Merge branch 'main' into fix/page-break-on-promptstudio

.gitignore

@@ -636,6 +636,7 @@ docker/*.env
 !docker/sample*.env
 docker/public_tools.json
 docker/proxy_overrides.yaml
+docker/compose.override.yaml
 docker/workflow_data/
 
 # Tool development

.pre-commit-config.yaml

@@ -80,7 +80,7 @@ repos:
             --settings-path=pyproject.toml,
           ]
   - repo: https://github.com/hadialqattan/pycln
-    rev: v2.4.0
+    rev: v2.5.0
     hooks:
       - id: pycln
         args: [--config=pyproject.toml]

backend/README.md

@@ -39,13 +39,21 @@ pdm install --dev -G lint
 pdm install --prod --no-editable
 ```
 
+#### Running scripts
+
 PDM allows you to run scripts applicable within the service dir.
 
 ```bash
 # List the possible scripts that can be executed
 pdm run -l
 ```
 
+For example to run the backend (dev mode is recommended to take advantage of gunicorn's `reload` feature)
+
+```bash
+pdm run backend --dev
+```
+
 #### Running commands
 
 - If you plan to run the django server locally, make sure the dependent services are up (either locally or through docker compose)

backend/api_v2/api_deployment_views.py

@@ -4,7 +4,7 @@
 
 from api_v2.constants import ApiExecution
 from api_v2.deployment_helper import DeploymentHelper
-from api_v2.exceptions import InvalidAPIRequest, NoActiveAPIKeyError
+from api_v2.exceptions import NoActiveAPIKeyError
 from api_v2.models import APIDeployment
 from api_v2.postman_collection.dto import PostmanCollection
 from api_v2.serializers import (
@@ -47,16 +47,14 @@ def initialize_request(
     def post(
         self, request: Request, org_name: str, api_name: str, api: APIDeployment
     ) -> Response:
-        file_objs = request.FILES.getlist(ApiExecution.FILES_FORM_DATA)
         serializer = ExecutionRequestSerializer(data=request.data)
         serializer.is_valid(raise_exception=True)
+        file_objs = serializer.validated_data.get(ApiExecution.FILES_FORM_DATA)
         timeout = serializer.validated_data.get(ApiExecution.TIMEOUT_FORM_DATA)
         include_metadata = serializer.validated_data.get(ApiExecution.INCLUDE_METADATA)
         include_metrics = serializer.validated_data.get(ApiExecution.INCLUDE_METRICS)
         use_file_history = serializer.validated_data.get(ApiExecution.USE_FILE_HISTORY)
         tag_names = serializer.validated_data.get(ApiExecution.TAGS)
-        if not file_objs or len(file_objs) == 0:
-            raise InvalidAPIRequest("File shouldn't be empty")
         response = DeploymentHelper.execute_workflow(
             organization_name=org_name,
             api=api,

backend/api_v2/key_helper.py

@@ -4,6 +4,7 @@
 from api_v2.exceptions import UnauthorizedKey
 from api_v2.models import APIDeployment, APIKey
 from api_v2.serializers import APIKeySerializer
+from django.core.exceptions import ValidationError
 from pipeline_v2.models import Pipeline
 from rest_framework.request import Request
 from workflow_manager.workflow_v2.workflow_helper import WorkflowHelper
@@ -29,7 +30,7 @@ def validate_api_key(
             api_key_instance: APIKey = APIKey.objects.get(api_key=api_key)
             if not KeyHelper.has_access(api_key_instance, instance):
                 raise UnauthorizedKey()
-        except APIKey.DoesNotExist:
+        except (APIKey.DoesNotExist, ValidationError):
             raise UnauthorizedKey()
 
     @staticmethod

backend/api_v2/serializers.py

@@ -9,8 +9,10 @@
 from rest_framework.serializers import (
     BooleanField,
     CharField,
+    FileField,
     IntegerField,
     JSONField,
+    ListField,
     ModelSerializer,
     Serializer,
     ValidationError,
@@ -115,12 +117,31 @@ class ExecutionRequestSerializer(TagParamsSerializer):
             e.g:'tag1,tag2-name,tag3_name'
     """
 
+    MAX_FILES_ALLOWED = 32
+
     timeout = IntegerField(
         min_value=-1, max_value=ApiExecution.MAXIMUM_TIMEOUT_IN_SEC, default=-1
     )
     include_metadata = BooleanField(default=False)
     include_metrics = BooleanField(default=False)
     use_file_history = BooleanField(default=False)
+    files = ListField(
+        child=FileField(),
+        required=True,
+        allow_empty=False,
+        error_messages={
+            "required": "At least one file must be provided.",
+            "empty": "The file list cannot be empty.",
+        },
+    )
+
+    def validate_files(self, value):
+        """Validate the files field."""
+        if len(value) == 0:
+            raise ValidationError("The file list cannot be empty.")
+        if len(value) > self.MAX_FILES_ALLOWED:
+            raise ValidationError(f"Maximum '{self.MAX_FILES_ALLOWED}' files allowed.")
+        return value
 
 
 class ExecutionQuerySerializer(Serializer):

backend/backend/public_urls_v2.py

@@ -20,9 +20,6 @@
 from django.conf.urls.static import static
 from django.urls import include, path
 
-path_prefix = settings.PATH_PREFIX
-api_path_prefix = settings.API_DEPLOYMENT_PATH_PREFIX
-
 urlpatterns = [
     path("", include("account_v2.urls")),
     # Connector OAuth

backend/backend/settings/base.py

@@ -323,7 +323,7 @@ def get_required_setting(
     "middleware.cache_control.CacheControlMiddleware",
 ]
 
-TENANT_SUBFOLDER_PREFIX = f"/{PATH_PREFIX}/unstract"
+TENANT_SUBFOLDER_PREFIX = f"{PATH_PREFIX}/unstract"
 SHOW_PUBLIC_IF_NO_TENANT_FOUND = True
 
 TEMPLATES = [
@@ -435,6 +435,11 @@ def get_required_setting(
         "django_filters.rest_framework.DjangoFilterBackend",
         "rest_framework.filters.OrderingFilter",
     ],
+    # For API versioning
+    "DEFAULT_VERSIONING_CLASS": "rest_framework.versioning.URLPathVersioning",
+    "DEFAULT_VERSION": "v1",
+    "ALLOWED_VERSIONS": ["v1"],
+    "VERSION_PARAM": "version",
 }
 
 # These paths will work without authentication

backend/backend/wsgi.py

@@ -6,7 +6,9 @@
 https://docs.djangoproject.com/en/4.2/howto/deployment/wsgi/
 """
 
+import logging
 import os
+import time
 
 from django.conf import settings
 from django.core.wsgi import get_wsgi_application
@@ -15,11 +17,17 @@
 
 load_dotenv()
 
+logger = logging.getLogger(__name__)
+
 os.environ.setdefault(
     "DJANGO_SETTINGS_MODULE",
     os.environ.get("DJANGO_SETTINGS_MODULE", "backend.settings.dev"),
 )
 
+wsgi_start_time = time.perf_counter()
 django_app = get_wsgi_application()
+wsgi_init_elapsed = time.perf_counter() - wsgi_start_time
+logger.info(f"WSGI application initialized in {wsgi_init_elapsed:.3f} seconds")
+
 
 application = start_server(django_app, f"{settings.PATH_PREFIX}/socket")

backend/entrypoint.sh

@@ -1,19 +1,50 @@
 #!/usr/bin/env bash
 
-cmd=$1
-if [ "$cmd" = "migrate" ]; then
+show_help() {
+    echo "Usage: ./entrypoint.sh [OPTIONS]"
+    echo ""
+    echo "Options:"
+    echo "  --migrate        Perform database migrations before starting the server."
+    echo "  --dev            Run Gunicorn in development mode with --reload and reduced graceful timeout (5s)."
+    echo "  --help, -h       Show this help message and exit."
+}
+
+# Parse arguments
+migrate=false
+dev=false
+
+while [[ "$#" -gt 0 ]]; do
+    case $1 in
+        --migrate) migrate=true ;;
+        --dev) dev=true ;;
+        --help|-h) show_help; exit 0 ;;
+        *) echo "Unknown argument: $1"; exit 1 ;;
+    esac
+    shift
+done
+
+# Perform database migration if --migrate is provided
+if [ "$migrate" = true ]; then
     echo "Migration initiated"
     .venv/bin/python manage.py migrate
 fi
 
-# NOTE: Leaving below for reference incase required in the future
-# python manage.py runserver 0.0.0.0:8000 --insecure
-# NOTE updated socket threads
-.venv/bin/gunicorn \
-    --bind 0.0.0.0:8000 \
-    --workers 2 \
-    --threads 2 \
-    --log-level debug \
-    --timeout 600 \
-    --access-logfile - \
-    backend.wsgi:application
+# Configure Gunicorn based on --dev flag
+gunicorn_args=(
+    --bind 0.0.0.0:8000
+    --workers 2
+    --threads 2
+    --log-level debug
+    --timeout 600
+    --access-logfile -
+)
+
+if [ "$dev" = true ]; then
+    echo "Running in development mode"
+    gunicorn_args+=(--reload --graceful-timeout 5)
+else
+    echo "Running in production mode"
+fi
+
+# Start Gunicorn
+.venv/bin/gunicorn "${gunicorn_args[@]}" backend.wsgi:application

backend/file_management/constants.py

@@ -6,9 +6,3 @@ class FileInformationKey:
     FILE_UPLOAD_MAX_SIZE = 100 * 1024 * 1024
     FILE_UPLOAD_ALLOWED_EXT = ["pdf"]
     FILE_UPLOAD_ALLOWED_MIME = ["application/pdf"]
-
-
-class FileViewTypes:
-    ORIGINAL = "ORIGINAL"
-    EXTRACT = "EXTRACT"
-    SUMMARIZE = "SUMMARIZE"

backend/middleware/organization_middleware.py

@@ -6,8 +6,7 @@
 
 class OrganizationMiddleware(MiddlewareMixin):
     def process_request(self, request):
-        tenant_prefix = settings.TENANT_SUBFOLDER_PREFIX.rstrip("/") + "/"
-        pattern = rf"^{tenant_prefix}(?P<org_id>[^/]+)/"
+        pattern = r"^/api/(?P<version>v[12])/unstract/(?P<org_id>[^/]+)/"
 
         # Check if the URL matches the pattern with organization ID
         match = re.match(pattern, request.path)
@@ -17,12 +16,12 @@ def process_request(self, request):
                 re.match(path, request.path)
                 for path in settings.ORGANIZATION_MIDDLEWARE_WHITELISTED_PATHS
             ):
-                request.path_info = "/" + request.path_info
                 return
 
             org_id = match.group("org_id")
+            version = match.group("version")
             request.organization_id = org_id
-            new_path = re.sub(pattern, "/" + tenant_prefix, request.path_info)
+            new_path = re.sub(pattern, f"/api/{version}/unstract/", request.path_info)
             request.path_info = new_path
         else:
             request.organization_id = None