cnhkmcp 2.1.1__tar.gz → 2.1.2__tar.gz

{cnhkmcp-2.1.1/cnhkmcp.egg-info → cnhkmcp-2.1.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cnhkmcp
-Version: 2.1.1
+Version: 2.1.2
 Summary: A comprehensive Model Context Protocol (MCP) server for quantitative trading platform integration
 Home-page: https://github.com/cnhk/cnhkmcp
 Author: CNHK

{cnhkmcp-2.1.1/cnhkmcp/untracked/APP/Tranformer → cnhkmcp-2.1.2/cnhkmcp/untracked/AI/346/241/214/351/235/242/346/217/222/344/273/266/get_knowledgeBase_tool}/helpful_functions.py

@@ -122,7 +122,7 @@ def save_simulation_result(result: dict) -> None:
 
     os.makedirs(folder_path, exist_ok=True)
 
-    with open(file_path, "w") as file:
+    with open(file_path, "w", encoding="utf-8") as file:
         json.dump(result, file)
 
 

{cnhkmcp-2.1.1 → cnhkmcp-2.1.2}/cnhkmcp/untracked/AI/346/241/214/351/235/242/346/217/222/344/273/266/process_knowledge_base.py

@@ -16,17 +16,18 @@ if TOOL_DIR not in sys.path:
 sys.path.insert(0, TOOL_DIR)
 import ace_lib
 from fetch_all_operators import fetch_operators, prompt_credentials
-from fetch_all_datasets import (
-    fetch_all_combinations,
-    fetch_datasets_for_combo,
-    merge_and_deduplicate,
-)
 from fetch_all_documentation import (
     fetch_tutorials,
     fetch_tutorial_pages,
     fetch_page,
     _extract_page_id,
 )
+# Dataset fetching currently disabled per request
+# from fetch_all_datasets import (
+#     fetch_all_combinations,
+#     fetch_datasets_for_combo,
+#     merge_and_deduplicate,
+# )
 
 
 def ensure_knowledge_dir():
@@ -107,62 +108,63 @@ def process_operators(session: ace_lib.SingleSession, knowledge_dir: str):
         print(f"✓ Created {filename} with {len(category_list)} operators")
 
 
-def process_datasets(session: ace_lib.SingleSession, dataset_dir: str):
-    """Fetch datasets and save one JSON per region."""
-    print("=== Processing Datasets ===")
-
-    print("Fetching valid instrument/region/delay/universe combinations...")
-    options_df = fetch_all_combinations(session)
-    if options_df is None or options_df.empty:
-        print("No simulation options fetched; aborting dataset fetch.")
-        return
-
-    all_datasets: list[pd.DataFrame] = []
-    combo_idx = 0
-
-    for _, row in options_df.iterrows():
-        instrument_type = row.get("InstrumentType")
-        region = row.get("Region")
-        delay = row.get("Delay")
-        universes = row.get("Universe") or []
-
-        for universe in universes:
-            combo_idx += 1
-            print(f"[{combo_idx}] {instrument_type} / {region} / D{delay} / {universe}")
-            try:
-                df = fetch_datasets_for_combo(session, instrument_type, region, delay, universe)
-                print(f"    -> {len(df)} rows")
-                all_datasets.append(df)
-            except Exception as exc:
-                print(f"    -> Failed: {exc}")
-
-    if not all_datasets:
-        print("No datasets fetched; nothing to save.")
-        return
-
-    combined_df = pd.concat([df for df in all_datasets if not df.empty], ignore_index=True)
-    if combined_df.empty:
-        print("No datasets fetched; nothing to save.")
-        return
-
-    regions = sorted(combined_df["param_region"].dropna().unique())
-    print(f"Found regions: {', '.join(regions)}")
-
-    for region in regions:
-        region_df = combined_df[combined_df["param_region"] == region]
-        region_unique = merge_and_deduplicate([region_df])
-
-        region_list = []
-        for _, row in region_unique.iterrows():
-            record = {col: to_jsonable(row[col]) for col in row.index}
-            region_list.append(record)
-
-        filename = f"{region.replace(' ', '_').lower()}_datasets.json"
-        filepath = os.path.join(dataset_dir, filename)
-        with open(filepath, "w", encoding="utf-8") as f:
-            json.dump(region_list, f, ensure_ascii=False, indent=2)
-
-        print(f"✓ Created {filename} with {len(region_list)} datasets")
+    # Dataset fetching intentionally disabled; keep for potential re-enable.
+    # def process_datasets(session: ace_lib.SingleSession, dataset_dir: str):
+    #     """Fetch datasets and save one JSON per region."""
+    #     print("=== Processing Datasets ===")
+    #
+    #     print("Fetching valid instrument/region/delay/universe combinations...")
+    #     options_df = fetch_all_combinations(session)
+    #     if options_df is None or options_df.empty:
+    #         print("No simulation options fetched; aborting dataset fetch.")
+    #         return
+    #
+    #     all_datasets: list[pd.DataFrame] = []
+    #     combo_idx = 0
+    #
+    #     for _, row in options_df.iterrows():
+    #         instrument_type = row.get("InstrumentType")
+    #         region = row.get("Region")
+    #         delay = row.get("Delay")
+    #         universes = row.get("Universe") or []
+    #
+    #         for universe in universes:
+    #             combo_idx += 1
+    #             print(f"[{combo_idx}] {instrument_type} / {region} / D{delay} / {universe}")
+    #             try:
+    #                 df = fetch_datasets_for_combo(session, instrument_type, region, delay, universe)
+    #                 print(f"    -> {len(df)} rows")
+    #                 all_datasets.append(df)
+    #             except Exception as exc:
+    #                 print(f"    -> Failed: {exc}")
+    #
+    #     if not all_datasets:
+    #         print("No datasets fetched; nothing to save.")
+    #         return
+    #
+    #     combined_df = pd.concat([df for df in all_datasets if not df.empty], ignore_index=True)
+    #     if combined_df.empty:
+    #         print("No datasets fetched; nothing to save.")
+    #         return
+    #
+    #     regions = sorted(combined_df["param_region"].dropna().unique())
+    #     print(f"Found regions: {', '.join(regions)}")
+    #
+    #     for region in regions:
+    #         region_df = combined_df[combined_df["param_region"] == region]
+    #         region_unique = merge_and_deduplicate([region_df])
+    #
+    #         region_list = []
+    #         for _, row in region_unique.iterrows():
+    #             record = {col: to_jsonable(row[col]) for col in row.index}
+    #             region_list.append(record)
+    #
+    #         filename = f"{region.replace(' ', '_').lower()}_datasets.json"
+    #         filepath = os.path.join(dataset_dir, filename)
+    #         with open(filepath, "w", encoding="utf-8") as f:
+    #             json.dump(region_list, f, ensure_ascii=False, indent=2)
+    #
+    #         print(f"✓ Created {filename} with {len(region_list)} datasets")
 
 
 def process_documentation(session: ace_lib.SingleSession, knowledge_dir: str):
@@ -240,7 +242,7 @@ def main():
     
     # Ensure knowledge directory exists
     knowledge_dir = ensure_knowledge_dir()
-    dataset_dir = knowledge_dir  # Save datasets directly under knowledge
+    # dataset_dir = knowledge_dir  # Save datasets directly under knowledge (disabled)
     print(f"Knowledge directory: {knowledge_dir}\n")
 
     # Process documentation (tutorials/pages)
@@ -262,16 +264,15 @@ def main():
         traceback.print_exc()
         return
 
-    # Process datasets by region
-    print("\nStarting dataset processing...\n")
-    try:
-        process_datasets(session, dataset_dir)
-    except Exception as exc:
-        print(f"✗ Failed to process datasets: {exc}")
-        import traceback
-        traceback.print_exc()
-        return
-
+    # Dataset processing disabled; re-enable by uncommenting the block below.
+    # print("\nStarting dataset processing...\n")
+    # try:
+    #     process_datasets(session, dataset_dir)
+    # except Exception as exc:
+    #     print(f"✗ Failed to process datasets: {exc}")
+    #     import traceback
+    #     traceback.print_exc()
+    #     return
 
     print("\n=== Processing Complete ===")
 

{cnhkmcp-2.1.1 → cnhkmcp-2.1.2}/cnhkmcp/untracked/AI/346/241/214/351/235/242/346/217/222/344/273/266/rag_engine.py

@@ -18,6 +18,7 @@ PREFERRED_MODELS = [
 # Final chosen model will be detected at runtime from supported list
 MODEL_NAME = None
 COLLECTION_NAME = "brain_kb_v5"
+BATCH_SIZE = 128  # batch upserts to avoid huge single writes
 
 # Optional imports for different file types
 try:
@@ -37,6 +38,8 @@ class KnowledgeBase:
         self.meta_path = os.path.join(self.db_path, "_meta.json")
         self.manifest_path = os.path.join(self.db_path, "_manifest.json")
         self._collection_reset_guard = False
+        self._query_reset_guard = False
+        self._sync_lock = threading.Lock()
         
         if not os.path.exists(self.kb_path):
             os.makedirs(self.kb_path)
@@ -83,6 +86,7 @@ class KnowledgeBase:
         
         # Initialize Vector DB
         self._init_collection()
+        self._healthcheck()
         
         # Initial sync
         self.sync_knowledge()
@@ -119,6 +123,20 @@ class KnowledgeBase:
         except Exception:
             pass  # Metadata failure should not block runtime
 
+    def _healthcheck(self):
+        """Validate index readability right after startup; rebuild if corrupted."""
+        try:
+            _ = self.collection.count()
+        except Exception as e:
+            msg = str(e).lower()
+            if any(x in msg for x in ["hnsw", "segment", "compaction", "backfill"]):
+                print("Detected index corruption on startup. Rebuilding vector_db...")
+                shutil.rmtree(self.db_path, ignore_errors=True)
+                self._init_collection(recreate=True)
+                self.sync_knowledge(allow_reset=False)
+            else:
+                print(f"Index healthcheck encountered an unexpected error: {e}")
+
     def _maybe_reset_for_incompatibility(self, chosen_model: str, embed_dim: int, chroma_version: str):
         """If existing index meta differs (model/dimension/chromadb), wipe it."""
         if not os.path.exists(self.db_path):
@@ -145,79 +163,88 @@ class KnowledgeBase:
 
     def sync_knowledge(self, allow_reset: bool = True):
         """Scans the knowledge folder and updates the vector database."""
+        if not self._sync_lock.acquire(blocking=False):
+            print("Sync already running, skip this trigger.")
+            return
+        
         print("Syncing knowledge base...")
         manifest = self._load_manifest()
         updated_manifest = {}
         supported_extensions = (".txt", ".md", ".pdf", ".docx", ".json")
         current_files = []
-        for filename in os.listdir(self.kb_path):
-            file_path = os.path.join(self.kb_path, filename)
-            if os.path.isfile(file_path) and filename.lower().endswith(supported_extensions):
-                current_files.append(filename)
-                mtime = os.path.getmtime(file_path)
-                size = os.path.getsize(file_path)
-                prev_meta = manifest.get(filename)
-                # Skip unchanged files
-                if prev_meta and prev_meta.get("mtime") == mtime and prev_meta.get("size") == size:
-                    updated_manifest[filename] = prev_meta
-                    continue
-                try:
-                    content = self._extract_text(file_path)
-                    if content:
-                        # Sliding window chunking on original text
-                        chunk_size = 400
-                        overlap = 80
-                        original_chunks = []
-                        for i in range(0, len(content), chunk_size - overlap):
-                            chunk = content[i:i + chunk_size].strip()
-                            if chunk:
-                                original_chunks.append(chunk)
-                        
-                        if original_chunks:
-                            # Normalize for embedding generation only (not for storage)
-                            normalized_chunks = [c.lower().replace('_', ' ') for c in original_chunks]
-                            
-                            ids = [f"{filename}_{i}" for i in range(len(original_chunks))]
-                            metadatas = [{"source": filename, "chunk": i} for i in range(len(original_chunks))]
-                            
-                            # Compute embeddings from normalized text
-                            embeddings = []
-                            for v in self.model.embed(normalized_chunks):
-                                try:
-                                    embeddings.append(v.tolist())
-                                except Exception:
-                                    embeddings.append(list(v))
+        try:
+            for filename in os.listdir(self.kb_path):
+                file_path = os.path.join(self.kb_path, filename)
+                if os.path.isfile(file_path) and filename.lower().endswith(supported_extensions):
+                    current_files.append(filename)
+                    mtime = os.path.getmtime(file_path)
+                    size = os.path.getsize(file_path)
+                    prev_meta = manifest.get(filename)
+                    # Skip unchanged files
+                    if prev_meta and prev_meta.get("mtime") == mtime and prev_meta.get("size") == size:
+                        updated_manifest[filename] = prev_meta
+                        continue
+                    try:
+                        content = self._extract_text(file_path)
+                        if content:
+                            # Sliding window chunking on original text
+                            chunk_size = 800
+                            overlap = 80
+                            original_chunks = []
+                            for i in range(0, len(content), chunk_size - overlap):
+                                chunk = content[i:i + chunk_size].strip()
+                                if chunk:
+                                    original_chunks.append(chunk)
                             
-                            # Store ORIGINAL text (not normalized) so users see the real content
-                            self.collection.upsert(
-                                documents=original_chunks,
-                                ids=ids,
-                                metadatas=metadatas,
-                                embeddings=embeddings
-                            )
-                            print(f"  ✓ Indexed {filename}: {len(original_chunks)} chunks")
-                            updated_manifest[filename] = {"mtime": mtime, "size": size}
+                            if original_chunks:
+                                # Normalize for embedding generation only (not for storage)
+                                normalized_chunks = [c.lower().replace('_', ' ') for c in original_chunks]
+                                
+                                ids = [f"{filename}_{i}" for i in range(len(original_chunks))]
+                                metadatas = [{"source": filename, "chunk": i} for i in range(len(original_chunks))]
+                                
+                                # Compute embeddings from normalized text
+                                embeddings = []
+                                for v in self.model.embed(normalized_chunks):
+                                    try:
+                                        embeddings.append(v.tolist())
+                                    except Exception:
+                                        embeddings.append(list(v))
+                                
+                                # Store ORIGINAL text (not normalized) so users see the real content
+                                for start in range(0, len(original_chunks), BATCH_SIZE):
+                                    end = start + BATCH_SIZE
+                                    self.collection.upsert(
+                                        documents=original_chunks[start:end],
+                                        ids=ids[start:end],
+                                        metadatas=metadatas[start:end],
+                                        embeddings=embeddings[start:end]
+                                    )
+                                print(f"  ✓ Indexed {filename}: {len(original_chunks)} chunks (batched)")
+                                updated_manifest[filename] = {"mtime": mtime, "size": size}
+                    except Exception as e:
+                        err_msg = str(e)
+                        print(f"Error processing {filename}: {err_msg}")
+                        # Auto-recover if HNSW/compaction/index errors occur
+                        if allow_reset and any(x in err_msg.lower() for x in ["hnsw", "compaction", "segment reader"]):
+                            if not self._collection_reset_guard:
+                                print("Detected index corruption. Rebuilding vector_db and retrying sync once...")
+                                self._collection_reset_guard = True
+                                self._init_collection(recreate=True)
+                                return self.sync_knowledge(allow_reset=False)
+            # Remove deleted files from the index
+            deleted_files = set(manifest.keys()) - set(current_files)
+            for filename in deleted_files:
+                try:
+                    self.collection.delete(where={"source": filename})
+                    print(f"  ✓ Removed deleted file from index: {filename}")
                 except Exception as e:
-                    err_msg = str(e)
-                    print(f"Error processing {filename}: {err_msg}")
-                    # Auto-recover if HNSW/compaction/index errors occur
-                    if allow_reset and any(x in err_msg.lower() for x in ["hnsw", "compaction", "segment reader"]):
-                        if not self._collection_reset_guard:
-                            print("Detected index corruption. Rebuilding vector_db and retrying sync once...")
-                            self._collection_reset_guard = True
-                            self._init_collection(recreate=True)
-                            return self.sync_knowledge(allow_reset=False)
-        # Remove deleted files from the index
-        deleted_files = set(manifest.keys()) - set(current_files)
-        for filename in deleted_files:
-            try:
-                self.collection.delete(where={"source": filename})
-                print(f"  ✓ Removed deleted file from index: {filename}")
-            except Exception as e:
-                print(f"  ! Failed to remove {filename}: {e}")
-        # Persist manifest
-        self._save_manifest(updated_manifest)
-        print("Knowledge base sync complete.")
+                    print(f"  ! Failed to remove {filename}: {e}")
+            # Persist manifest
+            self._save_manifest(updated_manifest)
+            print("Knowledge base sync complete.")
+        finally:
+            self._sync_lock.release()
 
     def _extract_text(self, file_path):
         ext = os.path.splitext(file_path)[1].lower()
@@ -249,7 +276,7 @@ class KnowledgeBase:
                 return json.dumps(data, ensure_ascii=False, indent=2)
         return None
 
-    def query(self, text, top_k=5, distance_threshold=0.8):
+    def query(self, text, top_k=5, distance_threshold=0.8, allow_reset: bool = True):
         """Retrieves relevant snippets from the knowledge base.
 
         Uses cosine distance (lower is better). A result is treated as a hit only
@@ -311,9 +338,25 @@ class KnowledgeBase:
             context = "\n---\n".join([h["text"] for h in hits]) if is_hit else ""
             return {"hit": is_hit, "context": context, "hits": hits}
         except Exception as e:
-            print(f"Query error: {e}")
+            err_msg = str(e)
+            print(f"Query error: {err_msg}")
             import traceback
             traceback.print_exc()
+
+            # Auto-recover if HNSW/compaction/backfill errors surface during query
+            if allow_reset and any(x in err_msg.lower() for x in ["hnsw", "compaction", "segment reader", "backfill"]):
+                if not self._query_reset_guard:
+                    print("Detected index corruption during query. Rebuilding vector_db and retrying once...")
+                    self._query_reset_guard = True
+                    try:
+                        self._init_collection(recreate=True)
+                        self.sync_knowledge(allow_reset=False)
+                        # Retry query once with guard disabled to avoid loops
+                        self._query_reset_guard = False
+                        return self.query(text, top_k=top_k, distance_threshold=distance_threshold, allow_reset=False)
+                    except Exception as inner_e:
+                        print(f"Auto-rebuild after query failure also failed: {inner_e}")
+                        self._query_reset_guard = False
             return {"hit": False, "context": "", "hits": []}
 
     def start_watcher(self):
@@ -343,14 +386,23 @@ class KBHandler(FileSystemEventHandler):
     def __init__(self, kb_instance):
         self.kb = kb_instance
         self.supported_extensions = (".txt", ".md", ".pdf", ".docx", ".json")
+        self._debounce_timer = None
+
+    def _trigger_sync(self):
+        def run():
+            self.kb.sync_knowledge()
+        if self._debounce_timer and self._debounce_timer.is_alive():
+            return
+        self._debounce_timer = threading.Timer(0.5, run)
+        self._debounce_timer.start()
 
     def on_modified(self, event):
         if not event.is_directory and event.src_path.lower().endswith(self.supported_extensions):
             print(f"File modified: {event.src_path}. Re-syncing...")
-            threading.Thread(target=self.kb.sync_knowledge).start()
+            self._trigger_sync()
 
     def on_created(self, event):
         if not event.is_directory and event.src_path.lower().endswith(self.supported_extensions):
             print(f"File created: {event.src_path}. Syncing...")
-            threading.Thread(target=self.kb.sync_knowledge).start()
+            self._trigger_sync()
 

{cnhkmcp-2.1.1 → cnhkmcp-2.1.2}/cnhkmcp/untracked/AI/346/241/214/351/235/242/346/217/222/344/273/266/requirements.txt

@@ -1,7 +1,7 @@
 openai
 Pillow
-fastembed>=0.3.4
-chromadb>=0.5.0
+fastembed>=0.3.6
+chromadb>=0.5.11
 watchdog
 pypdf
 python-docx

{cnhkmcp-2.1.1 → cnhkmcp-2.1.2}/cnhkmcp/untracked/APP/Tranformer/Transformer.py

@@ -4729,7 +4729,7 @@ async def main():
         input_str = input()
         if input_str == "":
             config_path = os.path.join(os.path.dirname(__file__), 'transformer_config.json')
-            with open(config_path, 'r') as f:
+            with open(config_path, 'r', encoding='utf-8') as f:
                 config = json.load(f)
             print("\n" + "="*60)
             print("✓ 已从 transformer_config.json 加载账号配置")

{cnhkmcp-2.1.1/cnhkmcp/untracked/APP/give_me_idea → cnhkmcp-2.1.2/cnhkmcp/untracked/APP/Tranformer}/helpful_functions.py

@@ -122,7 +122,7 @@ def save_simulation_result(result: dict) -> None:
 
     os.makedirs(folder_path, exist_ok=True)
 
-    with open(file_path, "w") as file:
+    with open(file_path, "w", encoding="utf-8") as file:
         json.dump(result, file)
 
 

{cnhkmcp-2.1.1/cnhkmcp/untracked/AI/346/241/214/351/235/242/346/217/222/344/273/266/get_knowledgeBase_tool → cnhkmcp-2.1.2/cnhkmcp/untracked/APP/give_me_idea}/helpful_functions.py

@@ -122,7 +122,7 @@ def save_simulation_result(result: dict) -> None:
 
     os.makedirs(folder_path, exist_ok=True)
 
-    with open(file_path, "w") as file:
+    with open(file_path, "w", encoding="utf-8") as file:
         json.dump(result, file)
 
 

{cnhkmcp-2.1.1 → cnhkmcp-2.1.2}/cnhkmcp/untracked/APP/helpful_functions.py

@@ -122,7 +122,7 @@ def save_simulation_result(result: dict) -> None:
 
     os.makedirs(folder_path, exist_ok=True)
 
-    with open(file_path, "w") as file:
+    with open(file_path, "w", encoding="utf-8") as file:
         json.dump(result, file)
 
 

{cnhkmcp-2.1.1 → cnhkmcp-2.1.2}/cnhkmcp/untracked/APP/hkSimulator/ace_lib.py

@@ -117,7 +117,7 @@ def get_credentials() -> tuple[str, str]:
     credentials_file_path = os.path.join(credentials_folder_path, "platform-brain.json")
 
     if Path(credentials_file_path).exists() and os.path.getsize(credentials_file_path) > 2:
-        with open(credentials_file_path) as file:
+        with open(credentials_file_path, encoding="utf-8") as file:
             data = json.loads(file.read())
     else:
         os.makedirs(credentials_folder_path, exist_ok=True)
@@ -128,7 +128,7 @@ def get_credentials() -> tuple[str, str]:
             email = input("Email:\n")
             password = getpass.getpass(prompt="Password:")
         data = {"email": email, "password": password}
-        with open(credentials_file_path, "w") as file:
+        with open(credentials_file_path, "w", encoding="utf-8") as file:
             json.dump(data, file)
     return (data["email"], data["password"])
 

{cnhkmcp-2.1.1 → cnhkmcp-2.1.2}/cnhkmcp/untracked/APP/hkSimulator/autosimulator.py

@@ -241,7 +241,7 @@ def get_user_json():
         json_path = re.sub(r'^["\']+|["\']+$', '', raw_path.strip())
         if os.path.exists(json_path):
             try:
-                with open(json_path, 'r') as f:
+                with open(json_path, 'r', encoding='utf-8') as f:
                     alpha_list = json.load(f)
                 # Check master log for previous progress
                 file_name = os.path.basename(json_path)
@@ -349,7 +349,7 @@ def simulation_worker(session, alpha_list, mode, json_path, location_path, retry
                 sent_count += len(batch)
                 update_master_log(json_path, sent_count - 1)
             locations[str(time.time())] = location
-            with open(location_path, 'w') as f:
+            with open(location_path, 'w', encoding='utf-8') as f:
                 json.dump(locations, f, indent=2)
             # Do NOT overwrite the input JSON file
             logger.info(f'Simulation sent, location(s) saved: {location}')
@@ -370,7 +370,7 @@ def result_worker(session, location_path, result_path, poll_interval=30):
         if not os.path.exists(location_path):
             time.sleep(poll_interval)
             continue
-        with open(location_path, 'r') as f:
+        with open(location_path, 'r', encoding='utf-8') as f:
             locations = json.load(f)
         for loc_key, loc_val in locations.items():
             if loc_key in results:
@@ -418,7 +418,7 @@ def result_worker(session, location_path, result_path, poll_interval=30):
                         alpha = session.get(f"https://api.worldquantbrain.com/alphas/{alpha_id}")
                         results[loc_key] = alpha.json()
                     logger.info(f"Result fetched for location {loc_val}")
-                with open(result_path, 'w') as f:
+                with open(result_path, 'w', encoding='utf-8') as f:
                     json.dump(results, f, indent=2)
             except Exception as e:
                 logger.error(f'Error fetching result for {loc_val}: {e}')