Refactor article retrieval to use article titles (urls)

Updated article retrieval logic to fetch actual article details from the database instead of using placeholder titles.
Now when we execute hybrid and hybrid_hyde workflows, we get the article links (with titles and not just article ids).

Author: Zafarhussain87

src/lex_llm/api/connectors/lex_db_connector.py

@@ -95,14 +95,20 @@ async def hybrid_search(
                         # Append additional chunks to the same article
                         articles_dict[result.article_id] += f"\n\n{result.chunk_text}"
 
-                return [
-                    LexArticle(
-                        id=article_id,
-                        title=f"Article {article_id}",  # We don't have the title in RetrievalResult
-                        text=text,
-                        url=f"https://lex.dk/article/{article_id}",
-                    )
-                    for article_id, text in articles_dict.items()
+                # Fetch actual article details from the database
+                search_results = lexdb_api.get_articles(
+                    ids=str(list(articles_dict.keys()))
+                )
+
+                if search_results.entries:
+                    return [
+                        LexArticle(
+                            id=result.id,
+                            title=result.title,
+                            text=result.xhtml_md,
+                            url=result.url,
+                        )
+                    for result in search_results.entries
                 ]
 
             return []
@@ -143,4 +149,4 @@ async def hyde_search(
             return []
         except httpx.RequestError as e:
             print(f"Error connecting to LexDB: {e}")
-            return []
+            return []