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

Add account amount after each transaction

Browse Source
This commit is contained in:
Benjamin Loison 2023-06-21 01:05:48 +02:00
parent 5f439c0e61
commit 02625cc71b
Signed by: Benjamin_Loison
SSH Key Fingerprint: SHA256:BtnEgYTlHdOg1u+RmYcDE0mnfz1rhv5dSbQ2gyxW8B8
1 changed files with 17 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
bnp_pdf_statement_parser.py
View File

@ -34,21 +34,25 @@ for folder in os.listdir():
for file in os.listdir(folder):
#folder = '2022'
#file = '20220321.pdf'
print(folder, file)
filePath = f'{folder}/{file}'
print(filePath)
content = getTextFromPdf(filePath)
lines = content.splitlines()
started = False
firstPage = True
payment = []
initialAmount = None
currentAmount = None
date = None
comment = []
for line in lines:
if not started:
# We are interested in the content after this line:
if soldeCrediteurAuRegex.match(line) != None or (line.startswith('Date Nature des opérations Valeur Débit Crédit') and not firstPage):
if soldeCrediteurAuRegex.match(line):
initialAmount = float(soldeCrediteurAuRegex.sub('', line).replace(',', '.').replace(' ', ''))
print(initialAmount)
currentAmount = initialAmount
print('Initial amount', initialAmount)
print()
started = True
continue
else:
@ -61,8 +65,16 @@ for folder in os.listdir():
elif line.startswith('TOTAL DES OPERATIONS'):
break
if firstLineOfPaymentRegex.match(line) != None:
if date != None:
print(date, valeur, amount, currentAmount)
print("\n".join(comment))
print()
print(line)
date, valeur, amount = line.split()
amount = float(amount.replace(',', '.'))
currentAmount -= amount
comment = []
else:
comment += [line]
break
break