feat: Add Zendesk connector integration

biome.json

@@ -1,5 +1,5 @@
 {
-	"$schema": "https://biomejs.dev/schemas/2.1.2/schema.json",
+	"$schema": "https://biomejs.dev/schemas/2.1.3/schema.json",
 	"vcs": {
 		"enabled": true,
 		"clientKind": "git",

docker-compose.override.yml

@@ -32,10 +32,4 @@ services:
       - NVIDIA_DRIVER_CAPABILITIES=compute,utility
       - LANGCHAIN_TRACING_V2=false
       - LANGSMITH_TRACING=false
-    deploy:
-      resources:
-        reservations:
-          devices:
-            - driver: nvidia
-              count: all
-              capabilities: [gpu]
+

surfsense_backend/alembic/versions/15_add_zendesk_connector_enums.py

@@ -0,0 +1,61 @@
+"""Add ZENDESK_CONNECTOR to enums
+
+Revision ID: 15
+Revises: 14
+"""
+
+from collections.abc import Sequence
+
+from alembic import op
+
+# revision identifiers, used by Alembic.
+revision: str = "15"
+down_revision: str | None = "14"
+branch_labels: str | Sequence[str] | None = None
+depends_on: str | Sequence[str] | None = None
+
+
+def upgrade() -> None:
+    """Safely add 'ZENDESK_CONNECTOR' to enum types if missing."""
+
+    # Add to searchsourceconnectortype enum
+    op.execute(
+        """
+    DO $$
+    BEGIN
+        IF NOT EXISTS (
+            SELECT 1 FROM pg_type t
+            JOIN pg_enum e ON t.oid = e.enumtypid
+            WHERE t.typname = 'searchsourceconnectortype' AND e.enumlabel = 'ZENDESK_CONNECTOR'
+        ) THEN
+            ALTER TYPE searchsourceconnectortype ADD VALUE 'ZENDESK_CONNECTOR';
+        END IF;
+    END
+    $$;
+    """
+    )
+
+    # Add to documenttype enum
+    op.execute(
+        """
+    DO $$
+    BEGIN
+        IF NOT EXISTS (
+            SELECT 1 FROM pg_type t
+            JOIN pg_enum e ON t.oid = e.enumtypid
+            WHERE t.typname = 'documenttype' AND e.enumlabel = 'ZENDESK_CONNECTOR'
+        ) THEN
+            ALTER TYPE documenttype ADD VALUE 'ZENDESK_CONNECTOR';
+        END IF;
+    END
+    $$;
+    """
+    )
+
+
+def downgrade() -> None:
+    """
+    Downgrade logic not implemented since PostgreSQL
+    does not support removing enum values.
+    """
+    pass

surfsense_backend/app/connectors/zendesk_connector.py

@@ -0,0 +1,45 @@
+from datetime import datetime
+
+import httpx
+
+
+class ZendeskConnector:
+    def __init__(self, subdomain: str, email: str, api_token: str):
+        if not subdomain or not email or not api_token:
+            raise ValueError("Subdomain, email, and API token cannot be empty.")
+        self.base_url = f"https://{subdomain}.zendesk.com/api/v2"
+        self.auth = (f"{email}/token", api_token)
+
+    async def get_tickets(
+        self, start_date: str | None = None, end_date: str | None = None
+    ):
+        tickets = []
+        async with httpx.AsyncClient() as client:
+            params = {}
+            if start_date:
+                # Zendesk API uses 'start_time' for filtering by updated_at
+                # Convert YYYY-MM-DD to Unix timestamp
+                start_timestamp = int(
+                    datetime.strptime(start_date, "%Y-%m-%d").timestamp()
+                )
+                params["start_time"] = start_timestamp
+
+            url = f"{self.base_url}/tickets.json"
+            while url:
+                try:
+                    response = await client.get(url, auth=self.auth, params=params)
+                    response.raise_for_status()
+                    data = response.json()
+                    tickets.extend(data["tickets"])
+                    url = data["next_page"]
+                    # Clear params for subsequent paginated requests as they are part of the next_page URL
+                    params = {}
+                except httpx.HTTPStatusError as e:
+                    print(
+                        f"HTTP error occurred: {e.response.status_code} - {e.response.text}"
+                    )
+                    break
+                except httpx.RequestError as e:
+                    print(f"An error occurred while requesting {e.request.url}: {e}")
+                    break
+        return tickets

surfsense_backend/app/db.py

@@ -44,10 +44,11 @@ class DocumentType(str, Enum):
     DISCORD_CONNECTOR = "DISCORD_CONNECTOR"
     JIRA_CONNECTOR = "JIRA_CONNECTOR"
     CONFLUENCE_CONNECTOR = "CONFLUENCE_CONNECTOR"
+    ZENDESK_CONNECTOR = "ZENDESK_CONNECTOR"
 
 
 class SearchSourceConnectorType(str, Enum):
-    SERPER_API = "SERPER_API"  # NOT IMPLEMENTED YET : DON'T REMEMBER WHY : MOST PROBABLY BECAUSE WE NEED TO CRAWL THE RESULTS RETURNED BY IT
+    SERPER_API = "SERPER_API"  # NOT IMPLEMENTED YET : DON'T REMEMBER WHY : MOST PROBABLY BECAUSE WE NEED TO CRAWL THE RESULTS RETURNed BY IT
     TAVILY_API = "TAVILY_API"
     LINKUP_API = "LINKUP_API"
     SLACK_CONNECTOR = "SLACK_CONNECTOR"
@@ -57,6 +58,7 @@ class SearchSourceConnectorType(str, Enum):
     DISCORD_CONNECTOR = "DISCORD_CONNECTOR"
     JIRA_CONNECTOR = "JIRA_CONNECTOR"
     CONFLUENCE_CONNECTOR = "CONFLUENCE_CONNECTOR"
+    ZENDESK_CONNECTOR = "ZENDESK_CONNECTOR"
 
 
 class ChatType(str, Enum):

surfsense_backend/app/routes/search_source_connectors_routes.py

@@ -43,6 +43,7 @@
     index_linear_issues,
     index_notion_pages,
     index_slack_messages,
+    index_zendesk_tickets,
 )
 from app.users import current_active_user
 from app.utils.check_ownership import check_ownership
@@ -488,6 +489,21 @@ async def index_connector_content(
             )
             response_message = "Discord indexing started in the background."
 
+        elif connector.connector_type == SearchSourceConnectorType.ZENDESK_CONNECTOR:
+            # Run indexing in background
+            logger.info(
+                f"Triggering Zendesk indexing for connector {connector_id} into search space {search_space_id} from {indexing_from} to {indexing_to}"
+            )
+            background_tasks.add_task(
+                run_zendesk_indexing_with_new_session,
+                connector_id,
+                search_space_id,
+                str(user.id),
+                indexing_from,
+                indexing_to,
+            )
+            response_message = "Zendesk indexing started in the background."
+
         else:
             raise HTTPException(
                 status_code=400,
@@ -960,3 +976,60 @@ async def run_confluence_indexing(
             exc_info=True,
         )
         # Optionally update status in DB to indicate failure
+
+
+async def run_zendesk_indexing_with_new_session(
+    connector_id: int,
+    search_space_id: int,
+    user_id: str,
+    start_date: str,
+    end_date: str,
+):
+    """Wrapper to run Zendesk indexing with its own database session."""
+    logger.info(
+        f"Background task started: Indexing Zendesk connector {connector_id} into space {search_space_id} from {start_date} to {end_date}"
+    )
+    async with async_session_maker() as session:
+        await run_zendesk_indexing(
+            session, connector_id, search_space_id, user_id, start_date, end_date
+        )
+    logger.info(f"Background task finished: Indexing Zendesk connector {connector_id}")
+
+
+async def run_zendesk_indexing(
+    session: AsyncSession,
+    connector_id: int,
+    search_space_id: int,
+    user_id: str,
+    start_date: str,
+    end_date: str,
+):
+    """Runs the Zendesk indexing task and updates the timestamp."""
+    try:
+        indexed_count, error_message = await index_zendesk_tickets(
+            session,
+            connector_id,
+            search_space_id,
+            user_id,
+            start_date,
+            end_date,
+            update_last_indexed=False,
+        )
+        if error_message:
+            logger.error(
+                f"Zendesk indexing failed for connector {connector_id}: {error_message}"
+            )
+            # Optionally update status in DB to indicate failure
+        else:
+            logger.info(
+                f"Zendesk indexing successful for connector {connector_id}. Indexed {indexed_count} documents."
+            )
+            # Update the last indexed timestamp only on success
+            await update_connector_last_indexed(session, connector_id)
+            await session.commit()  # Commit timestamp update
+    except Exception as e:
+        logger.error(
+            f"Critical error in run_zendesk_indexing for connector {connector_id}: {e}",
+            exc_info=True,
+        )
+        # Optionally update status in DB to indicate failure

surfsense_backend/app/schemas/search_source_connector.py

@@ -167,6 +167,26 @@ def validate_config_for_connector_type(
             if not config.get("CONFLUENCE_BASE_URL"):
                 raise ValueError("CONFLUENCE_BASE_URL cannot be empty")
 
+        elif connector_type == SearchSourceConnectorType.ZENDESK_CONNECTOR:
+            # For ZENDESK_CONNECTOR, require ZENDESK_SUBDOMAIN, ZENDESK_EMAIL, and ZENDESK_API_TOKEN
+            allowed_keys = ["ZENDESK_SUBDOMAIN", "ZENDESK_EMAIL", "ZENDESK_API_TOKEN"]
+            if set(config.keys()) != set(allowed_keys):
+                raise ValueError(
+                    f"For ZENDESK_CONNECTOR connector type, config must only contain these keys: {allowed_keys}"
+                )
+
+            # Ensure the subdomain is not empty
+            if not config.get("ZENDESK_SUBDOMAIN"):
+                raise ValueError("ZENDESK_SUBDOMAIN cannot be empty")
+
+            # Ensure the email is not empty
+            if not config.get("ZENDESK_EMAIL"):
+                raise ValueError("ZENDESK_EMAIL cannot be empty")
+
+            # Ensure the API token is not empty
+            if not config.get("ZENDESK_API_TOKEN"):
+                raise ValueError("ZENDESK_API_TOKEN cannot be empty")
+
         return config