add bmo parser, some refactors

parsers.py

@@ -1,12 +1,14 @@
 import re
-from abc import ABC
-from abc import abstractmethod
-from datetime import datetime
+from abc import ABC, abstractmethod
+from datetime import date, datetime
 from decimal import Decimal
-from email.message import EmailMessage
-from email.message import Message
+from email.message import EmailMessage, Message
+from logging import info
+from typing import Optional
 from uuid import UUID
 
+from bs4 import BeautifulSoup
+
 from model import Transaction
 
 
@@ -25,7 +27,7 @@ class TransactionParser(ABC):
         pass
 
     @abstractmethod
-    def extract(self, msg: EmailMessage) -> Transaction:
+    def extract(self, msg: EmailMessage) -> Optional[Transaction]:
         """
         Extracts transaction details from the given email message.
 
@@ -59,11 +61,11 @@ class RogersBankParser(TransactionParser):
             and msg["Subject"] == "Purchase amount alert"
         )
 
-    def extract(self, msg: EmailMessage) -> Transaction:
-        body = msg.get_body()
+    def extract(self, msg: EmailMessage) -> Optional[Transaction]:
+        body = msg.get_content()
         if body is None:
             raise TransactionParsingFailed("No body of message found")
-        matches = self.EXTRACT_RE.search(body.as_string())
+        matches = self.EXTRACT_RE.search(body)
         if matches is None:
             raise TransactionParsingFailed("No matches for extraction RE")
         amount = Decimal(matches[1])
@@ -88,27 +90,68 @@ class MBNAParser(TransactionParser):
     def __init__(self, account_id: UUID):
         self.account_id = account_id
 
-    def match(self, msg: Message):
+    def match(self, msg: Message) -> bool:
         return (
             msg["From"] == "MBNA Notifications <noreply@mbna.ca>"
             and msg["Subject"] == "MBNA - Transaction Alert"
         )
 
-    def extract(self, msg: EmailMessage) -> Transaction:
-        body = msg.get_body()
+    def extract(self, msg: EmailMessage) -> Optional[Transaction]:
+        body = msg.get_content()
         if body is None:
             raise TransactionParsingFailed("No body of message found")
-        matches = self.EXTRACT_RE.search(body.as_string())
+        matches = self.EXTRACT_RE.search(body)
         if matches is None:
             raise TransactionParsingFailed("No matches for extraction RE")
         amount = Decimal(matches[1])
         payee = matches[2]
-        date = matches[5]
+        date_raw = matches[5]
         return Transaction(
             account=self.account_id,
-            date=date,
+            date=date.fromisoformat(date_raw),
             amount=amount,
             payee=payee,
             notes="via email",
             imported_id=msg["Message-ID"],
         )
+
+
+class BMOParser(TransactionParser):
+    EXTRACT_RE = re.compile(
+        r"We want to let you know that a (withdrawal|deposit) of\s+\$(\d+\.\d{2})\s+has been made (?:to|from) your account ending\s+in\s+(\d{3})",  # noqa: E501
+        flags=re.MULTILINE,
+    )
+
+    def __init__(self, account_map: dict[int, UUID]):
+        self._account_map = account_map
+
+    def match(self, msg: Message) -> bool:
+        return msg["From"] == "bmoalerts@bmo.com"
+
+    def extract(self, msg: EmailMessage) -> Optional[Transaction]:
+        body = msg.get_content()
+        if body is None:
+            raise TransactionParsingFailed("No body of message found")
+        soup = BeautifulSoup(body, "html.parser")
+        matches = self.EXTRACT_RE.search(soup.get_text())
+        if matches is None:
+            raise TransactionParsingFailed("No matches for extraction RE")
+
+        amount = Decimal(matches[2])
+        if matches[1] == "withdrawal":
+            amount = amount * -1
+        date_raw = msg["Date"]
+        date = datetime.strptime(date_raw, "%a, %d %b %Y %H:%M:%S %z").date()
+        account_ref = int(matches[3])
+        if account_ref not in self._account_map:
+            info("Account %s not in account map", account_ref)
+            return None
+        account_id = self._account_map[account_ref]
+        return Transaction(
+            account=account_id,
+            date=date,
+            amount=amount,
+            payee="",
+            notes="via email",
+            imported_id=msg["Message-ID"],
+        )