = self.valid_header.index("Item Title")
stmt_line = StatementLine()
stmt_line.id = row[id_idx]
stmt_line.date = datetime.strptime(row[date_idx], self.date_format)
stmt_line.memo = row[memo_idx]
if self.analyze:
memo_parts = [row[memo_idx]]
payee = row[payee_idx]
if payee and (payee.lower() == '*****@*****.**'):
memo_parts.append(row[title_idx])
stmt_line.memo = ' / '.join(filter(bool, memo_parts))
stmt_line.refnum = row[refnum_idx]
stmt_line.amount = atof(row[amount_idx].replace(" ", ""), self.locale)
return stmt_line
def _parse_line(self, ntry):
sline = StatementLine()
crdeb = _find(ntry, 'CdtDbtInd').text
amtnode = _find(ntry, 'Amt')
amt = self._parse_amount(amtnode, self.statement.currency)
if crdeb == CD_DEBIT:
amt = -amt
payee = _find(ntry, 'NtryDtls/TxDtls/RltdPties/Cdtr/Nm')
else:
payee = _find(ntry, 'NtryDtls/TxDtls/RltdPties/Dbtr/Nm')
sline.payee = payee.text
sline.amount = amt
dt = _find(ntry, 'ValDt')
sline.date = self._parse_date(dt)
bookdt = _find(ntry, 'BookgDt')
sline.date_user = self._parse_date(bookdt)
svcref = _find(ntry, 'NtryDtls/TxDtls/Refs/AcctSvcrRef')
sline.refnum = svcref.text
rmtinf = _find(ntry, 'NtryDtls/TxDtls/RmtInf/Ustrd')
sline.memo = rmtinf.text
return sline
def parse_record(self, line):
"""Parse given transaction line and return StatementLine object
"""
#print(line)
sl = StatementLine()
sl = super(PostFinanceBankParser, self).parse_record(line)
if not line[3]:
sl.amount = self.parse_float(line[2])
sl.trntype = 'CREDIT'
else:
sl.amount = self.parse_float("-" + line[3])
sl.trntype = 'DEBIT'
sl.payee = sl.memo
if payeesplit: #TODO: enable via settings
found = False
for regex in payeefinder:
m = regex.match(sl.payee)
if m:
found = True
sl.payee = m.group(1)
#if not found:
# print ("Found No PAYEE!" + sl.memo)
return sl
def parse_record(self, row):
self.row_num += 1
line = StatementLine()
line.date = self.parse_datetime(row[0].value)
line.date_user = self.parse_datetime(row[1].value)
line.refnum = str(self.row_num)
line.memo = row[2].value
line.amount = row[3].value
line.trntype = self.get_type(line)
line.id = generate_transaction_id(line)
return line
def parse_record(self, line):
# Free Headerline
if self.cur_record <= 1:
return None
stmt_line = StatementLine()
stmt_line.date = self.parse_datetime(line[0].strip())
# Amount
paid_out = -self.parse_amount(line[2])
paid_in = self.parse_amount(line[3])
stmt_line.amount = paid_out or paid_in
reference = line[1].strip()
trntype = False
for prefix, transaction_type in TRANSACTION_TYPES.items():
if reference.startswith(prefix):
trntype = transaction_type
break
if not trntype:
trntype = 'POS' # Default: Debit card payment
# It's ... pretty ugly, but I see no other way to do this than parse
# the reference string because that's all the data we have.
stmt_line.trntype = trntype
if trntype == 'POS':
stmt_line.payee, stmt_line.memo = self.parse_payee_memo(reference)
elif reference.startswith('Cash at '):
stmt_line.payee, stmt_line.memo = self.parse_payee_memo(
reference[8:])
elif reference.startswith('To ') or reference.startswith('From '):
stmt_line.payee = self.parse_value(
reference[reference.find(' '):], 'payee'
)
else:
stmt_line.memo = self.parse_value(reference, 'memo')
# Notes (from Apr-2018)
if len(line) > 8 and line[8].strip():
if not stmt_line.memo:
stmt_line.memo = u''
elif len(stmt_line.memo.strip()) > 0:
stmt_line.memo += u' '
stmt_line.memo += u'({})'.format(line[8].strip())
return stmt_line
def _parse_line(self, ntry: ET.Element) -> Optional[StatementLine]:
sline = StatementLine()
crdeb = self._findstrict(ntry, "CdtDbtInd").text
amtnode = self._findstrict(ntry, "Amt")
amt_ccy = amtnode.get("Ccy")
if amt_ccy != self.statement.currency:
# We can't include amounts with incompatible currencies into the
# statement.
return None
amt = self._parse_amount(amtnode)
if crdeb == CD_DEBIT:
amt = -amt
payee = self._find(ntry, "NtryDtls/TxDtls/RltdPties/Cdtr/Nm")
else:
payee = self._find(ntry, "NtryDtls/TxDtls/RltdPties/Dbtr/Nm")
sline.payee = payee.text if payee is not None else None
sline.amount = amt
dt = self._find(ntry, "ValDt")
sline.date = self._parse_date(dt)
bookdt = self._find(ntry, "BookgDt")
sline.date_user = self._parse_date(bookdt)
svcref = self._find(ntry, "NtryDtls/TxDtls/Refs/AcctSvcrRef")
if svcref is None:
svcref = self._find(ntry, "AcctSvcrRef")
if svcref is not None:
sline.refnum = svcref.text
# Try to find memo from different possible locations
refinf = self._find(ntry, "NtryDtls/TxDtls/RmtInf/Strd/CdtrRefInf/Ref")
rmtinf = self._find(ntry, "NtryDtls/TxDtls/RmtInf/Ustrd")
addinf = self._find(ntry, "AddtlNtryInf")
if refinf is not None:
sline.memo = refinf.text
elif rmtinf is not None:
sline.memo = rmtinf.text
elif addinf is not None:
sline.memo = addinf.text
return sline
def parse_record(self, line):
# Namespace stuff
namespaces = {'ns': line.tag[1:].partition("}")[0]}
# Get all fields
type_code = line.find('ns:TypeCode', namespaces=namespaces).text
date = line.find('ns:BookDate', namespaces=namespaces).text
c_or_d = line.find('ns:CorD', namespaces=namespaces).text
amount = line.find('ns:AccAmt', namespaces=namespaces).text
id = line.find('ns:BankRef', namespaces=namespaces).text
note = line.find('ns:PmtInfo', namespaces=namespaces).text
# Payee name
payee_name = None
payee = line.find('ns:CPartySet', namespaces=namespaces)
if payee:
payee_account = payee.find('ns:AccHolder', namespaces=namespaces)
if payee_account:
payee_name = payee_account.find('ns:Name', namespaces=namespaces).text
# Create statement line
stmt_line = StatementLine(id, self.parse_datetime(date), note, self.parse_float(amount))
stmt_line.payee = payee_name
# Credit & Debit stuff
stmt_line.trntype = "DEP"
if c_or_d == 'D':
stmt_line.amount = -stmt_line.amount
stmt_line.trntype = "DEBIT"
# Various types
if type_code == 'CHOU':
stmt_line.trntype = "ATM"
elif type_code == 'MEMD':
stmt_line.trntype = "SRVCHG"
elif type_code == 'OUTP':
stmt_line.trntype = "PAYMENT"
elif type_code == 'INP':
stmt_line.trntype = "XFER"
# # Check if paid by card
# m = CARD_PURCHASE_RE.match(stmt_line.memo)
# if m:
# # this is an electronic purchase. extract some useful
# # information from memo field
# date = m.group(1).split('/')
# date = '%s-%s-%s' % (date[2], date[1], date[0])
# stmt_line.date_user = self.parse_datetime(date)
print(stmt_line, stmt_line.trntype)
return stmt_line
def parse(self):
statement = Statement(bank_id='321081669', currency='USD')
with open(self.filename) as f:
for row in csv.DictReader(f):
line = StatementLine(id=row['Transaction Number'],
date=self.parse_datetime(row['Date']),
memo=row['Statement Description'],
amount=Decimal(row['Debit']
or row['Credit']))
line.payee = row['Description']
line.check_no = row['Check Number']
line.trntype = self.guess_type(line.payee, line.amount)
statement.lines.append(line)
return statement
def parse_record(self, line):
"""Extracts the transaction data from the given line parsed from the JSON file.
:param line: A transaction line from the parsed JSON file.
:return: A new StatementLine filled by the given line's data.
"""
amount = to_float(line['amounts']['amount'])
if self.statement.filter_zeros and is_zero(amount):
return None
memo = line['description']
datetime = ts_to_datetime(line['times']['when_recorded'])
id = line['uuid']
ledger_amount = convert_debit(amount, line['bookkeeping_type'])
stmt_line = StatementLine(id=id, date=datetime, memo=memo, amount=ledger_amount)
stmt_line.end_balance = to_float(line['running_balance'])
return stmt_line
def parse_record(self, line):
# Namespace stuff
namespaces = {'ns': line.tag[1:].partition("}")[0]}
# Get all fields
type_code = line.find('ns:TypeCode', namespaces=namespaces).text
date = line.find('ns:BookDate', namespaces=namespaces).text
c_or_d = line.find('ns:CorD', namespaces=namespaces).text
amount = line.find('ns:AccAmt', namespaces=namespaces).text
id = line.find('ns:BankRef', namespaces=namespaces).text
note = line.find('ns:PmtInfo', namespaces=namespaces).text
# Payee name
payee_name = None
payee = line.find('ns:CPartySet', namespaces=namespaces)
if payee:
payee_account = payee.find('ns:AccHolder', namespaces=namespaces)
if payee_account:
payee_name = payee_account.find('ns:Name', namespaces=namespaces).text
# Create statement line
stmt_line = StatementLine(id, self.parse_datetime(date), note, self.parse_float(amount))
stmt_line.payee = payee_name
# Credit & Debit stuff
stmt_line.trntype = "DEP"
if c_or_d == 'D':
stmt_line.amount = -stmt_line.amount
stmt_line.trntype = "DEBIT"
# Various types
if type_code == 'MEMD':
stmt_line.trntype = "SRVCHG"
elif type_code == 'OUTP':
stmt_line.trntype = "PAYMENT"
# Check if paid by card
m = CARD_PURCHASE_RE.match(stmt_line.memo)
if m:
# this is an electronic purchase. extract some useful
# information from memo field
date = m.group(1).split('/')
date = '%s-%s-%s' % (date[2], date[1], date[0])
stmt_line.date_user = self.parse_datetime(date)
# DEBUG
if self.debug:
print(stmt_line, stmt_line.trntype)
return stmt_line
def parse_record(self, line):
# Namespace stuff
namespaces = {'ns': line.tag[1:].partition("}")[0]}
# Get all fields
type_code = line.find('ns:TypeCode', namespaces=namespaces).text
date = line.find('ns:BookDate', namespaces=namespaces).text
c_or_d = line.find('ns:CorD', namespaces=namespaces).text
amount = line.find('ns:AccAmt', namespaces=namespaces).text
id = line.find('ns:BankRef', namespaces=namespaces).text
note = line.find('ns:PmtInfo', namespaces=namespaces).text
# Payee name
payee_name = None
payee = line.find('ns:CPartySet', namespaces=namespaces)
if payee:
payee_account = payee.find('ns:AccHolder', namespaces=namespaces)
if payee_account:
payee_name = payee_account.find('ns:Name', namespaces=namespaces).text
# Create statement line
stmt_line = StatementLine(id, self.parse_datetime(date), note, self.parse_float(amount))
stmt_line.payee = payee_name
# Credit & Debit stuff
stmt_line.trntype = "DEP"
if c_or_d == 'D':
stmt_line.amount = -stmt_line.amount
stmt_line.trntype = "DEBIT"
# Various types
if type_code == 'CHOU':
stmt_line.trntype = "ATM"
elif type_code == 'MEMD':
stmt_line.trntype = "SRVCHG"
elif type_code == 'OUTP':
stmt_line.trntype = "PAYMENT"
elif type_code == 'INP':
stmt_line.trntype = "XFER"
# DEBUG
if self.debug:
print(stmt_line, stmt_line.trntype)
return stmt_line
def _parse_line(self, ntry):
sline = StatementLine()
crdeb = _find(ntry, 'CdtDbtInd').text
amtnode = _find(ntry, 'Amt')
amt = self._parse_amount(amtnode, self.statement.currency)
if crdeb == CD_DEBIT:
amt = -amt
payee = _find(ntry, 'NtryDtls/TxDtls/RltdPties/Cdtr/Nm')
else:
payee = _find(ntry, 'NtryDtls/TxDtls/RltdPties/Dbtr/Nm')
sline.payee = payee.text
sline.amount = amt
dt = _find(ntry, 'ValDt')
sline.date = self._parse_date(dt)
bookdt = _find(ntry, 'BookgDt')
sline.date_user = self._parse_date(bookdt)
svcref = _find(ntry, 'NtryDtls/TxDtls/Refs/AcctSvcrRef')
sline.refnum = svcref.text
rmtinf = _find(ntry, 'NtryDtls/TxDtls/RmtInf/Ustrd')
sline.memo = rmtinf.text
return sline
def parse_record(self, line):
statement_line = StatementLine()
for field, col_number in self.mappings.items():
state = State(field, self.parse_value)
raw_value = self.get_nth_value(line, col_number)
value = raw_value.flatmap(state.execute)
setattr(statement_line, field, value)
return statement_line
def parse_record(self, line):
stmt_line = StatementLine()
for field, col in self.mappings.items():
if col >= len(line):
raise ValueError("Cannot find column %s in line of %s items " % (col, len(line)))
rawvalue = line[col]
value = self.parse_value(rawvalue, field)
setattr(stmt_line, field, value)
return stmt_line
def parse_record(self, row):
stmt_line = StatementLine()
stmt_line.date = self.parse_datetime(row[0])
stmt_line.date_user = self.parse_datetime(row[1])
stmt_line.memo = row[2]
stmt_line.amount = self.parse_float(row[3])
stmt_line.id = generate_transaction_id(stmt_line)
return stmt_line
def test_ofxWriter(self):
# Create sample statement:
statement = Statement("BID", "ACCID", "LTL")
statement.lines.append(StatementLine(
"1", datetime(2012, 2, 12), "Sample 1", 15.4))
line = StatementLine("2", datetime(2012, 2, 12), "Sample 2", 25.0)
line.payee = ''
line.bank_account_to = BankAccount("SNORAS", "LT1232")
line.bank_account_to.branch_id = "VNO"
statement.lines.append(line)
# Create writer:
writer = ofx.OfxWriter(statement)
# Set the generation time so it is always predictable
writer.genTime = datetime(2012, 3, 3, 0, 0, 0)
assert prettyPrint(writer.toxml()) == SIMPLE_OFX
def parse_record(self, line):
"""Extracts the transaction data from the given line parsed from the JSON file.
:param line: A transaction line from the parsed JSON file.
:return: A new StatementLine filled by the given line's data.
"""
amount = to_float(line['amounts']['amount'])
if self.statement.filter_zeros and is_zero(amount):
return None
memo = line['description']
datetime = ts_to_datetime(line['times']['when_recorded'])
id = line['uuid']
ledger_amount = convert_debit(amount, line['bookkeeping_type'])
stmt_line = StatementLine(id=id,
date=datetime,
memo=memo,
amount=ledger_amount)
stmt_line.end_balance = to_float(line['running_balance'])
return stmt_line
def _parse_line(self, ntry):
sline = StatementLine()
crdeb = self._find(ntry, 'CdtDbtInd').text
amtnode = self._find(ntry, 'Amt')
amt_ccy = amtnode.get('Ccy')
if amt_ccy != self.statement.currency:
# We can't include amounts with incompatible currencies into the
# statement.
return None
amt = self._parse_amount(amtnode)
if crdeb == CD_DEBIT:
amt = -amt
payee = self._find(ntry, 'NtryDtls/TxDtls/RltdPties/Cdtr/Nm')
else:
payee = self._find(ntry, 'NtryDtls/TxDtls/RltdPties/Dbtr/Nm')
sline.payee = payee.text if payee is not None else None
sline.amount = amt
dt = self._find(ntry, 'ValDt')
sline.date = self._parse_date(dt)
bookdt = self._find(ntry, 'BookgDt')
sline.date_user = self._parse_date(bookdt)
svcref = self._find(ntry, 'NtryDtls/TxDtls/Refs/AcctSvcrRef')
if svcref is None:
svcref = self._find(ntry, 'AcctSvcrRef')
if svcref is None:
svcref = self._find(ntry, 'NtryDtls/TxDtls/Refs/MsgId')
if svcref is not None:
sline.refnum = svcref.text
# Try to find memo from different possible locations
rmtinf = self._find(ntry, 'NtryDtls/TxDtls/RmtInf/Ustrd')
addinf = self._find(ntry, 'AddtlNtryInf')
if rmtinf is not None:
sline.memo = rmtinf.text
elif addinf is not None:
sline.memo = addinf.text
return sline
def try_cache(self, stmt_line: StatementLine) -> None:
account_id: str = str(self.statement.account_id)
key: TransactionKey
data: TransactionData
# The PDF may differ from the OFX by a different date
# or a switch from payee to memo or otherwise: forget them
for dt in [
stmt_line.accounting_date, stmt_line.operation_date,
stmt_line.value_date
]:
check_no: Optional[str]
name: Optional[str]
if stmt_line.payee == 'VIREMENT SEPA' and not (stmt_line.check_no):
check_no = None
name = stmt_line.memo
else:
check_no = stmt_line.check_no
name = None
key = Parser.transaction_key(account_id, check_no, dt,
stmt_line.amount, name)
if key in self.cache:
data = self.cache[key]
logger.debug('Found data %r for key %r', data, key)
stmt_line.date = dt
stmt_line.id = data.id
stmt_line.payee = data.name
stmt_line.memo = data.memo
return
else:
logger.debug('Did not find value for key %r', key)
if not self.cache_printed:
self.print_cache('try_cache: no data found')
self.cache_printed = True
return
def parse_record(self, line):
"""Parse given transaction line and return StatementLine object
"""
stmt_line = StatementLine()
# date field
stmt_line.date = self.xls_date(int(line[0]))
# amount field
if line[2]:
income = line[2]
outcome = 0
elif line[3]:
outcome = line[3]
income = 0
stmt_line.amount = income - outcome
# transaction type field
if (stmt_line.amount < 0):
stmt_line.trntype = "DEBIT"
else:
stmt_line.trntype = "CREDIT"
# name field
# set <NAME> field with content of column 'Descrizione'
# only if proper option is active
if self.info2name:
stmt_line.payee = line[4]
# memo field
stmt_line.memo = line[5]
# concat "Descrizione" column at the end of <MEMO> field
# if proper option is present
if self.info2memo:
if stmt_line.memo != '' and line[2] != '':
stmt_line.memo += ' - '
stmt_line.memo += line[2]
# id field
stmt_line.id = generate_transaction_id(stmt_line)
#print(str(stmt_line))
return stmt_line
def parse_record(self, line):
# Namespace stuff
namespaces = {"ns": line.tag[1:].partition("}")[0]}
# Get all fields
type_code = line.find("ns:TypeCode", namespaces=namespaces).text
date = line.find("ns:BookDate", namespaces=namespaces).text
c_or_d = line.find("ns:CorD", namespaces=namespaces).text
amount = line.find("ns:AccAmt", namespaces=namespaces).text
id = line.find("ns:BankRef", namespaces=namespaces).text
note = line.find("ns:PmtInfo", namespaces=namespaces).text
# Payee name
payee_name = None
payee = line.find("ns:CPartySet", namespaces=namespaces)
if payee:
payee_account = payee.find("ns:AccHolder", namespaces=namespaces)
if payee_account:
payee_name = payee_account.find("ns:Name", namespaces=namespaces).text
# Create statement line
stmt_line = StatementLine(id, self.parse_datetime(date), note, self.parse_float(amount))
stmt_line.payee = payee_name
# Credit & Debit stuff
stmt_line.trntype = "DEP"
if c_or_d == "D":
stmt_line.amount = -stmt_line.amount
stmt_line.trntype = "DEBIT"
# Various types
if type_code == "MEMD":
stmt_line.trntype = "SRVCHG"
elif type_code == "OUTP":
stmt_line.trntype = "PAYMENT"
# Check if paid by card
m = CARD_PURCHASE_RE.match(stmt_line.memo)
if m:
# this is an electronic purchase. extract some useful
# information from memo field
date = m.group(1).split("/")
date = "%s-%s-%s" % (date[2], date[1], date[0])
stmt_line.date_user = self.parse_datetime(date)
# print(stmt_line)
return stmt_line
def parse_record(self, row):
self.row_num += 1
line = StatementLine()
line.date = self.parse_datetime(row[0].value)
line.date_user = self.parse_datetime(row[1].value)
line.refnum = str(self.row_num)
line.memo = row[2].value
line.amount = row[3].value
line.trntype = self.get_type(line)
if self.statement.start_balance is None and self.row_num == 1:
self.statement.start_balance = row[4].value - line.amount
self.statement.start_date = line.date_user
self.statement.end_balance = row[4].value
line.id = self.generate_transaction_id(line)
if line.id in self.seen:
log.warn(
"Transaction with duplicate FITID generated:\n%s\n%s\n\n" %
(line, self.seen[line.id]))
else:
self.seen[line.id] = line
return line
def parse_record(self, line):
"""Parse given transaction line and return StatementLine object
"""
#print(line)
sl = StatementLine()
sl = super(PostFinanceCreditParser, self).parse_record(line)
#print(line)
if not line[3]:
sl.amount = self.parse_float(line[4])
sl.trntype = 'CREDIT'
else:
sl.amount = self.parse_float("-" + line[3])
sl.trntype = 'DEBIT'
sl.payee = sl.memo
return sl
def parse_record(self, stmt_line: StatementLine) -> StatementLine:
"""Parse given transaction line and return StatementLine object
"""
def add_years(d: date, years: int) -> date:
"""Return a date that's `years` years after the date
object `d`. Return the same calendar date (month and day) in the
destination year, if it exists, otherwise use the following day
(thus changing February 29 to March 1).
"""
return d.replace(year=d.year + years, month=3, day=1) \
if d.month == 2 and d.day == 29 \
else d.replace(year=d.year + years)
def get_date(d_m: str) -> date:
assert self.statement.end_date
# Without a year it will be 1900 so add the year
d_m_y: str = "{}/{}".format(d_m, self.statement.end_date.year)
d: date = datetime.strptime(d_m_y, '%d/%m/%Y').date()
if d > self.statement.end_date:
d = add_years(d, -1)
assert d <= self.statement.end_date
return d
logger.debug('Statement line: %r', stmt_line)
# Remove zero-value notifications
if stmt_line.amount == 0: # pragma: no cover
return None
stmt_line.accounting_date = get_date(stmt_line.accounting_date)
stmt_line.operation_date = get_date(stmt_line.operation_date)
stmt_line.value_date = get_date(stmt_line.value_date)
self.try_cache(stmt_line)
if not stmt_line.id:
stmt_line.date = stmt_line.accounting_date
account_id = self.statement.account_id
stmt_line.id = \
generate_unique_transaction_id(stmt_line,
self.unique_id_sets[account_id])
m = re.match(r'([0-9a-f]+)(-\d+)?$', stmt_line.id)
assert m, "Id should match hexadecimal digits, \
optionally followed by a minus and a counter: '{}'".format(stmt_line.id)
if m.group(2):
counter = int(m.group(2)[1:])
# include counter so the memo gets unique
stmt_line.memo = stmt_line.memo + ' #' + str(counter + 1)
return stmt_line
def _parse_line(self, ntry):
sline = StatementLine()
crdeb = _find(ntry, 'CdtDbtInd').text
amtnode = _find(ntry, 'Amt')
amt = self._parse_amount(amtnode)
if crdeb == CD_DEBIT:
amt = -amt
payee = _find(ntry, 'NtryDtls/TxDtls/RltdPties/Cdtr/Nm')
else:
payee = _find(ntry, 'NtryDtls/TxDtls/RltdPties/Dbtr/Nm')
if payee is not None:
payee = payee.text
sline.payee = payee
sline.amount = amt
dt = _find(ntry, 'ValDt')
sline.date = self._parse_date(dt)
bookdt = _find(ntry, 'BookgDt')
sline.date_user = self._parse_date(bookdt)
svcref = _find(ntry, 'NtryDtls/TxDtls/Refs/AcctSvcrRef')
sline.refnum = getattr(svcref, 'text', None)
rmtinf = _find(ntry, 'NtryDtls/TxDtls/RmtInf/Ustrd')
sline.memo = rmtinf.text if rmtinf.text else ''
addtlinf_node = _find(ntry, 'NtryDtls/TxDtls/AddtlTxInf')
addtlinf = self._parse_addtlinf(addtlinf_node)
if 'VÁSÁRLÁS KÁRTYÁVAL' == addtlinf and not sline.payee:
sline.payee = _trim_payee(sline.memo)
sline.memo += ' ' + addtlinf
return sline
def test_ofxWriter(self) -> None:
# Create sample statement:
statement = Statement("BID", "ACCID", "LTL")
statement.lines.append(
StatementLine("1", datetime(2012, 2, 12), "Sample 1",
Decimal("15.4")))
line = StatementLine("2", datetime(2012, 2, 12), "Sample 2",
Decimal("25.0"))
line.payee = ""
line.bank_account_to = BankAccount("SNORAS", "LT1232")
line.bank_account_to.branch_id = "VNO"
line.currency = Currency("USD")
line.orig_currency = Currency("EUR", Decimal("3.4543"))
statement.lines.append(line)
# Create writer:
writer = ofx.OfxWriter(statement)
# Set the generation time so it is always predictable
writer.genTime = datetime(2012, 3, 3, 0, 0, 0)
assert prettyPrint(writer.toxml()) == SIMPLE_OFX
def parse_record(self, transaction: Transaction) -> StatementLine:
"""Parse given transaction line and return StatementLine object
"""
logger.debug('transaction:\n' + pformat(transaction, indent=4))
stmt_line = None
# Use str() to prevent rounding errors
bank_account_to = transaction.data.get('customer_reference')
amount = Decimal(str(transaction.data['amount'].amount))
memo = transaction.data['transaction_details']
memo = memo.replace("\n", '')
memo = memo.replace(transaction.data['customer_reference'], '', 1)
memo = memo.replace(transaction.data['extra_details'], '', 1).strip()
memo = memo if memo != '' else 'UNKNOWN'
payee = None
if transaction.data['customer_reference'] != ''\
and transaction.data['extra_details'] != '':
payee = "{1} ({0})".format(transaction.data['customer_reference'],
transaction.data['extra_details'])
date = transaction.data['date']
# Remove zero-value notifications
if amount != 0:
stmt_line = StatementLine(date=date, memo=memo, amount=amount)
stmt_line.id = \
generate_unique_transaction_id(stmt_line, self.unique_id_set)
m = re.match(r'([0-9a-f]+)(-\d+)?$', stmt_line.id)
assert m, "Id should match hexadecimal digits, \
optionally followed by a minus and a counter: '{}'".format(stmt_line.id)
if m.group(2):
counter = int(m.group(2)[1:])
# include counter so the memo gets unique
stmt_line.memo = stmt_line.memo + ' #' + str(counter + 1)
stmt_line.payee = payee
if bank_account_to:
stmt_line.bank_account_to = \
BankAccount(bank_id=None,
acct_id=bank_account_to)
return stmt_line
def parse(self):
DATA_DATE = 'data_date'
DATA_AMT = 'data_amt'
DATA_DSC = 'data_dsc'
DATA_MEMO = 'data_mem'
DATA_BAL = 'data_bal'
def parser_old(filename):
def parse_tr(tr):
def get_header(tr, n):
return str(
tr.find_all(headers='header' + str(n))[0].string)
hdrs = tr.find_all(headers=True)
assert len(hdrs) in [0, 7]
if len(hdrs) == 0:
cont = True
date_or_comm = str(tr.find_all(colspan='5')[0].string)
desc = None
paym = None
depo = None
bal = None
else:
cont = False
date_or_comm = dateutil.parser.parse(get_header(tr, 1),
dayfirst=True)
desc = get_header(tr, 2)
paym = self.get_float(get_header(tr, 5))
depo = self.get_float(get_header(tr, 6))
assert (depo is None) != (
paym is None
), 'Either depo or paym need to exist, but not both: ' + str(
tr)
bal = self.get_float(get_header(tr, 7))
return cont, date_or_comm, desc, paym, depo, bal
with open(filename, 'r', encoding='iso-8859-8') as f:
soup = BeautifulSoup(f, 'lxml')
statement = Statement(currency='ILS')
data = []
for tr in soup.find_all(
'table',
id='mytable_body')[0].find_all(id='TR_ROW_BANKTABLE'):
cont, date_or_comm, desc, paym, depo, bal = parse_tr(tr)
if cont:
data[-1][DATA_MEMO] = date_or_comm
else:
new_line = {}
new_line[DATA_DATE] = date_or_comm
new_line[DATA_DSC] = desc
new_line[
DATA_AMT] = -paym if paym is not None else depo
new_line[DATA_BAL] = bal
new_line[DATA_MEMO] = None
data.append(new_line)
return data
def parser_new(filename):
with open(filename, 'r', encoding='iso-8859-8') as f:
bs = BeautifulSoup(f, 'lxml')
trs = bs.find_all('table', i__d='trBlueOnWhite12')[0].find_all(
'tr', recursive=False)
data = []
for tr in trs[1:]:
new_line = {}
tds = tr.find_all('td')
if len(tds) != 7:
continue
new_line[DATA_DATE] = dateutil.parser.parse(str(tds[0].string),
dayfirst=True)
new_line[DATA_DSC] = str(tds[1].string)
paym = self.get_float(str(tds[4].string))
depo = self.get_float(str(tds[5].string))
assert (depo is None) != (
paym is None
), 'Either depo or paym need to exist, but not both: ' + str(
tds)
new_line[DATA_AMT] = -paym if paym is not None else depo
new_line[DATA_BAL] = self.get_float(str(tds[6].string))
new_line[DATA_MEMO] = None
data.append(new_line)
return data
def parser_xslx(filename):
with open(filename, 'rb') as f:
ws = openpyxl.load_workbook(f).worksheets[0]
iterrows = iter(ws.rows)
for _ in range(6):
next(iterrows)
data = []
for row in iterrows:
new_line = {}
new_line[DATA_DATE] = dateutil.parser.parse(str(row[0].value),
dayfirst=False)
new_line[DATA_DSC] = str(row[1].value)
paym = self.get_float(str(row[3].value))
depo = self.get_float(str(row[4].value))
assert (depo is None) != (
paym is
None), 'Either depo or paym need to exist, but not both: '
new_line[DATA_AMT] = -paym if paym is not None else depo
new_line[DATA_BAL] = self.get_float(str(row[5].value))
new_line[DATA_MEMO] = " ".join(
[str(c.value) for c in row[7:] if c.value is not None])
data.append(new_line)
return data
PARSERS = [parser_old, parser_new, parser_xslx]
v = self.detect_version()
assert v is not None, "Unsupported file %s" % self.filename
self.log("Detected file of version %d" % v)
data = PARSERS[v](self.filename)
self.log('Found %d transactions' % len(data))
stmnt = Statement(currency='ILS')
for d in data:
stmt_line = StatementLine(date=d[DATA_DATE], amount=d[DATA_AMT])
#stmt_line.end_balance = d[DATA_BAL] #TODO: conf
stmt_line.payee = d[DATA_DSC]
if d[DATA_MEMO] is not None:
stmt_line.memo = d[DATA_MEMO]
if stmt_line.payee.startswith('משיכה'):
stmt_line.trntype = 'ATM'
elif stmt_line.payee.startswith('שיק'):
stmt_line.trntype = 'CHECK'
else:
stmt_line.trntype = "CASH" if d[DATA_AMT] < 0 else "DEP"
stmt_line.assert_valid()
stmnt.lines.append(stmt_line)
return stmnt
def parse_record(self, line):
if self.cur_record < 2:
return None
m = None
parse_info = None
if len(line) == 11:
m = self.mt940_mappings
parse_info = self.parse_transaction_info_mt940
elif len(line) == 17:
m = self.camt_mappings
parse_info = self.parse_transaction_info_camt
else:
raise ValueError("invalid input line: '%s'" % line)
if self.statement.account_id is None:
self.statement.account_id = line[m["accid"]]
sl = StatementLine()
sl.date_avail = self.parse_datetime(line[m["valdate"]])
if len(line[m["date"]]) > 0:
sl.date = self.parse_datetime(line[m["date"]])
else
sl.date = sl.date_avail
sl.amount = self.parse_float(line[m["amount"]])
sl.trntype = self.parse_transaction_type(sl.amount,
line[m["btext"]])
# remove leading or all) zeros
line[m["toaccid"]] = line[m["toaccid"]].lstrip('0')
if len(line[m["toaccid"]]) > 0 and len(line[m["tobankid"]]) > 0:
# additional bank information is present
splitted = self.parse_iban(line[m["toaccid"]])
if splitted:
sl.bank_account_to = BankAccount(**splitted)
else:
sl.bank_account_to = BankAccount(line[m["tobankid"]],
line[m["toaccid"]])
if line[m["currency"]] != self.statement.currency:
# different currency is used
sl.currency = line[m["currency"]]
# remove additional spaces in the payee
sl.payee = re.sub(' +', ' ', line[m["payee"]])[:32]
info = parse_info(line)
# remove additional spaces in the memo
sl.memo = "%s: %s" % (line[m["btext"]],
re.sub(' +', ' ', info["memo"].strip()))
# we need to generate an ID because nothing is given
sl.id = generate_stable_transaction_id(sl)
return sl
def testAmount(self):
"""Only amount is set"""
sline = StatementLine()
sline.amount = 123.45
self.assertEqual(self.amt_hash, generate_stable_transaction_id(sline))
def parse(self):
DATA_DATE = 'data_date'
DATA_AMT = 'data_amt'
DATA_DSC = 'data_dsc'
DATA_MEMO = 'data_mem'
DATA_BAL = 'data_bal'
def parser_old(filename):
def parse_tr(tr):
def get_header(tr, n):
return str(tr.find_all(headers='header'+str(n))[0].string)
hdrs = tr.find_all(headers=True)
assert len(hdrs) in [0, 7]
if len(hdrs) == 0:
cont = True
date_or_comm = str(tr.find_all(colspan='5')[0].string)
desc = None
paym = None
depo = None
bal = None
else:
cont = False
date_or_comm = dateutil.parser.parse(get_header(tr, 1), dayfirst=True)
desc = get_header(tr, 2)
paym = self.get_float(get_header(tr, 5))
depo = self.get_float(get_header(tr, 6))
assert (depo is None) != (paym is None), 'Either depo or paym need to exist, but not both: ' + str(tr)
bal = self.get_float(get_header(tr, 7))
return cont, date_or_comm, desc, paym, depo, bal
with open(filename, 'r', encoding='iso-8859-8') as f:
soup = BeautifulSoup(f, 'lxml')
statement = Statement(currency='ILS')
data = []
for tr in soup.find_all('table', id='mytable_body')[0].find_all(id='TR_ROW_BANKTABLE'):
cont, date_or_comm, desc, paym, depo, bal = parse_tr(tr)
if cont:
data[-1][DATA_MEMO] = date_or_comm
else:
new_line = {}
new_line[DATA_DATE] = date_or_comm
new_line[DATA_DSC] = desc
new_line[DATA_AMT] = -paym if paym is not None else depo
new_line[DATA_BAL] = bal
new_line[DATA_MEMO] = None
data.append(new_line)
return data
def parser_new(filename):
with open(filename, 'r', encoding='iso-8859-8') as f:
bs = BeautifulSoup(f, 'lxml')
trs = bs.find_all('table', i__d='trBlueOnWhite12')[0].find_all('tr', recursive=False)
data = []
for tr in trs[1:]:
new_line = {}
tds = tr.find_all('td')
if len(tds) != 7:
continue
new_line[DATA_DATE] = dateutil.parser.parse(str(tds[0].string), dayfirst=True)
new_line[DATA_DSC] = str(tds[1].string)
paym = self.get_float(str(tds[4].string))
depo = self.get_float(str(tds[5].string))
assert (depo is None) != (paym is None), 'Either depo or paym need to exist, but not both: ' + str(tds)
new_line[DATA_AMT] = -paym if paym is not None else depo
new_line[DATA_BAL] = self.get_float(str(tds[6].string))
new_line[DATA_MEMO] = None
data.append(new_line)
return data
def parser_xslx(filename):
with open(filename, 'rb') as f:
ws = openpyxl.load_workbook(f).worksheets[0]
iterrows = iter(ws.rows)
for _ in range(6):
next(iterrows)
data = []
for row in iterrows:
new_line = {}
new_line[DATA_DATE] = dateutil.parser.parse(str(row[0].value), dayfirst=False)
new_line[DATA_DSC] = str(row[1].value)
paym = self.get_float(str(row[3].value))
depo = self.get_float(str(row[4].value))
assert (depo is None) != (paym is None), 'Either depo or paym need to exist, but not both: '
new_line[DATA_AMT] = -paym if paym is not None else depo
new_line[DATA_BAL] = self.get_float(str(row[5].value))
new_line[DATA_MEMO] = " ".join([str(c.value) for c in row[7:] if c.value is not None])
data.append(new_line)
return data
PARSERS = [parser_old, parser_new, parser_xslx]
v = self.detect_version()
assert v is not None, "Unsupported file %s" % self.filename
self.log("Detected file of version %d" % v)
data = PARSERS[v](self.filename)
self.log('Found %d transactions' % len(data))
stmnt = Statement(currency='ILS')
for d in data:
stmt_line = StatementLine(date=d[DATA_DATE], amount=d[DATA_AMT])
#stmt_line.end_balance = d[DATA_BAL] #TODO: conf
stmt_line.payee = d[DATA_DSC]
if d[DATA_MEMO] is not None:
stmt_line.memo = d[DATA_MEMO]
if stmt_line.payee.startswith('משיכה'):
stmt_line.trntype = 'ATM'
elif stmt_line.payee.startswith('שיק'):
stmt_line.trntype = 'CHECK'
else:
stmt_line.trntype = "CASH" if d[DATA_AMT] < 0 else "DEP"
stmt_line.assert_valid()
stmnt.lines.append(stmt_line)
return stmnt
