Discourse Edge Case (#1685)

backend/danswer/connectors/discourse/connector.py

@@ -101,7 +101,11 @@ class DiscourseConnector(PollConnector):
             if end and end < last_time_dt:
                 continue
 
-            if valid_categories and topic.get("category_id") not in valid_categories:
+            if (
+                self.categories
+                and valid_categories
+                and topic.get("category_id") not in valid_categories
+            ):
                 continue
 
             topic_ids.append(topic["id"])
@@ -138,10 +142,16 @@ class DiscourseConnector(PollConnector):
             sections.append(
                 Section(link=topic_url, text=parse_html_page_basic(post["cooked"]))
             )
+        category_name = self.category_id_map.get(topic["category_id"])
+
+        metadata: dict[str, str | list[str]] = (
+            {
+                "category": category_name,
+            }
+            if category_name
+            else {}
+        )
 
-        metadata: dict[str, str | list[str]] = {
-            "category": self.category_id_map[topic["category_id"]],
-        }
         if topic.get("tags"):
             metadata["tags"] = topic["tags"]
 

backend/danswer/db/tag.py

@@ -91,9 +91,10 @@ def create_or_add_document_tag_list(
             new_tags.append(new_tag)
             existing_tag_values.add(tag_value)
 
-    logger.debug(
-        f"Created new tags: {', '.join([f'{tag.tag_key}:{tag.tag_value}' for tag in new_tags])}"
-    )
+    if new_tags:
+        logger.debug(
+            f"Created new tags: {', '.join([f'{tag.tag_key}:{tag.tag_value}' for tag in new_tags])}"
+        )
 
     all_tags = existing_tags + new_tags