Honor explicit allauth opt-in

.github/scripts/discover_test_matrix.py

@@ -20,7 +20,13 @@ def main() -> None:
                 "name": f"main/{path.stem}",
                 "paths": [test_path],
                 "paths_arg": test_path,
-                "extra": "ldap" if test_path.endswith("/test_auth_ldap.py") else "",
+                "extra": (
+                    "ldap"
+                    if test_path.endswith("/test_auth_ldap.py")
+                    else "allauth"
+                    if test_path.endswith("/test_allauth_integration.py")
+                    else ""
+                ),
                 "count": 1,
             },
         )

archivebox/core/admin_site.py

@@ -11,8 +11,10 @@ from admin_data_views.admin import (
 from admin_data_views.admin import (
     get_app_list as adv_get_app_list,
 )
+from django.conf import settings
 from django.contrib import admin
-from django.contrib.auth import REDIRECT_FIELD_NAME, get_user_model, login as auth_login
+from django.contrib.auth import REDIRECT_FIELD_NAME, get_user_model
+from django.contrib.auth import login as auth_login
 from django.contrib.auth.decorators import login_not_required
 from django.contrib.auth.forms import UserCreationForm
 from django.contrib.auth.views import LoginView
@@ -31,7 +33,6 @@ from archivebox.core.routes_util import is_allowed_archivebox_redirect_url
 if TYPE_CHECKING:
     from admin_data_views.typing import AppDict
     from django.http import HttpRequest
-    from django.template.response import TemplateResponse
     from django.urls import URLPattern, URLResolver
 
 
@@ -114,6 +115,7 @@ class ArchiveBoxAdmin(admin.AdminSite):
             "app_path": request.get_full_path(),
             "username": request.user.get_username(),
             "first_admin_setup": first_admin_setup,
+            "allauth_enabled": settings.ALLAUTH_ENABLED,
         }
         if REDIRECT_FIELD_NAME not in request.GET and REDIRECT_FIELD_NAME not in request.POST:
             context[REDIRECT_FIELD_NAME] = reverse("admin:index", current_app=self.name)

archivebox/core/settings.py

@@ -43,6 +43,7 @@ WSGI_APPLICATION = "archivebox.core.wsgi.application"
 ASGI_APPLICATION = "archivebox.core.asgi.application"
 ROOT_URLCONF = "archivebox.core.urls"
 
+LOGIN_URL = "/accounts/login/"
 LOGOUT_REDIRECT_URL = CONFIG.LOGOUT_REDIRECT_URL
 
 PASSWORD_RESET_URL = "/accounts/password_reset/"
@@ -151,16 +152,9 @@ try:
                 "email": CONFIG.LDAP_EMAIL_ATTR,
             }
 
-            # Use custom LDAP backend that supports LDAP_CREATE_SUPERUSER
-            # Include allauth backend first if allauth is installed
-            try:
-                import allauth as _allauth_check  # noqa: F401
-
-                _allauth_backend = ["allauth.account.auth_backends.AuthenticationBackend"]
-            except ImportError:
-                _allauth_backend = []
-
-            AUTHENTICATION_BACKENDS = _allauth_backend + [
+            # Use custom LDAP backend that supports LDAP_CREATE_SUPERUSER.
+            # The allauth block below prepends its backend when explicitly enabled.
+            AUTHENTICATION_BACKENDS = [
                 "archivebox.ldap.auth.ArchiveBoxLDAPBackend",
                 "django.contrib.auth.backends.RemoteUserBackend",
                 "django.contrib.auth.backends.ModelBackend",
@@ -184,10 +178,16 @@ except ImportError:
 
 ################################################################################
 ### django-allauth Configuration
-# Conditionally loaded if django-allauth is installed
+# Installing an optional dependency must not silently change authentication.
+# ALLAUTH_ENABLED is the single switch for apps, middleware, backend, and routes.
 ################################################################################
-try:
-    import allauth  # noqa: F401
+ALLAUTH_ENABLED = CONFIG.ALLAUTH_ENABLED
+
+if ALLAUTH_ENABLED:
+    try:
+        import allauth  # noqa: F401
+    except ImportError as err:
+        raise ImportError("ALLAUTH_ENABLED=True requires the archivebox[allauth] optional dependency") from err
 
     INSTALLED_APPS += [
         "archivebox.auth",
@@ -213,7 +213,7 @@ try:
         "allauth.account.middleware.AccountMiddleware",
     ]
 
-    # Prepend allauth backend to the list (only if not already present, e.g., from LDAP block)
+    # Prepend allauth so its email authentication runs before Django's username backend.
     _allauth_auth_backend = "allauth.account.auth_backends.AuthenticationBackend"
     if _allauth_auth_backend not in AUTHENTICATION_BACKENDS:
         AUTHENTICATION_BACKENDS = [_allauth_auth_backend] + AUTHENTICATION_BACKENDS
@@ -239,13 +239,9 @@ try:
     # e.g. SOCIALACCOUNT_PROVIDERS='{"google": {"APP": {"client_id": "...", "secret": "..."}}}'
     SOCIALACCOUNT_PROVIDERS = CONFIG.SOCIALACCOUNT_PROVIDERS
 
-    LOGIN_URL = "/accounts/login/"
     LOGIN_REDIRECT_URL = "/admin/"
     ACCOUNT_LOGOUT_REDIRECT_URL = LOGOUT_REDIRECT_URL or "/"
 
-except ImportError:
-    pass
-
 ################################################################################
 ### Staticfile and Template Settings
 ################################################################################

archivebox/core/urls.py

@@ -27,6 +27,17 @@ from archivebox.opencode.views import opencode_proxy_view
 from archivebox.progressmonitor.views import live_progress_view
 from archivebox.search.views import public_snapshot_search_stream_view
 
+if settings.ALLAUTH_ENABLED:
+    account_urlpatterns = [path("accounts/", include("allauth.urls"))]
+else:
+    # Preserve ArchiveBox's built-in auth surface unless allauth was explicitly
+    # enabled. Optional package installation alone must not change login behavior.
+    account_urlpatterns = [
+        path("accounts/login/", RedirectView.as_view(url="/admin/login/", query_string=True)),
+        path("accounts/logout/", RedirectView.as_view(url="/admin/logout/", query_string=True)),
+        path("accounts/", include("django.contrib.auth.urls")),
+    ]
+
 urlpatterns = [
     re_path(r"^static/(?P<path>.*)$", serve_static),
     path("robots.txt", static.serve, {"document_root": CONSTANTS.STATIC_DIR, "path": "robots.txt"}),
@@ -61,12 +72,8 @@ urlpatterns = [
     path("admin/core/snapshot/add/", RedirectView.as_view(url="/add/")),
     path("admin/core/snapshot/replay-auth/", SnapshotReplayAuthView.as_view(), name="snapshot-replay-auth"),
     path("add/", AddView.as_view(), name="add"),
-    # ``query_string=True`` preserves the ``?next=…`` param that Django's
-    # auth/login mixins append, so e.g. ``UserPassesTestMixin`` redirecting
-    # an unauthenticated ``/add`` visitor to ``/accounts/login/?next=/add/``
-    # carries the ``next`` through to ``/admin/login/`` and lands them at
-    # ``/add/`` after login instead of the admin homepage.
-    path("accounts/", include("allauth.urls")),
+    # The disabled-mode login redirect preserves Django's ``?next=…`` query.
+    *account_urlpatterns,
     path("progress.json", live_progress_view, name="live_progress"),
     path("admin/", archivebox_admin.urls),
     path("api/", include("archivebox.api.urls"), name="api"),

archivebox/templates/admin/login.html

@@ -56,9 +56,11 @@
       <label>&nbsp;</label><input type="submit" value="{% trans 'Create admin and continue' %}">
     </div>
   </form>
+  {% if allauth_enabled %}
   <p><a href="/accounts/login/?next={{ next|default:'/admin/'|urlencode }}">Use a configured identity provider instead</a></p>
+  {% endif %}
 </div>
-{% else %}
+{% elif allauth_enabled %}
 <script>
   // Keep Django's destination while handing authentication to allauth.
   var next = "{{ next|default:"/admin/"|escapejs }}";
@@ -67,5 +69,43 @@
 <noscript>
   <p><a href="/accounts/login/?next={{ next|default:'/admin/'|urlencode }}">Click here to log in</a></p>
 </noscript>
+{% else %}
+{% if form.errors and not form.non_field_errors %}
+<p class="errornote">
+{% if form.errors.items|length == 1 %}{% trans "Please correct the error below." %}{% else %}{% trans "Please correct the errors below." %}{% endif %}
+</p>
+{% endif %}
+
+{% if form.non_field_errors %}
+{% for error in form.non_field_errors %}
+<p class="errornote">{{ error }}</p>
+{% endfor %}
+{% endif %}
+
+<div id="content-main">
+  {% if user.is_authenticated %}
+  <p class="errornote">
+  {% blocktrans trimmed %}
+      You are authenticated as {{ username }}, but are not authorized to
+      access this page. Would you like to login to a different account?
+  {% endblocktrans %}
+  </p>
+  {% endif %}
+
+  <form action="{{ app_path }}" method="post" id="login-form">{% csrf_token %}
+    <div class="form-row" style="gap: 0;">
+      {{ form.username.errors }}
+      {{ form.username.label_tag }} {{ form.username }}
+    </div>
+    <div class="form-row" style="gap: 0;">
+      {{ form.password.errors }}
+      {{ form.password.label_tag }} {{ form.password }}
+      <input type="hidden" name="next" value="{{ next }}">
+    </div>
+    <div class="submit-row" style="border: none;">
+      <label>&nbsp;</label><input type="submit" value="{% trans 'Log in' %}">
+    </div>
+  </form>
+</div>
 {% endif %}
 {% endblock %}

archivebox/tests/conftest.py

@@ -1,26 +1,27 @@
 """archivebox/tests/conftest.py - Pytest fixtures for CLI tests."""
 
-import os
+import ctypes
 import json
+import os
 import re
 import secrets
-import signal
 import select
+import shutil
+import signal
 import socket
 import subprocess
 import sys
 import tempfile
 import textwrap
 import time
-import shutil
-import ctypes
+from collections.abc import Callable
 from datetime import timedelta
 from http.server import BaseHTTPRequestHandler, ThreadingHTTPServer
+from importlib.util import find_spec
 from pathlib import Path
 from threading import Event, Thread
 from types import SimpleNamespace
 from typing import Any
-from collections.abc import Callable
 
 import psutil
 import pytest
@@ -35,6 +36,11 @@ SESSION_DATA_DIR = Path(
 
 os.environ["ARCHIVEBOX_PYTEST_SESSION_DATA_DIR"] = str(SESSION_DATA_DIR)
 os.environ["DATA_DIR"] = str(SESSION_DATA_DIR)
+# When the optional dependency is present, exercise the opt-in auth stack in
+# the normal test process. The disabled-by-default contract is verified in a
+# clean subprocess by test_allauth_config.py.
+if find_spec("allauth") is not None:
+    os.environ.setdefault("ALLAUTH_ENABLED", "true")
 (SESSION_DATA_DIR / "tests").mkdir(parents=True, exist_ok=True)
 os.chdir(SESSION_DATA_DIR)
 os.environ.setdefault("DJANGO_SETTINGS_MODULE", "archivebox.core.settings")
@@ -1153,6 +1159,7 @@ def run_python_cwd(
         input=script,
         capture_output=True,
         text=True,
+        check=False,
         cwd=cwd,
         env=base_env,
         timeout=timeout,
@@ -1355,9 +1362,10 @@ def get_http_response(
 
 
 def make_latest_schedule_due(cwd: Path) -> None:
+    from django.utils import timezone
+
     from archivebox.crawls.models import Crawl, CrawlSchedule
     from archivebox.tests.test_orm_helpers import use_archivebox_db
-    from django.utils import timezone
 
     with use_archivebox_db(cwd):
         schedule = CrawlSchedule.objects.order_by("-created_at").select_related("template").first()
@@ -1653,6 +1661,7 @@ def resolve_abxpkg_binary_env(
         command,
         capture_output=True,
         text=True,
+        check=False,
         env=command_env,
     )
     assert result.returncode == 0, result.stderr or result.stdout

archivebox/tests/test_allauth_config.py

@@ -34,18 +34,15 @@ import json
 django.setup()
 
 from django.conf import settings
-from django.urls import Resolver404, resolve
+from django.urls import resolve
 
-try:
-    resolve('/accounts/login/')
-    accounts_route_enabled = True
-except Resolver404:
-    accounts_route_enabled = False
+login_match = resolve('/accounts/login/')
+route_uses_allauth = login_match.url_name == 'account_login'
 
 print(json.dumps({
     'app_enabled': 'allauth.account' in settings.INSTALLED_APPS,
     'backend_enabled': 'allauth.account.auth_backends.AuthenticationBackend' in settings.AUTHENTICATION_BACKENDS,
-    'route_enabled': accounts_route_enabled,
+    'route_uses_allauth': route_uses_allauth,
 }))
 """,
         ],
@@ -56,7 +53,7 @@ print(json.dumps({
         text=True,
     )
     state = json.loads(result.stdout.splitlines()[-1])
-    assert state == {"app_enabled": False, "backend_enabled": False, "route_enabled": False}
+    assert state == {"app_enabled": False, "backend_enabled": False, "route_uses_allauth": False}
 
 
 def test_allauth_config_from_env(monkeypatch):

archivebox/tests/test_allauth_integration.py

@@ -43,13 +43,11 @@ def test_admin_login_preserves_first_admin_setup(client):
 
 @pytest.mark.django_db
 def test_email_password_login_works(client, django_user_model, monkeypatch):
-    # Disable subdomain routing so AdminCookieIsolationMiddleware does not
-    # strip the session cookie from responses issued to the test client's
-    # default "testserver" host.
     from archivebox.config.common import get_config
+    from archivebox.core.routes_util import get_admin_host
 
     config = get_config()
-    monkeypatch.setattr(type(config), "USES_SUBDOMAIN_ROUTING", property(lambda self: False))
+    admin_host = get_admin_host(config=config)
 
     user = django_user_model.objects.create_user(
         username="logintest",
@@ -70,6 +68,7 @@ def test_email_password_login_works(client, django_user_model, monkeypatch):
             "login": "logintest@example.com",
             "password": "testpassword123",
         },
+        HTTP_HOST=admin_host,
         follow=True,
     )
     assert response.status_code == 200
@@ -78,6 +77,9 @@ def test_email_password_login_works(client, django_user_model, monkeypatch):
 
 @pytest.mark.django_db
 def test_registration_mode_approval_creates_inactive_user(client, settings, monkeypatch):
+    from archivebox.config.common import get_config
+    from archivebox.core.routes_util import get_admin_host
+
     settings.ACCOUNT_ADAPTER = "archivebox.auth.adapters.ArchiveBoxAccountAdapter"
     import archivebox.auth.adapters as m
 
@@ -90,6 +92,7 @@ def test_registration_mode_approval_creates_inactive_user(client, settings, monk
             "password1": "ComplexPass999!",
             "password2": "ComplexPass999!",
         },
+        HTTP_HOST=get_admin_host(config=get_config()),
     )
     from django.contrib.auth import get_user_model