fix(search-service): Pass structured_answer through SearchOutput

Add structured_answer field to SearchOutput creation to ensure citations
and structured output data flows through the search pipeline.

**Changes:**

1. **SearchOutput Creation** (search_service.py:586):
   - Add `structured_answer=result_context.structured_answer` to SearchOutput
   - Ensures structured output (with citations) is included in search results
   - Previously: Field existed in schema but not populated from result_context

2. **Debug Logging**:
   - Log document_metadata count before SearchOutput creation
   - Log first document name for debugging
   - Helps track data flow through search pipeline

**Why This Matters:**

- SearchOutput schema has `structured_answer: StructuredAnswer | None` field
- generation_stage.py creates structured output and adds to result_context
- But SearchService wasn't passing it through to SearchOutput
- Result: Structured output generated but lost before returning to caller

**Data Flow:**

```
generation_stage.py
    ↓
result_context.structured_answer = StructuredAnswer(...)
    ↓
SearchService._search_with_executor()
    ↓
SearchOutput(
    answer=...,
    documents=...,
    structured_answer=result_context.structured_answer  ← ADDED
)
    ↓
MessageProcessingOrchestrator
    ↓
Frontend (citations display)
```

**Testing:**
- Structured output now included in SearchOutput
- Citations data flows through to conversation API response
- No breaking changes (field is optional, None if not generated)

**Dependencies:**
- Requires PR #626 (Structured Output schema) for StructuredAnswer field definition
- Works with PR #631 (Conversation API config) to enable user-controlled structured output

**Related:**
- Part of Issue #629 fix (citations not displaying)
- Small but critical piece of the structured output pipeline

Author: manavgup

backend/rag_solution/services/search_service.py

@@ -569,6 +569,11 @@ async def _search_with_pipeline(self, search_input: SearchInput) -> SearchOutput
         # Convert cot_output from ChainOfThoughtOutput to dict if present
         cot_output_dict = result_context.cot_output.model_dump() if result_context.cot_output else None
 
+        # Debug: Log document_metadata before creating SearchOutput
+        logger.info("📊 SEARCH_SERVICE: result_context.document_metadata has %d items", len(result_context.document_metadata))
+        if result_context.document_metadata:
+            logger.info("📊 SEARCH_SERVICE: First doc_metadata = %s", result_context.document_metadata[0].document_name if hasattr(result_context.document_metadata[0], 'document_name') else 'NO DOCUMENT_NAME')
+
         search_output = SearchOutput(
             answer=cleaned_answer,
             documents=result_context.document_metadata,
@@ -578,6 +583,7 @@ async def _search_with_pipeline(self, search_input: SearchInput) -> SearchOutput
             execution_time=result_context.execution_time,
             cot_output=cot_output_dict,
             token_warning=result_context.token_warning,
+            structured_answer=result_context.structured_answer,
             metadata={
                 "pipeline_architecture": "v2_stage_based",
                 "stages_executed": executor.get_stage_names(),