BNP_PDF_statement_parser/bnp_pdf_statement_parser.py

#!/usr/bin/env python

import os
import matplotlib.pyplot as plt
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/'

os.chdir(PATH)

PRINT_TRANSACTIONS = False

MAIN_BANK_ACCOUNT = 'compte_de_cheques'

allTransactions = []
monthlyTransactions = {}

def appendTransactions(transactions, bankAccount):
    global allTransactions
    for transaction in transactions:
        transaction['bank account'] = bankAccount
        allTransactions += [transaction]
        date = transaction['date'].replace(day = 1)
        monthlyTransactions[date] = monthlyTransactions.get(date, []) + [transaction]

for folder in os.listdir():
    if folder != MAIN_BANK_ACCOUNT:
        for file in os.listdir(folder):
            filePath = f'{folder}/{file}'
            print(filePath)
            transactions = readPdfBankStatement(filePath)[3]
            pprint(transactions)
            appendTransactions(transactions, folder)

os.chdir(f'{MAIN_BANK_ACCOUNT}/')

for folder in sorted(os.listdir()):
    for file in sorted(os.listdir(folder)):
        filePath = f'{folder}/{file}'
        print(filePath)
        transactions = readPdfBankStatement(filePath)[3]
        appendTransactions(transactions, MAIN_BANK_ACCOUNT)
        if PRINT_TRANSACTIONS:
            pprint(transactions)
        #break
    #break

allTransactions.sort(key = operator.itemgetter('date'))
print(len(allTransactions))
#pprint(allTransactions)

import re
VIRT_A_CPTE_EMIS_SUR_LE_REGEX = re.compile('VIRT CPTE A CPTE EMIS SUR LE\n(CEL|LEP|LVJ|L\.A|LDD)\\d{23}')

# Could precise bank account to restrict own account comments.
def isTransactionFromOwnAccounts(comment):
    #if comment.startswith('DEPOT INITIAL DU COMPTE\n'):
    #    print(comment)
    #if comment.startswith('VIR CPTE A CPTE EMIS /MOTIF '):
    #    print(comment)
    return comment.startswith('DEPOT INITIAL DU COMPTE\n') or \
           comment.startswith('VIR CPTE A CPTE EMIS /MOTIF ') or \
           VIRT_A_CPTE_EMIS_SUR_LE_REGEX.fullmatch(comment)#comment.startswith('VIRT CPTE A CPTE EMIS SUR LE')
           #comment.startswith('VIR CPTE A CPTE RECU /DE ') or \
           #  and comment.endswith('/REFDO /REFBEN')

#allTransactions = [transaction for transaction in allTransactions if not isTransactionFromOwnAccounts(transaction['comment'])]
sortedMonths = sorted(monthlyTransactions.keys())
for month in sortedMonths:
    monthlyTransactions[month] = [transaction for transaction in monthlyTransactions[month] if not isTransactionFromOwnAccounts(transaction['comment'])]
totalMonthlyDebits = [sum([min(transaction['amount'], 0) for transaction in monthlyTransactions[month]]) for month in sortedMonths]
totalMonthlyCredits = [sum([max(transaction['amount'], 0) for transaction in monthlyTransactions[month]]) for month in sortedMonths]
totalMonthlyDifferences = [sum([transaction['amount'] for transaction in monthlyTransactions[month]]) for month in sortedMonths]
totals = [monthlyTransactions[sortedMonths[0]][0]['current amount'] + sum(totalMonthlyDifferences[:monthIndex + 1]) for monthIndex in range(len(sortedMonths))]

fig, ax = plt.subplots()
plt.title('BNP accounts monthly debits and credits')
plt.xlabel('Date')
plt.ylabel('€')
ALPHA = 0.5

xTicks = range(getMonthIndexSinceEpoch(sortedMonths[0]), getMonthIndexSinceEpoch(sortedMonths[-1]) + 1)
totalMonthlyAmountAndLabel = (
    (totalMonthlyDebits, 'Debit'),
    (totalMonthlyCredits, 'Credit'),
    (totalMonthlyDifferences, 'Difference'),
    (totals, 'Total'),
)
for totalMonthlyAmount, totalMonthlyLabel in totalMonthlyAmountAndLabel:
    plt.bar(xTicks, totalMonthlyAmount, alpha = ALPHA, label = totalMonthlyLabel)
plt.legend()

#plt.yscale('symlog')
ax.yaxis.set_major_formatter(ticker.StrMethodFormatter('{x:,}'))

ticksLabels = [getMonthNameFromMonthIndex(monthIndex) for monthIndex in xTicks]
plt.xticks(xTicks, ticksLabels, rotation = 90)
#plt.tight_layout()
# How to show the horizontal lines for subticks?
plt.grid(axis = 'y')

plt.show()