Benjamin_Loison/BNP_PDF_statement_parser
1
0
Fork 0
Code Issues 2 Pull Requests Packages Projects Releases Wiki Activity

#2: Use pdftotext -layout instead of pdftotext -raw to distinguish debit from credit

Browse Source
This commit is contained in:
Benjamin Loison 2024-10-04 00:53:33 +02:00
parent 69d6442966
commit 2d4a10da05
Signed by: Benjamin_Loison
SSH Key Fingerprint: SHA256:BtnEgYTlHdOg1u+RmYcDE0mnfz1rhv5dSbQ2gyxW8B8
2 changed files with 28 additions and 20 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
bnp_pdf_statement_parser.py
View File

@ -6,6 +6,7 @@ import matplotlib.ticker as ticker
from datetime import datetime
from utils import getDatetimeFromFileName, getMonthIndexSinceEpoch, getMonthNameFromMonthIndex, readPdfBankStatement
import operator
from pprint import pprint
PATH = f'/home/benjamin/Desktop/bens_folder/bazaar/documents/bnp/bank_statements/'
@ -34,11 +35,12 @@ for folder in os.listdir():
filePath = f'{folder}/{file}'
print(filePath)
print(readPdfBankStatement(filePath))
exit(0)
#exit(0)
transactions = readPdfBankStatement(filePath)[3]
pprint(transactions)
appendTransactions(transactions, folder)
#exit(1)
exit(1)
os.chdir(f'{MAIN_BANK_ACCOUNT}/')
for folder in sorted(os.listdir()):
@ -53,7 +55,6 @@ for folder in sorted(os.listdir()):
#break
#break
from pprint import pprint
allTransactions.sort(key = operator.itemgetter('date'))
print(len(allTransactions))
pprint(allTransactions)

27
utils.py
View File

@ -2,17 +2,18 @@ import subprocess
from datetime import datetime
import re
FIRST_LINE_OF_PAYMENT_REGEX = re.compile('(\\d{2}\\.\\d{2}) (\\d{2}\\.\\d{2}) ([\\d ]+,\\d{2})')
# Source: [the Stack Overflow answer 766377](https://stackoverflow.com/a/766377)
FIRST_LINE_OF_PAYMENT_REGEX = re.compile('\\ +(\\d{2}\\.\\d{2})\\ +([A-Z\\d /.()*]+?)\\ +(\\d{2}\\.\\d{2})\\ +([\\d ]+,\\d{2})')
END_PAGE_AFTER_THE_FIRST_ONE_REGEX = re.compile('P\\. \\d+/\\d+')
SOLDE_CREDITEUR_AU_REGEX = re.compile('SOLDE CREDITEUR AU (\\d{2}\\.\\d{2}\\.\\d{4}) ([\\d ]+,\\d{2})')
TOTAL_DES_OPERATIONS_REGEX = re.compile('TOTAL\\ DES\\ OPERATIONS\\ ([\\d ]+,\\d{2})\\ ([\\d ]+,\\d{2})')
TOTAL_DES_OPERATIONS_CREDIT_ONLY_REGEX = re.compile('TOTAL\\ DES\\ OPERATIONS\\ ([\\d ]+,\\d{2})')
SOLDE_CREDITEUR_AU_REGEX = re.compile('\\ +SOLDE CREDITEUR AU (\\d{2}\\.\\d{2}\\.\\d{4})\\ +([\\d ]+,\\d{2})')
TOTAL_DES_OPERATIONS_REGEX = re.compile('\\ +TOTAL\\ DES\\ OPERATIONS\\ +([\\d ]+,\\d{2})\\ +([\\d ]+,\\d{2})')
TOTAL_DES_OPERATIONS_CREDIT_ONLY_REGEX = re.compile('\\ +TOTAL\\ DES\\ OPERATIONS\\ +([\\d ]+,\\d{2})')
def execute(command):
return subprocess.check_output(command).decode('utf-8')
def getTextFromPdf(pdfPath):
return execute(['pdftotext', '-raw', pdfPath, '-'])
return execute(['pdftotext', '-layout', pdfPath, '-'])
def getDatetimeFromFileName(aDatetimeStr):
#aDatetime = datetime.strptime(aDatetimeStr, '%Y%m%d.pdf')
@ -56,6 +57,7 @@ def readPdfBankStatement(filePath):
if soldeCrediteurAuRegexMatch is not None:
initialDate = datetime.strptime(soldeCrediteurAuRegexMatch.group(1), '%d.%m.%Y')
initialAmount = toFloat(soldeCrediteurAuRegexMatch.group(2))
print(f'{initialAmount=}')
#currentAmount = initialAmount
started = True
continue
@ -66,15 +68,19 @@ def readPdfBankStatement(filePath):
started = False
continue
# We aren't interested in the content after this line
elif line.startswith('TOTAL DES OPERATIONS'):
else:
totalDesOperationsRegexMatch = TOTAL_DES_OPERATIONS_REGEX.match(line)
totalDesOperationsCreditOnlyRegexMatch = TOTAL_DES_OPERATIONS_CREDIT_ONLY_REGEX.match(line)
if totalDesOperationsRegexMatch is not None or totalDesOperationsCreditOnlyRegexMatch is not None:
# Note that transfer between accounts will be noted in both debits and credits, as trying to cancel would make benefits show as negative debit which does not make sense.
# Cannot just consider January as benefits only as `20240122.pdf` also contains an additional transfer between my accounts.
if totalDesOperationsRegexMatch is not None:
totalMonthlyDebit, totalMonthlyCredit = [toFloat(group) for group in totalDesOperationsRegexMatch.groups()]
else:
totalMonthlyCredit = toFloat(TOTAL_DES_OPERATIONS_CREDIT_ONLY_REGEX.match(line).group(1))
totalMonthlyCredit = toFloat(totalDesOperationsCreditOnlyRegexMatch.group(1))
totalMonthlyDebit = 0
print(f'{totalMonthlyDebit=}')
print(f'{totalMonthlyCredit=}')
break
firstLineOfPaymentRegexMatch = FIRST_LINE_OF_PAYMENT_REGEX.match(line)
if firstLineOfPaymentRegexMatch is not None:
@ -88,11 +94,12 @@ def readPdfBankStatement(filePath):
'comment': '\n'.join(comment)
}]
date = None
date, valeur, amount = firstLineOfPaymentRegexMatch.groups()
date, firstCommentLine, valeur, amount = firstLineOfPaymentRegexMatch.groups()
amount = toFloat(amount)
#currentAmount -= amount
comment = []
else:
comment = [firstCommentLine]
elif line != '':
print(f'comment: {line}')
comment += [line]
if date is not None:
transactions += [{