Extract information from TradeRepublic statements PDF

tr_extract.py

#!/usr/bin/env python3
from pypdf import PdfReader
import re
import pprint

re_label = "Ausführung Handel Direktkauf Kauf"
re_date = r"\d{2}\.\d{2}\.\d{4}"
re_isin = r"[A-Z]{2}[A-Z0-9]{9}[0-9]"
re_amount = r"-?\d{1,3}(?:\.\d{3})*(?:,\d{2})"

line_pattern = re.compile(
    rf"^({re_date}){re_label}\s({re_isin})\s.*?\s?({re_amount})?$"
)

amount_pattern = re.compile(rf"({re_amount})$")

reader = PdfReader("stmt2.pdf")

full_text = ""
for page in reader.pages:
    full_text += page.extract_text() + "\n"

lines = full_text.split("\n")

records = {}
for index, line in enumerate(lines):
    match = line_pattern.search(line)
    if match:
        date = match.group(1)
        isin = match.group(2)
        amount = 0
        if match.group(3):
            amount = match.group(3)
        else:  # Amount is on the next line
            match = amount_pattern.search(lines[index + 1])
            if match:
                amount = match.group(1)
        if isin not in records:
            records[isin] = []

        records[isin].append({"date": date, "amount": amount})

pprint.pprint(records)