fix: address PR INCF#93 review feedback - move RRF import to top, remove ID overwrite, restore local fuzzy search

Author: Zohaib

backend/agents.py

@@ -10,6 +10,7 @@
 
 from ks_search_tool import general_search, general_search_async, global_fuzzy_keyword_search
 from retrieval import get_retriever
+from rrf import reciprocal_rank_fusion
 
 #  LLM (Gemini) client setup 
 try:
@@ -425,8 +426,6 @@ async def execute_search(state: AgentState) -> Dict[str, Any]:
     return {"ks_results": all_ks_results, "vector_results": vec_results}
 
 
-from rrf import reciprocal_rank_fusion
-
 def fuse_results(state: AgentState) -> AgentState:
     print("--- Node: Result Fusion (RRF) ---")
     ks_results = state.get("ks_results", [])

backend/ks_search_tool.py

@@ -93,30 +93,27 @@ def search_across_all_fields(query: str, all_configs: dict, threshold: float = 0
 
 def global_fuzzy_keyword_search(keywords: Iterable[str], top_k: int = 20) -> List[dict]:
     """
-    Keyword search utilizing the public API with explicitly extracted keywords.
+    For each keyword, run search_across_all_fields across all datasources_config and combine unique hits.
     """
-    if not keywords:
-        return []
-        
-    query_str = " OR ".join([f'"{kw}"' if ' ' in kw else kw for kw in keywords if kw])
-    if not query_str:
-        return []
-        
-    try:
-        # Search via public API using the combined keyword OR-query string
-        res = general_search(query_str, top_k=top_k, enrich_details=True)
-        out = res.get("combined_results", [])
-        
-        # Differentiate IDs for RRF matching, though RRF
-        # also naturally merges duplicate links/IDs.
-        for i, item in enumerate(out):
-            item["_id"] = f"fuzzy_{i}"
-            item["id"] = f"fuzzy_{i}"
-            
-        return out
-    except Exception as e:
-        print(f"  -> Error in global_fuzzy_keyword_search: {e}")
+    config_path = "datasources_config.json"
+    if not os.path.exists(config_path):
         return []
+    with open(config_path, "r", encoding="utf-8") as fh:
+        all_configs = json.load(fh)
+    out: List[dict] = []
+    seen = set()
+    for kw in keywords or []:
+        if not kw:
+            continue
+        results = search_across_all_fields(kw, all_configs, threshold=0.8)
+        for r in results:
+            rid = r.get("_id") or r.get("id")
+            if rid and rid not in seen:
+                seen.add(rid)
+                out.append(r)
+        if len(out) >= top_k:
+            break
+    return out[:top_k]
 
 
 def extract_datasource_info_from_link(link: str) -> tuple: