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

Specify compte_de_cheques/

Browse Source
This commit is contained in:
Benjamin Loison 2024-10-03 17:36:07 +02:00
parent f090374caf
commit ee6e850174
Signed by: Benjamin_Loison
SSH Key Fingerprint: SHA256:BtnEgYTlHdOg1u+RmYcDE0mnfz1rhv5dSbQ2gyxW8B8
1 changed files with 8 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
bnp_pdf_statement_parser.py
View File

@ -9,7 +9,7 @@ import matplotlib.pyplot as plt
import matplotlib.ticker as ticker import matplotlib.ticker as ticker
from datetime import datetime from datetime import datetime
path = f'/home/benjamin/Desktop/bens_folder/bazaar/documents/bnp/bank_statements/' path = f'/home/benjamin/Desktop/bens_folder/bazaar/documents/bnp/bank_statements/compte_de_cheques/'
os.chdir(path) os.chdir(path)
@ -45,12 +45,11 @@ firstDatetime = None
lastDatetime = None lastDatetime = None
for folder in sorted(os.listdir()): for folder in sorted(os.listdir()):
for file in sorted(os.listdir(folder)): for file in sorted(os.listdir(folder)):
#folder = '2022'
#file = '20220321.pdf'
filePath = f'{folder}/{file}' filePath = f'{folder}/{file}'
print(filePath) print(filePath)
currentDatetime = getDatetime(file)
if firstDatetime is None: if firstDatetime is None:
firstDatetime = file firstDatetime = currentDatetime
content = getTextFromPdf(filePath) content = getTextFromPdf(filePath)
lines = content.splitlines() lines = content.splitlines()
started = False started = False
@ -86,7 +85,6 @@ for folder in sorted(os.listdir()):
totalMonthlyDebits += [totalMonthlyDebit] totalMonthlyDebits += [totalMonthlyDebit]
totalMonthlyCredits += [totalMonthlyCredit] totalMonthlyCredits += [totalMonthlyCredit]
totalMonthlyDifference = totalMonthlyCredit - totalMonthlyDebit totalMonthlyDifference = totalMonthlyCredit - totalMonthlyDebit
#print(totalMonthlyDifference)
totalMonthlyDifferences += [totalMonthlyDifference] totalMonthlyDifferences += [totalMonthlyDifference]
break break
if firstLineOfPaymentRegex.match(line) is not None: if firstLineOfPaymentRegex.match(line) is not None:
@ -102,7 +100,7 @@ for folder in sorted(os.listdir()):
comment += [line] comment += [line]
#break #break
#break #break
lastDatetime = file lastDatetime = getDatetime(file)
fig, ax = plt.subplots() fig, ax = plt.subplots()
plt.title('Monthly debits and credits') plt.title('Monthly debits and credits')
@ -110,15 +108,14 @@ plt.xlabel('Date')
plt.ylabel('€') plt.ylabel('€')
ALPHA = 0.5 ALPHA = 0.5
def getMonthIndex(aDatetimeStr): def getDatetime(aDatetimeStr):
aDatetime = datetime.strptime(aDatetimeStr, '%Y%m%d.pdf') return datetime.strptime(aDatetimeStr, '%Y%m%d.pdf')
def getMonthIndex(aDatetime):
return aDatetime.year * 12 + aDatetime.month return aDatetime.year * 12 + aDatetime.month
xTicks = range(getMonthIndex(firstDatetime), getMonthIndex(lastDatetime) + 1) xTicks = range(getMonthIndex(firstDatetime), getMonthIndex(lastDatetime) + 1)
# sign does not seem respected for `totalMonthlyDifferences`. # sign does not seem respected for `totalMonthlyDifferences`.
#print(totalMonthlyDifferences)
#print(f'{min(totalMonthlyDifferences)}')
#print(f'{max(totalMonthlyDifferences)}')
totalMonthlyAmountAndLabel = ( totalMonthlyAmountAndLabel = (
#(totalMonthlyDebits, 'Debit'), #(totalMonthlyDebits, 'Debit'),
#(totalMonthlyCredits, 'Credit'), #(totalMonthlyCredits, 'Credit'),
@ -129,7 +126,6 @@ for totalMonthlyAmount, totalMonthlyLabel in totalMonthlyAmountAndLabel:
plt.bar(xTicks, totalMonthlyAmount, alpha = ALPHA, label = totalMonthlyLabel) plt.bar(xTicks, totalMonthlyAmount, alpha = ALPHA, label = totalMonthlyLabel)
plt.legend() plt.legend()
# Hides negative
#plt.yscale('symlog') #plt.yscale('symlog')
ax.yaxis.set_major_formatter(ticker.StrMethodFormatter('{x:,}')) ax.yaxis.set_major_formatter(ticker.StrMethodFormatter('{x:,}'))