Compare commits

..

2 Commits

Author SHA1 Message Date
Andreas Knuth 397a2f7d98 Merge branch 'main' of git.bizmatch.net:aknuth/email-amazon 2026-01-12 17:19:54 -06:00
Andreas Knuth be9c5b4ceb cleanwork 2026-01-12 17:19:44 -06:00
17 changed files with 0 additions and 324 deletions

13
TESTS
View File

@ -1,13 +0,0 @@
# Via AWS SES CLI
aws ses send-email \
--from "sender@example.com" \
--destination "ToAddresses=test@andreasknuth.de" \
--message "Subject={Data='Test',Charset=utf-8},Body={Text={Data='Test message',Charset=utf-8}}" \
--region us-east-2
# Mail an mehrere Domains
aws ses send-email \
--from "sender@example.com" \
--destination "ToAddresses=test@andreasknuth.de,test@bizmatch.net" \
--message "Subject={Data='Multi-Domain Test',Charset=utf-8},Body={Text={Data='Testing multiple domains',Charset=utf-8}}" \
--region us-east-2

View File

@ -1,24 +0,0 @@
import sys
import email
from email.utils import getaddresses
if len(sys.argv) < 2:
print("Usage: python3 extract_email_headers.py <email_file>")
sys.exit(1)
file_path = sys.argv[1]
with open(file_path, 'rb') as f:
msg = email.message_from_bytes(f.read())
# From: Nur die reine Email-Adresse
from_header = msg.get('From', '')
from_name, from_addr = email.utils.parseaddr(from_header)
# Recipients: Alle To und Cc, nur reine Emails, kommagetrennt
to_addrs = getaddresses(msg.get_all('To', []))
cc_addrs = getaddresses(msg.get_all('Cc', []))
all_addrs = to_addrs + cc_addrs
recipients = ','.join([addr for name, addr in all_addrs if addr]) if all_addrs else ''
print(f'FROM:{from_addr}')
print(f'RECIPIENTS:{recipients}')

View File

@ -1,123 +0,0 @@
import json
import os
import boto3
import uuid
import logging
from datetime import datetime
from botocore.exceptions import ClientError
import time
import random
# Logging konfigurieren
logger = logging.getLogger()
logger.setLevel(logging.INFO)
sqs = boto3.client('sqs')
# Retry-Konfiguration
MAX_RETRIES = 3
BASE_BACKOFF = 1 # Sekunden
def exponential_backoff(attempt):
"""Exponential Backoff mit Jitter"""
return BASE_BACKOFF * (2 ** attempt) + random.uniform(0, 1)
def get_queue_url(domain):
"""
Generiert Queue-Namen aus Domain und holt URL.
Konvention: domain.tld -> domain-tld-queue
"""
queue_name = domain.replace('.', '-') + '-queue'
try:
response = sqs.get_queue_url(QueueName=queue_name)
return response['QueueUrl']
except ClientError as e:
if e.response['Error']['Code'] == 'AWS.SimpleQueueService.NonExistentQueue':
logger.error(f"Queue nicht gefunden für Domain: {domain}")
raise ValueError(f"Keine Queue für Domain {domain}")
else:
raise
def lambda_handler(event, context):
"""
Nimmt SES Event entgegen, extrahiert Domain dynamisch,
verpackt Metadaten als 'Fake SNS' und sendet an die domain-spezifische SQS.
Mit integrierter Retry-Logik für SQS-Send.
"""
try:
records = event.get('Records', [])
logger.info(f"Received event with {len(records)} records.")
for record in records:
ses_data = record.get('ses', {})
if not ses_data:
logger.warning(f"Invalid SES event: Missing 'ses' in record: {record}")
continue
mail = ses_data.get('mail', {})
receipt = ses_data.get('receipt', {})
# Domain extrahieren (aus erstem Recipient)
recipients = receipt.get('recipients', []) or mail.get('destination', [])
if not recipients:
logger.warning("No recipients in event - skipping")
continue
first_recipient = recipients[0]
domain = first_recipient.split('@')[-1].lower()
if not domain:
logger.error("Could not extract domain from recipient")
continue
# Wichtige Metadaten loggen
msg_id = mail.get('messageId', 'unknown')
source = mail.get('source', 'unknown')
logger.info(f"Processing Message-ID: {msg_id} for domain: {domain}")
logger.info(f" From: {source}")
logger.info(f" To: {recipients}")
# SES JSON als String serialisieren
ses_json_string = json.dumps(ses_data)
# Payload Größe loggen und checken (Safeguard)
payload_size = len(ses_json_string.encode('utf-8'))
logger.info(f" Metadata Payload Size: {payload_size} bytes")
if payload_size > 200000: # Arbitrary Limit < SQS 256KB
raise ValueError("Payload too large for SQS")
# Fake SNS Payload
fake_sns_payload = {
"Type": "Notification",
"MessageId": str(uuid.uuid4()),
"TopicArn": "arn:aws:sns:ses-shim:global-topic",
"Subject": "Amazon SES Email Receipt Notification",
"Message": ses_json_string,
"Timestamp": datetime.utcnow().isoformat() + "Z"
}
# Queue URL dynamisch holen
queue_url = get_queue_url(domain)
# SQS Send mit Retries
attempt = 0
while attempt < MAX_RETRIES:
try:
sqs.send_message(
QueueUrl=queue_url,
MessageBody=json.dumps(fake_sns_payload)
)
logger.info(f"✅ Successfully forwarded {msg_id} to SQS: {queue_url}")
break
except ClientError as e:
attempt += 1
error_code = e.response['Error']['Code']
logger.warning(f"Retry {attempt}/{MAX_RETRIES} for SQS send: {error_code} - {str(e)}")
if attempt == MAX_RETRIES:
raise
time.sleep(exponential_backoff(attempt))
return {'status': 'ok'}
except Exception as e:
logger.error(f"❌ Critical Error in Lambda Shim: {str(e)}", exc_info=True)
raise e

View File

@ -1,164 +0,0 @@
#!/usr/bin/env python3
"""
Test script für Message-ID Extraktion - VERBESSERTE VERSION
Kann lokal ausgeführt werden ohne AWS-Verbindung
"""
import re
from email.parser import BytesParser
from email.policy import SMTP as SMTPPolicy
def log(message: str, level: str = 'INFO'):
"""Dummy log für Tests"""
print(f"[{level}] {message}")
def extract_original_message_id(parsed):
"""
Extrahiert Original SES Message-ID aus Email
SES Format: 010f[hex32]-[hex8]-[hex4]-[hex4]-[hex4]-[hex12]-000000
"""
import re
# SES Message-ID Pattern (endet immer mit -000000)
ses_pattern = re.compile(r'010f[0-9a-f]{12}-[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}-000000')
# Die Message-ID der aktuellen Email (Bounce selbst) - diese wollen wir NICHT
current_msg_id = (parsed.get('Message-ID') or '').strip()
current_match = ses_pattern.search(current_msg_id)
current_id = current_match.group(0) if current_match else None
log(f"Current Message-ID: {current_id}", 'DEBUG')
# 1. Versuche Standard-Header (In-Reply-To, References)
for header in ['In-Reply-To', 'References']:
value = (parsed.get(header) or '').strip()
if value:
match = ses_pattern.search(value)
if match:
found_id = match.group(0)
# Nur nehmen wenn es NICHT die aktuelle Bounce-ID ist
if found_id != current_id:
log(f" Found Message-ID in {header}: {found_id}")
return found_id
# 2. Durchsuche den kompletten Email-Body (inkl. ALLE Attachments/Parts)
try:
body_text = ''
# Hole den kompletten Body als String
if parsed.is_multipart():
for part in parsed.walk():
content_type = part.get_content_type()
# SPEZIALFALL: message/rfc822 (eingebettete Messages)
if content_type == 'message/rfc822':
log(f" Processing embedded message/rfc822", 'DEBUG')
try:
# get_payload() gibt eine Liste mit einem EmailMessage-Objekt zurück!
payload = part.get_payload()
if isinstance(payload, list) and len(payload) > 0:
embedded_msg = payload[0]
# Hole Message-ID aus dem eingebetteten Message
embedded_id = (embedded_msg.get('Message-ID') or '').strip()
match = ses_pattern.search(embedded_id)
if match:
found_id = match.group(0)
log(f" Found ID in embedded msg: {found_id}", 'DEBUG')
# Nur nehmen wenn es NICHT die aktuelle Bounce-ID ist
if found_id != current_id:
log(f" ✓ Found Message-ID in embedded message: {found_id}")
return found_id
# Fallback: Konvertiere eingebettete Message zu String
body_text += embedded_msg.as_string()
except Exception as e:
log(f" Warning: Could not process embedded message: {e}", 'WARNING')
# Durchsuche ALLE anderen Parts (außer Binärdaten wie images)
elif content_type.startswith('text/') or content_type.startswith('application/'):
try:
payload = part.get_payload(decode=True)
if payload:
# Versuche als UTF-8, fallback auf Latin-1
try:
body_text += payload.decode('utf-8', errors='ignore')
except:
try:
body_text += payload.decode('latin-1', errors='ignore')
except:
# Letzter Versuch: als ASCII mit ignore
body_text += str(payload, errors='ignore')
except:
# Falls decode fehlschlägt, String-Payload holen
payload = part.get_payload()
if isinstance(payload, str):
body_text += payload
else:
# Nicht-Multipart Message
payload = parsed.get_payload(decode=True)
if payload:
try:
body_text = payload.decode('utf-8', errors='ignore')
except:
body_text = payload.decode('latin-1', errors='ignore')
# Suche alle SES Message-IDs im Body
matches = ses_pattern.findall(body_text)
if matches:
log(f" Found {len(matches)} total IDs in body: {matches}", 'DEBUG')
# Filtere die aktuelle Bounce-ID raus
candidates = [m for m in matches if m != current_id]
if candidates:
# Nehme die ERSTE der verbleibenden (meist die Original-ID)
log(f" Found {len(matches)} SES Message-ID(s) in body, using first (not bounce): {candidates[0]}")
return candidates[0]
else:
log(f" Found {len(matches)} SES Message-ID(s) but all match the bounce ID")
except Exception as e:
log(f" Warning: Could not search body for Message-ID: {e}", 'WARNING')
return None
def test_with_file(filepath: str):
"""Test mit einer echten Email-Datei"""
print(f"\n{'='*70}")
print(f"Testing: {filepath}")
print('='*70)
with open(filepath, 'rb') as f:
raw_bytes = f.read()
parsed = BytesParser(policy=SMTPPolicy).parsebytes(raw_bytes)
print(f"\nEmail Headers:")
print(f" From: {parsed.get('From')}")
print(f" To: {parsed.get('To')}")
print(f" Subject: {parsed.get('Subject')}")
print(f" Message-ID: {parsed.get('Message-ID')}")
print(f" In-Reply-To: {parsed.get('In-Reply-To')}")
print(f" References: {parsed.get('References')}")
print(f"\n--- EXTRACTION ---")
result = extract_original_message_id(parsed)
print(f"\n{'='*70}")
print(f"RESULT: {result}")
print('='*70)
return result
if __name__ == '__main__':
import sys
if len(sys.argv) > 1:
# Email-Datei als Argument
result = test_with_file(sys.argv[1])
# Exit code: 0 = success (ID found), 1 = failure (no ID)
sys.exit(0 if result else 1)
else:
print("Usage: python3 test_extract_v2.py <email_file>")
sys.exit(1)