fix(security): address reported unsafe pickle.load usages (#2099)

qlib/contrib/rolling/ddgda.py

@@ -14,6 +14,7 @@ from qlib.model.meta.task import MetaTask
 from qlib.model.trainer import TrainerR
 from qlib.typehint import Literal
 from qlib.utils import init_instance_by_config
+from qlib.utils.pickle_utils import restricted_pickle_load
 from qlib.workflow import R
 from qlib.workflow.task.utils import replace_task_handler_with_cache
 
@@ -298,7 +299,7 @@ class DDGDA(Rolling):
         # but their task test segment are not aligned! It worked in my previous experiment.
         # So the misalignment will not affect the effectiveness of the method.
         with self._internal_data_path.open("rb") as f:
-            internal_data = pickle.load(f)
+            internal_data = restricted_pickle_load(f)
 
         md = MetaDatasetDS(exp_name=internal_data, **kwargs)
 
@@ -360,7 +361,7 @@ class DDGDA(Rolling):
         )
 
         with self._internal_data_path.open("rb") as f:
-            internal_data = pickle.load(f)
+            internal_data = restricted_pickle_load(f)
         mds = MetaDatasetDS(exp_name=internal_data, **kwargs)
 
         # 3) meta model make inference and get new qlib task

qlib/contrib/tuner/tuner.py

@@ -8,7 +8,6 @@ import os
 import yaml
 import json
 import copy
-import pickle
 import logging
 import importlib
 import subprocess
@@ -18,6 +17,7 @@ import numpy as np
 from abc import abstractmethod
 
 from ...log import get_module_logger, TimeInspector
+from ...utils.pickle_utils import restricted_pickle_load
 from hyperopt import fmin, tpe
 from hyperopt import STATUS_OK, STATUS_FAIL
 
@@ -136,7 +136,7 @@ class QLibTuner(Tuner):
         exp_result_dir = os.path.join(self.ex_dir, QLibTuner.EXP_RESULT_DIR.format(estimator_ex_id))
         exp_result_path = os.path.join(exp_result_dir, QLibTuner.EXP_RESULT_NAME)
         with open(exp_result_path, "rb") as fp:
-            analysis_df = pickle.load(fp)
+            analysis_df = restricted_pickle_load(fp)
 
         # 4. Get the backtest factor which user want to optimize, if user want to maximize the factor, then reverse the result
         res = analysis_df.loc[self.optim_config.report_type].loc[self.optim_config.report_factor]

qlib/data/cache.py

@@ -30,6 +30,7 @@ from ..utils import (
     normalize_cache_fields,
     normalize_cache_instruments,
 )
+from ..utils.pickle_utils import restricted_pickle_load
 
 from ..log import get_module_logger
 from .base import Feature
@@ -225,7 +226,7 @@ class CacheUtils:
             cache_path = Path(cache_path)
             meta_path = cache_path.with_suffix(".meta")
             with meta_path.open("rb") as f:
-                d = pickle.load(f)
+                d = restricted_pickle_load(f)
             with meta_path.open("wb") as f:
                 try:
                     d["meta"]["last_visit"] = str(time.time())
@@ -592,7 +593,7 @@ class DiskExpressionCache(ExpressionCache):
 
         with CacheUtils.writer_lock(self.r, f"{str(C.dpm.get_data_uri())}:expression-{cache_uri}"):
             with meta_path.open("rb") as f:
-                d = pickle.load(f)
+                d = restricted_pickle_load(f)
             instrument = d["info"]["instrument"]
             field = d["info"]["field"]
             freq = d["info"]["freq"]
@@ -959,7 +960,7 @@ class DiskDatasetCache(DatasetCache):
         im = DiskDatasetCache.IndexManager(cp_cache_uri)
         with CacheUtils.writer_lock(self.r, f"{str(C.dpm.get_data_uri())}:dataset-{cache_uri}"):
             with meta_path.open("rb") as f:
-                d = pickle.load(f)
+                d = restricted_pickle_load(f)
             instruments = d["info"]["instruments"]
             fields = d["info"]["fields"]
             freq = d["info"]["freq"]

qlib/data/dataset/loader.py

@@ -2,7 +2,6 @@
 # Licensed under the MIT License.
 
 import abc
-import pickle
 from pathlib import Path
 import warnings
 import pandas as pd
@@ -11,6 +10,7 @@ from typing import Tuple, Union, List, Dict
 
 from qlib.data import D
 from qlib.utils import load_dataset, init_instance_by_config, time_to_slc_point
+from qlib.utils.pickle_utils import restricted_pickle_load
 from qlib.log import get_module_logger
 from qlib.utils.serial import Serializable
 
@@ -283,7 +283,7 @@ class StaticDataLoader(DataLoader, Serializable):
                 self._data = pd.read_parquet(self._config, engine="pyarrow")
             else:
                 with Path(self._config).open("rb") as f:
-                    self._data = pickle.load(f)
+                    self._data = restricted_pickle_load(f)
         elif isinstance(self._config, pd.DataFrame):
             self._data = self._config
 

qlib/utils/mod.py

@@ -11,7 +11,6 @@ import contextlib
 import importlib
 import os
 from pathlib import Path
-import pickle
 import pkgutil
 import re
 import sys
@@ -20,6 +19,7 @@ from typing import Any, Dict, List, Tuple, Union
 from urllib.parse import urlparse
 
 from qlib.typehint import InstConf
+from qlib.utils.pickle_utils import restricted_pickle_load
 
 
 def get_module_by_module_path(module_path: Union[str, ModuleType]):
@@ -168,10 +168,10 @@ def init_instance_by_config(
 
                 pr_path = os.path.join(pr.netloc, path) if bool(pr.path) else pr.netloc
                 with open(os.path.normpath(pr_path), "rb") as f:
-                    return pickle.load(f)
+                    return restricted_pickle_load(f)
         else:
             with config.open("rb") as f:
-                return pickle.load(f)
+                return restricted_pickle_load(f)
 
     klass, cls_kwargs = get_callable_kwargs(config, default_module=default_module)
 

qlib/utils/objm.py

@@ -6,6 +6,7 @@ import tempfile
 from pathlib import Path
 
 from qlib.config import C
+from qlib.utils.pickle_utils import restricted_pickle_load
 
 
 class ObjManager:
@@ -116,7 +117,7 @@ class FileManager(ObjManager):
 
     def load_obj(self, name):
         with (self.path / name).open("rb") as f:
-            return pickle.load(f)
+            return restricted_pickle_load(f)
 
     def exists(self, name):
         return (self.path / name).exists()

qlib/workflow/task/manage.py

@@ -28,6 +28,7 @@ from tqdm.cli import tqdm
 
 from .utils import get_mongodb
 from ...config import C
+from ...utils.pickle_utils import restricted_pickle_loads
 
 
 class TaskManager:
@@ -131,7 +132,7 @@ class TaskManager:
         for prefix in self.ENCODE_FIELDS_PREFIX:
             for k in list(task.keys()):
                 if k.startswith(prefix):
-                    task[k] = pickle.loads(task[k])
+                    task[k] = restricted_pickle_loads(task[k])
         return task
 
     def _dict_to_str(self, flt):