add neo parser

update node & actual

.pre-commit-config.yaml

@@ -1,14 +1,14 @@
 # .pre-commit-config.yaml
 repos:
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.1.5  # Check for the latest version
+    rev: v0.15.12  # Check for the latest version
     hooks:
       - id: ruff
         args: ["check", "--select", "I", "--fix"]
       - id: ruff-format
 
   - repo: https://github.com/pre-commit/pre-commit-hooks
-    rev: v4.5.0  # Check for the latest version
+    rev: v6.0.0  # Check for the latest version
     hooks:
       - id: check-yaml
       - id: end-of-file-fixer
@@ -16,6 +16,6 @@ repos:
       - id: requirements-txt-fixer
 
   - repo: https://github.com/PyCQA/flake8
-    rev: 6.0.0  # Check for the latest version
+    rev: 7.3.0  # Check for the latest version
     hooks:
       - id: flake8

Dockerfile

@@ -5,7 +5,7 @@ RUN apk add py3-uv
 RUN uv build
 
 # Use an official Node.js runtime as a base
-FROM node:20-alpine AS runtime
+FROM node:25-alpine AS runtime
 
 # Create app directory
 WORKDIR /app

actual_imap_poll/model.py

@@ -1,5 +1,5 @@
+import datetime
 from dataclasses import dataclass
-from datetime import date
 from decimal import Decimal
 from uuid import UUID
 
@@ -12,7 +12,7 @@ class Transaction:
     """
 
     account: UUID
-    date: date
+    date: datetime.date
     amount: Decimal  # Note: decimal dollars, JS shim will convert to cents as described in the API
     payee: str  # imported_payee in API
     notes: str

actual_imap_poll/parsers.py

@@ -1,10 +1,11 @@
 import re
 from abc import ABC, abstractmethod
+from collections.abc import Sequence
 from datetime import date, datetime
 from decimal import Decimal
 from email.message import EmailMessage, Message
-from logging import info, warning
-from typing import Any, Optional, Sequence
+from logging import debug, info, warning
+from typing import Any, Optional
 from uuid import UUID
 
 from bs4 import BeautifulSoup
@@ -58,6 +59,18 @@ class TransactionParser(ABC):
             raise TransactionParsingFailed("No content of message found")
         return content
 
+    @staticmethod
+    def strip_html(msg: EmailMessage) -> str:
+        body = msg.get_body(preferencelist=("html", "plain"))
+        if body is None:
+            raise TransactionParsingFailed("No HTML body of message found")
+        content = body.get_content()
+        if content is None:
+            raise TransactionParsingFailed("No content of message found")
+
+        soup = BeautifulSoup(content, "html.parser")
+        return soup.get_text()
+
 
 class TransactionParsingFailed(Exception):
     pass
@@ -65,7 +78,7 @@ class TransactionParsingFailed(Exception):
 
 class RogersBankParser(TransactionParser):
     EXTRACT_RE = re.compile(
-        r"Attempt of \$([0-9,]+\.\d{2}) was made on ([A-z]{3} \d{1,2}, \d{4})[^<]*at ([^<]+) in ([^<]+)."  # noqa: E501
+        r"\$([0-9,]+\.\d{2}) on ((Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec) [0-9]{2}, [0-9]{4}) on your credit card ending in ([0-9]{4}). Details: ([^.]+)"  # noqa: E501
     )
 
     def __init__(self, account_id: UUID):
@@ -78,17 +91,16 @@ class RogersBankParser(TransactionParser):
         )
 
     def extract(self, msg: EmailMessage) -> Optional[Transaction]:
-        content = self.get_content(msg)
+        content = self.strip_html(msg)
         matches = self.EXTRACT_RE.search(content)
         if matches is None:
             raise TransactionParsingFailed("No matches for extraction RE")
         amount = Decimal(matches[1].replace(",", "")) * -1
 
         date_raw = matches[2]
-        payee = matches[3]
-        location = matches[4]
+        payee = matches[5]
 
-        if "Rebate" == location and "CashBack" in payee:
+        if "CashBack" in payee:
             amount = amount * -1
 
         date = datetime.strptime(date_raw, "%b %d, %Y").date()
@@ -97,7 +109,7 @@ class RogersBankParser(TransactionParser):
             date=date,
             amount=amount,
             payee=payee,
-            notes=f"in {location} (via email)",
+            notes="(via email)",
             imported_id=msg["Message-ID"],
         )
 
@@ -136,8 +148,8 @@ class MBNAParser(TransactionParser):
 
 class BMOParser(TransactionParser):
     EXTRACT_RE = re.compile(
-        r"We want to let you know that a (withdrawal|deposit) of\s+\$([0-9,]+\.\d{2})\s+has been made (?:to|from) your account ending\s+in\s+(\d{3})",  # noqa: E501
-        flags=re.MULTILINE,
+        r"There was a (withdrawal|deposit).*Amount:\s*\$([0-9,]+\.\d{2}).*Account:\s*Ending in ([0-9]+)",  # noqa: E501
+        flags=re.DOTALL,
     )
 
     def __init__(self, account_map: dict[int, UUID]):
@@ -147,9 +159,8 @@ class BMOParser(TransactionParser):
         return msg["From"] == "bmoalerts@bmo.com"
 
     def extract(self, msg: EmailMessage) -> Optional[Transaction]:
-        content = self.get_content(msg)
-        soup = BeautifulSoup(content, "html.parser")
-        matches = self.EXTRACT_RE.search(soup.get_text())
+        content = self.strip_html(msg)
+        matches = self.EXTRACT_RE.search(content)
         if matches is None:
             raise TransactionParsingFailed("No matches for extraction RE")
 
@@ -298,3 +309,42 @@ class ScotiaBankParser(TransactionParser):
                 return self.extract_transaction(msg)
         info("Subject `%s` didn't match any extractors", msg["Subject"])
         return None
+
+
+class NeoFinancialParser(TransactionParser):
+    PURCHASE_EXTRACT_RE = re.compile(
+        r"your purchase of \$(?P<amount>[0-9]+\.[0-9]{2}) at (?P<payee>.+) on (?P<date>(January|February|March|April|May|June|July|August|September|October|November|December) \d{1,2}(st|nd|rd|th)?, \d{4})",  # noqa: E501
+        re.MULTILINE,
+    )
+
+    def __init__(self, account_id: UUID):
+        self.account_id = account_id
+
+    def match(self, msg: Message) -> bool:
+        return msg["From"] == "Neo Financial <info@neofinancial.com>" and msg[
+            "Subject"
+        ].startswith("You earned")
+
+    def extract_purchase(self, msg: EmailMessage) -> Optional[Transaction]:
+        content = self.get_content(msg, preferencelist=["plain"])
+        matches = self.PURCHASE_EXTRACT_RE.search(content)
+        if matches is None:
+            debug(f"No match on\n\n{content}")
+            raise TransactionParsingFailed("no matches for purchase extraction RE")
+
+        amount = Decimal(matches["amount"].replace(",", "")) * -1
+        # Use date from email header, it seems to match the body and includes TZ info
+        date = parse_email_time(msg["Date"]).date()
+        return Transaction(
+            account=self.account_id,
+            date=date,
+            amount=amount,
+            payee=matches["payee"],
+            notes="via email",
+            imported_id=msg["Message-ID"],
+        )
+
+    def extract(self, msg: EmailMessage) -> Optional[Transaction]:
+        if msg["Subject"].startswith("You earned"):
+            return self.extract_purchase(msg)
+        return None

actual_imap_poll/watcher.py

@@ -1,6 +1,7 @@
 #!/usr/bin/env python
 import asyncio
 import email.policy
+import importlib.metadata
 import importlib.util
 import logging
 import os
@@ -129,6 +130,11 @@ class App:
 
 def main():
     logging.basicConfig(level=getenv("LOG_LEVEL", "INFO"))
+    logging.info(
+        "actual-imap-parser %s starting up",
+        importlib.metadata.version("actual-imap-poll"),
+    )
     app_config = AppConfig()
+    logging.debug("Config: %s", app_config)
     app = App(app_config, CONFIG_PATH)
     asyncio.run(app.run())

package.json

@@ -14,7 +14,7 @@
   "license": "ISC",
   "type": "commonjs",
   "dependencies": {
-    "@actual-app/api": "^25.11.0",
+    "@actual-app/api": "^26.5.2",
     "yargs": "^17.7.2"
   }
 }

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "actual-imap-poll"
-version = "0.1.1"
+version = "0.2.0"
 description = "Poll an IMAP mailbox looking for transactions to submit to Actual Budget"
 authors = [
     {name = "Keenan Tims",email = "ktims@gotroot.ca"}