4.42.2

app/app/alias_utils.py

@@ -308,28 +308,29 @@ def delete_alias(alias: Alias, user: User):
     Delete an alias and add it to either global or domain trash
     Should be used instead of Alias.delete, DomainDeletedAlias.create, DeletedAlias.create
     """
-    # save deleted alias to either global or domain trash
+    LOG.i(f"User {user} has deleted alias {alias}")
+    # save deleted alias to either global or domain tra
     if alias.custom_domain_id:
         if not DomainDeletedAlias.get_by(
             email=alias.email, domain_id=alias.custom_domain_id
         ):
-            LOG.d("add %s to domain %s trash", alias, alias.custom_domain_id)
+            domain_deleted_alias = DomainDeletedAlias(
-            Session.add(
-                DomainDeletedAlias(
                 user_id=user.id,
                 email=alias.email,
                 domain_id=alias.custom_domain_id,
             )
-            )
+            Session.add(domain_deleted_alias)
             Session.commit()
-
+            LOG.i(
+                f"Moving {alias} to domain {alias.custom_domain_id} trash {domain_deleted_alias}"
+            )
     else:
         if not DeletedAlias.get_by(email=alias.email):
-            LOG.d("add %s to global trash", alias)
+            deleted_alias = DeletedAlias(email=alias.email)
-            Session.add(DeletedAlias(email=alias.email))
+            Session.add(deleted_alias)
             Session.commit()
+            LOG.i(f"Moving {alias} to global trash {deleted_alias}")
 
-    LOG.i("delete alias %s", alias)
     Alias.filter(Alias.id == alias.id).delete()
     Session.commit()
 

app/app/config.py

@@ -434,6 +434,8 @@ HIBP_MAX_ALIAS_CHECK = 10_000
 HIBP_RPM = 100
 HIBP_SKIP_PARTNER_ALIAS = os.environ.get("HIBP_SKIP_PARTNER_ALIAS")
 
+KEEP_OLD_DATA_DAYS = 30
+
 POSTMASTER = os.environ.get("POSTMASTER")
 
 # store temporary files, especially for debugging

app/app/dashboard/views/enter_sudo.py

@@ -14,7 +14,7 @@ from app.models import PartnerUser, SocialAuth
 from app.proton.utils import get_proton_partner
 from app.utils import sanitize_next_url
 
-_SUDO_GAP = 900
+_SUDO_GAP = 120
 
 
 class LoginForm(FlaskForm):

app/app/dashboard/views/index.py

@@ -141,7 +141,7 @@ def index():
                 )
 
             if request.form.get("form-name") == "delete-alias":
-                LOG.d("delete alias %s", alias)
+                LOG.i(f"User {current_user} requested deletion of alias {alias}")
                 email = alias.email
                 alias_utils.delete_alias(alias, current_user)
                 flash(f"Alias {email} has been deleted", "success")

app/app/email/headers.py

@@ -21,6 +21,7 @@ LIST_UNSUBSCRIBE = "List-Unsubscribe"
 LIST_UNSUBSCRIBE_POST = "List-Unsubscribe-Post"
 RETURN_PATH = "Return-Path"
 AUTHENTICATION_RESULTS = "Authentication-Results"
+SL_QUEUE_ID = "X-SL-Queue-Id"
 
 # headers used to DKIM sign in order of preference
 DKIM_HEADERS = [

app/app/email_utils.py

@@ -494,9 +494,10 @@ def delete_header(msg: Message, header: str):
 
 def sanitize_header(msg: Message, header: str):
     """remove trailing space and remove linebreak from a header"""
+    header_lowercase = header.lower()
     for i in reversed(range(len(msg._headers))):
         header_name = msg._headers[i][0].lower()
-        if header_name == header.lower():
+        if header_name == header_lowercase:
             # msg._headers[i] is a tuple like ('From', 'hey@google.com')
             if msg._headers[i][1]:
                 msg._headers[i] = (

app/app/handler/dmarc.py

@@ -30,7 +30,9 @@ def apply_dmarc_policy_for_forward_phase(
 ) -> Tuple[Message, Optional[str]]:
     spam_result = SpamdResult.extract_from_headers(msg, Phase.forward)
     if not DMARC_CHECK_ENABLED or not spam_result:
+        LOG.i("DMARC check disabled")
         return msg, None
+    LOG.i(f"Spam check result in {spam_result}")
 
     from_header = get_header_unicode(msg[headers.FROM])
 
@@ -150,8 +152,10 @@ def apply_dmarc_policy_for_reply_phase(
 ) -> Optional[str]:
     spam_result = SpamdResult.extract_from_headers(msg, Phase.reply)
     if not DMARC_CHECK_ENABLED or not spam_result:
+        LOG.i("DMARC check disabled")
         return None
 
+    LOG.i(f"Spam check result is {spam_result}")
     if spam_result.dmarc not in (
         DmarcCheckResult.quarantine,
         DmarcCheckResult.reject,

app/app/import_utils.py

@@ -30,7 +30,10 @@ def handle_batch_import(batch_import: BatchImport):
 
     LOG.d("Download file %s from %s", batch_import.file, file_url)
     r = requests.get(file_url)
-    lines = [line.decode("utf-8") for line in r.iter_lines()]
+    # Replace invisible character
+    lines = [
+        line.decode("utf-8").replace("\ufeff", "").strip() for line in r.iter_lines()
+    ]
 
     import_from_csv(batch_import, user, lines)
 

app/app/rate_limiter.py

@@ -30,7 +30,9 @@ def check_bucket_limit(
     try:
         value = lock_redis.incr(bucket_lock_name, bucket_seconds)
         if value > max_hits:
-            LOG.i(f"Rate limit hit for {bucket_lock_name} -> {value}/{max_hits}")
+            LOG.i(
+                f"Rate limit hit for {lock_name} (bucket id {bucket_id}) -> {value}/{max_hits}"
+            )
             newrelic.agent.record_custom_event(
                 "BucketRateLimit",
                 {"lock_name": lock_name, "bucket_seconds": bucket_seconds},

app/app/s3.py

@@ -5,19 +5,9 @@ from typing import Optional
 import boto3
 import requests
 
-from app.config import (
+from app import config
-    AWS_REGION,
-    BUCKET,
-    AWS_ACCESS_KEY_ID,
-    AWS_SECRET_ACCESS_KEY,
-    LOCAL_FILE_UPLOAD,
-    UPLOAD_DIR,
-    URL,
-    AWS_ENDPOINT_URL,
-)
 from app.log import LOG
 
-
 _s3_client = None
 
 
@@ -25,12 +15,12 @@ def _get_s3client():
     global _s3_client
     if _s3_client is None:
         args = {
-            "aws_access_key_id": AWS_ACCESS_KEY_ID,
+            "aws_access_key_id": config.AWS_ACCESS_KEY_ID,
-            "aws_secret_access_key": AWS_SECRET_ACCESS_KEY,
+            "aws_secret_access_key": config.AWS_SECRET_ACCESS_KEY,
-            "region_name": AWS_REGION,
+            "region_name": config.AWS_REGION,
         }
-        if AWS_ENDPOINT_URL:
+        if config.AWS_ENDPOINT_URL:
-            args["endpoint_url"] = AWS_ENDPOINT_URL
+            args["endpoint_url"] = config.AWS_ENDPOINT_URL
         _s3_client = boto3.client("s3", **args)
     return _s3_client
 
@@ -38,8 +28,8 @@ def _get_s3client():
 def upload_from_bytesio(key: str, bs: BytesIO, content_type="application/octet-stream"):
     bs.seek(0)
 
-    if LOCAL_FILE_UPLOAD:
+    if config.LOCAL_FILE_UPLOAD:
-        file_path = os.path.join(UPLOAD_DIR, key)
+        file_path = os.path.join(config.UPLOAD_DIR, key)
         file_dir = os.path.dirname(file_path)
         os.makedirs(file_dir, exist_ok=True)
         with open(file_path, "wb") as f:
@@ -47,7 +37,7 @@ def upload_from_bytesio(key: str, bs: BytesIO, content_type="application/octet-s
 
     else:
         _get_s3client().put_object(
-            Bucket=BUCKET,
+            Bucket=config.BUCKET,
             Key=key,
             Body=bs,
             ContentType=content_type,
@@ -57,8 +47,8 @@ def upload_from_bytesio(key: str, bs: BytesIO, content_type="application/octet-s
 def upload_email_from_bytesio(path: str, bs: BytesIO, filename):
     bs.seek(0)
 
-    if LOCAL_FILE_UPLOAD:
+    if config.LOCAL_FILE_UPLOAD:
-        file_path = os.path.join(UPLOAD_DIR, path)
+        file_path = os.path.join(config.UPLOAD_DIR, path)
         file_dir = os.path.dirname(file_path)
         os.makedirs(file_dir, exist_ok=True)
         with open(file_path, "wb") as f:
@@ -66,7 +56,7 @@ def upload_email_from_bytesio(path: str, bs: BytesIO, filename):
 
     else:
         _get_s3client().put_object(
-            Bucket=BUCKET,
+            Bucket=config.BUCKET,
             Key=path,
             Body=bs,
             # Support saving a remote file using Http header
@@ -77,12 +67,12 @@ def upload_email_from_bytesio(path: str, bs: BytesIO, filename):
 
 
 def download_email(path: str) -> Optional[str]:
-    if LOCAL_FILE_UPLOAD:
+    if config.LOCAL_FILE_UPLOAD:
-        file_path = os.path.join(UPLOAD_DIR, path)
+        file_path = os.path.join(config.UPLOAD_DIR, path)
         with open(file_path, "rb") as f:
             return f.read()
     resp = _get_s3client().get_object(
-        Bucket=BUCKET,
+        Bucket=config.BUCKET,
         Key=path,
     )
     if not resp or "Body" not in resp:
@@ -96,29 +86,30 @@ def upload_from_url(url: str, upload_path):
 
 
 def get_url(key: str, expires_in=3600) -> str:
-    if LOCAL_FILE_UPLOAD:
+    if config.LOCAL_FILE_UPLOAD:
-        return URL + "/static/upload/" + key
+        return config.URL + "/static/upload/" + key
     else:
         return _get_s3client().generate_presigned_url(
             ExpiresIn=expires_in,
             ClientMethod="get_object",
-            Params={"Bucket": BUCKET, "Key": key},
+            Params={"Bucket": config.BUCKET, "Key": key},
         )
 
 
 def delete(path: str):
-    if LOCAL_FILE_UPLOAD:
+    if config.LOCAL_FILE_UPLOAD:
-        os.remove(os.path.join(UPLOAD_DIR, path))
+        file_path = os.path.join(config.UPLOAD_DIR, path)
+        os.remove(file_path)
     else:
-        _get_s3client().delete_object(Bucket=BUCKET, Key=path)
+        _get_s3client().delete_object(Bucket=config.BUCKET, Key=path)
 
 
 def create_bucket_if_not_exists():
     s3client = _get_s3client()
     buckets = s3client.list_buckets()
     for bucket in buckets["Buckets"]:
-        if bucket["Name"] == BUCKET:
+        if bucket["Name"] == config.BUCKET:
             LOG.i("Bucket already exists")
             return
-    s3client.create_bucket(Bucket=BUCKET)
+    s3client.create_bucket(Bucket=config.BUCKET)
-    LOG.i(f"Bucket {BUCKET} created")
+    LOG.i(f"Bucket {config.BUCKET} created")

app/cron.py

@@ -61,6 +61,9 @@ from app.pgp_utils import load_public_key_and_check, PGPException
 from app.proton.utils import get_proton_partner
 from app.utils import sanitize_email
 from server import create_light_app
+from tasks.cleanup_old_imports import cleanup_old_imports
+from tasks.cleanup_old_jobs import cleanup_old_jobs
+from tasks.cleanup_old_notifications import cleanup_old_notifications
 
 DELETE_GRACE_DAYS = 30
 
@@ -1221,6 +1224,13 @@ def clear_users_scheduled_to_be_deleted(dry_run=False):
         Session.commit()
 
 
+def delete_old_data():
+    oldest_valid = arrow.now().shift(days=-config.KEEP_OLD_DATA_DAYS)
+    cleanup_old_imports(oldest_valid)
+    cleanup_old_jobs(oldest_valid)
+    cleanup_old_notifications(oldest_valid)
+
+
 if __name__ == "__main__":
     LOG.d("Start running cronjob")
     parser = argparse.ArgumentParser()
@@ -1235,6 +1245,7 @@ if __name__ == "__main__":
             "notify_manual_subscription_end",
             "notify_premium_end",
             "delete_logs",
+            "delete_old_data",
             "poll_apple_subscription",
             "sanity_check",
             "delete_old_monitoring",
@@ -1263,6 +1274,9 @@ if __name__ == "__main__":
         elif args.job == "delete_logs":
             LOG.d("Deleted Logs")
             delete_logs()
+        elif args.job == "delete_old_data":
+            LOG.d("Delete old data")
+            delete_old_data()
         elif args.job == "poll_apple_subscription":
             LOG.d("Poll Apple Subscriptions")
             poll_apple_subscription()

app/crontab.yml

@@ -37,6 +37,12 @@ jobs:
     schedule: "15 5 * * *"
     captureStderr: true
 
+  - name: SimpleLogin Delete Old data
+    command: python /code/cron.py -j delete_old_data
+    shell: /bin/bash
+    schedule: "30 5 * * *"
+    captureStderr: true
+
   - name: SimpleLogin Poll Apple Subscriptions
     command: python /code/cron.py -j poll_apple_subscription
     shell: /bin/bash

app/email_handler.py

@@ -875,6 +875,7 @@ def forward_email_to_mailbox(
         # References and In-Reply-To are used for keeping the email thread
         headers.REFERENCES,
         headers.IN_REPLY_TO,
+        headers.SL_QUEUE_ID,
         headers.LIST_UNSUBSCRIBE,
         headers.LIST_UNSUBSCRIBE_POST,
     ] + headers.MIME_HEADERS
@@ -2040,10 +2041,11 @@ def handle(envelope: Envelope, msg: Message) -> str:
         return status.E204
 
     # sanitize email headers
-    sanitize_header(msg, "from")
+    sanitize_header(msg, headers.FROM)
-    sanitize_header(msg, "to")
+    sanitize_header(msg, headers.TO)
-    sanitize_header(msg, "cc")
+    sanitize_header(msg, headers.CC)
-    sanitize_header(msg, "reply-to")
+    sanitize_header(msg, headers.REPLY_TO)
+    sanitize_header(msg, headers.MESSAGE_ID)
 
     LOG.d(
         "==>> Handle mail_from:%s, rcpt_tos:%s, header_from:%s, header_to:%s, "

app/tasks/cleanup_old_imports.py

@@ -0,0 +1,19 @@
+import arrow
+
+from app import s3
+from app.log import LOG
+from app.models import BatchImport
+
+
+def cleanup_old_imports(oldest_allowed: arrow.Arrow):
+    LOG.i(f"Deleting imports older than {oldest_allowed}")
+    for batch_import in (
+        BatchImport.filter(BatchImport.created_at < oldest_allowed).yield_per(500).all()
+    ):
+        LOG.i(
+            f"Deleting batch import {batch_import} with file {batch_import.file.path}"
+        )
+        file = batch_import.file
+        if file is not None:
+            s3.delete(file.path)
+        BatchImport.delete(batch_import.id, commit=True)

app/tasks/cleanup_old_jobs.py

@@ -0,0 +1,24 @@
+import arrow
+from sqlalchemy import or_, and_
+
+from app import config
+from app.db import Session
+from app.log import LOG
+from app.models import Job, JobState
+
+
+def cleanup_old_jobs(oldest_allowed: arrow.Arrow):
+    LOG.i(f"Deleting jobs older than {oldest_allowed}")
+    count = Job.filter(
+        or_(
+            Job.state == JobState.done.value,
+            Job.state == JobState.error.value,
+            and_(
+                Job.state == JobState.taken.value,
+                Job.attempts >= config.JOB_MAX_ATTEMPTS,
+            ),
+        ),
+        Job.updated_at < oldest_allowed,
+    ).delete()
+    Session.commit()
+    LOG.i(f"Deleted {count} jobs")

app/tasks/cleanup_old_notifications.py

@@ -0,0 +1,12 @@
+import arrow
+
+from app.db import Session
+from app.log import LOG
+from app.models import Notification
+
+
+def cleanup_old_notifications(oldest_allowed: arrow.Arrow):
+    LOG.i(f"Deleting notifications older than {oldest_allowed}")
+    count = Notification.filter(Notification.created_at < oldest_allowed).delete()
+    Session.commit()
+    LOG.i(f"Deleted {count} notifications")

app/tests/cron/test_get_alias_for_hibp.py

@@ -28,7 +28,7 @@ def test_get_alias_for_free_user_has_no_alias():
     assert len(aliases) == 0
 
 
-def test_get_alias_for_lifetime():
+def test_get_alias_for_lifetime_with_null_hibp_date():
     user = create_new_user()
     user.lifetime = True
     alias_id = Alias.create_new_random(user).id
@@ -39,6 +39,19 @@ def test_get_alias_for_lifetime():
     assert alias_id == aliases[0].id
 
 
+def test_get_alias_for_lifetime_with_old_hibp_date():
+    user = create_new_user()
+    user.lifetime = True
+    alias = Alias.create_new_random(user)
+    alias.hibp_last_check = arrow.now().shift(days=-1)
+    alias_id = alias.id
+    Session.commit()
+    aliases = list(
+        cron.get_alias_to_check_hibp(arrow.now(), [], alias_id, alias_id + 1)
+    )
+    assert alias_id == aliases[0].id
+
+
 def create_partner_sub(user: User):
     pu = PartnerUser.create(
         partner_id=get_proton_partner().id,
@@ -114,3 +127,16 @@ def test_skipped_user_is_not_checked():
         cron.get_alias_to_check_hibp(arrow.now(), [user.id], alias_id, alias_id + 1)
     )
     assert len(aliases) == 0
+
+
+def test_already_checked_is_not_checked():
+    user = create_new_user()
+    user.lifetime = True
+    alias = Alias.create_new_random(user)
+    alias.hibp_last_check = arrow.now().shift(days=1)
+    alias_id = alias.id
+    Session.commit()
+    aliases = list(
+        cron.get_alias_to_check_hibp(arrow.now(), [user.id], alias_id, alias_id + 1)
+    )
+    assert len(aliases) == 0

app/tests/tasks/test_cleanup_old_imports.py

@@ -0,0 +1,35 @@
+import tempfile
+from io import BytesIO
+
+import arrow
+
+from app import s3, config
+from app.models import File, BatchImport
+from tasks.cleanup_old_imports import cleanup_old_imports
+from tests.utils import random_token, create_new_user
+
+
+def test_cleanup_old_imports():
+    BatchImport.filter().delete()
+    with tempfile.TemporaryDirectory() as tmpdir:
+        config.UPLOAD_DIR = tmpdir
+        user = create_new_user()
+        path = random_token()
+        s3.upload_from_bytesio(path, BytesIO("data".encode("utf-8")))
+        file = File.create(path=path, commit=True)  # noqa: F821
+        now = arrow.now()
+        delete_batch_import_id = BatchImport.create(
+            user_id=user.id,
+            file_id=file.id,
+            created_at=now.shift(minutes=-1),
+            flush=True,
+        ).id
+        keep_batch_import_id = BatchImport.create(
+            user_id=user.id,
+            file_id=file.id,
+            created_at=now.shift(minutes=+1),
+            commit=True,
+        ).id
+        cleanup_old_imports(now)
+        assert BatchImport.get(id=delete_batch_import_id) is None
+        assert BatchImport.get(id=keep_batch_import_id) is not None

app/tests/tasks/test_cleanup_old_jobs.py

@@ -0,0 +1,72 @@
+import arrow
+
+from app import config
+from app.models import Job, JobState
+from tasks.cleanup_old_jobs import cleanup_old_jobs
+
+
+def test_cleanup_old_jobs():
+    Job.filter().delete()
+    now = arrow.now()
+    delete_ids = [
+        Job.create(
+            updated_at=now.shift(minutes=-1),
+            state=JobState.done.value,
+            name="",
+            payload="",
+            flush=True,
+        ).id,
+        Job.create(
+            updated_at=now.shift(minutes=-1),
+            state=JobState.error.value,
+            name="",
+            payload="",
+            flush=True,
+        ).id,
+        Job.create(
+            updated_at=now.shift(minutes=-1),
+            state=JobState.taken.value,
+            attempts=config.JOB_MAX_ATTEMPTS,
+            name="",
+            payload="",
+            flush=True,
+        ).id,
+    ]
+
+    keep_ids = [
+        Job.create(
+            updated_at=now.shift(minutes=+1),
+            state=JobState.done.value,
+            name="",
+            payload="",
+            flush=True,
+        ).id,
+        Job.create(
+            updated_at=now.shift(minutes=+1),
+            state=JobState.error.value,
+            name="",
+            payload="",
+            flush=True,
+        ).id,
+        Job.create(
+            updated_at=now.shift(minutes=+1),
+            state=JobState.taken.value,
+            attempts=config.JOB_MAX_ATTEMPTS,
+            name="",
+            payload="",
+            flush=True,
+        ).id,
+        Job.create(
+            updated_at=now.shift(minutes=-1),
+            state=JobState.taken.value,
+            attempts=config.JOB_MAX_ATTEMPTS - 1,
+            name="",
+            payload="",
+            flush=True,
+        ).id,
+    ]
+    cleanup_old_jobs(now)
+    for delete_id in delete_ids:
+        assert Job.get(id=delete_id) is None
+    for keep_id in keep_ids:
+        assert Job.get(id=keep_id) is not None

app/tests/tasks/test_cleanup_old_notifications.py

@@ -0,0 +1,26 @@
+import arrow
+
+from app.models import Notification
+from tasks.cleanup_old_notifications import cleanup_old_notifications
+from tests.utils import create_new_user
+
+
+def test_cleanup_old_notifications():
+    Notification.filter().delete()
+    user = create_new_user()
+    now = arrow.now()
+    delete_id = Notification.create(
+        user_id=user.id,
+        created_at=now.shift(minutes=-1),
+        message="",
+        flush=True,
+    ).id
+    keep_id = Notification.create(
+        user_id=user.id,
+        created_at=now.shift(minutes=+1),
+        message="",
+        flush=True,
+    ).id
+    cleanup_old_notifications(now)
+    assert Notification.get(id=delete_id) is None
+    assert Notification.get(id=keep_id) is not None

app/tests/test_email_handler.py

@@ -384,3 +384,30 @@ def test_break_loop_alias_as_mailbox(flask_client):
     msg[headers.SUBJECT] = random_string()
     result = email_handler.handle(envelope, msg)
     assert result == status.E525
+
+
+@mail_sender.store_emails_test_decorator
+def test_preserve_headers(flask_client):
+    headers_to_keep = [
+        headers.SUBJECT,
+        headers.DATE,
+        headers.MESSAGE_ID,
+        headers.REFERENCES,
+        headers.IN_REPLY_TO,
+        headers.SL_QUEUE_ID,
+    ] + headers.MIME_HEADERS
+    user = create_new_user()
+    alias = Alias.create_new_random(user)
+    envelope = Envelope()
+    envelope.mail_from = "somewhere@lo.cal"
+    envelope.rcpt_tos = [alias.email]
+    msg = EmailMessage()
+    for header in headers_to_keep:
+        msg[header] = header + "keep"
+    result = email_handler.handle(envelope, msg)
+    assert result == status.E200
+    sent_mails = mail_sender.get_stored_emails()
+    assert len(sent_mails) == 1
+    msg = sent_mails[0].msg
+    for header in headers_to_keep:
+        assert msg[header] == header + "keep"