Fix regulation_filter bypass for chunks without regulation_code

Chunks without a regulation_code were silently passing through the filter
in _scan_rag(), causing unrelated documents (e.g. Data Act, legal templates)
to be included in filtered generation jobs. Now chunks without reg_code are
skipped when regulation_filter is active.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

backend-compliance/compliance/services/control_generator.py

@@ -806,7 +806,9 @@ class ControlGeneratorPipeline:
                                 or payload.get("source_code", ""))
 
                     # Filter by regulation_code if configured
-                    if config.regulation_filter and reg_code:
+                    if config.regulation_filter:
+                        if not reg_code:
+                            continue  # Skip chunks without regulation code
                         code_lower = reg_code.lower()
                         if not any(code_lower.startswith(f.lower()) for f in config.regulation_filter):
                             continue
@@ -852,10 +854,16 @@ class ControlGeneratorPipeline:
                 collection, collection_total, collection_new,
             )
 
-        logger.info(
-            "RAG scroll complete: %d total unique seen, %d new unprocessed to process",
-            len(seen_hashes), len(all_results),
-        )
+        if config.regulation_filter:
+            logger.info(
+                "RAG scroll complete: %d total unique seen, %d passed regulation_filter %s",
+                len(seen_hashes), len(all_results), config.regulation_filter,
+            )
+        else:
+            logger.info(
+                "RAG scroll complete: %d total unique seen, %d new unprocessed to process",
+                len(seen_hashes), len(all_results),
+            )
         return all_results
 
     def _get_processed_hashes(self, hashes: list[str]) -> set[str]: