[Python] Improve hygiene around .read() on file-like objects

GitOrigin-RevId: 7af5e6d8b0c9c710127ece31c867674d6c713d9b

src/gretel_synthetics/batch.py

@@ -135,7 +135,7 @@ def load_validator_from_file(self):
         p = Path(self.checkpoint_dir) / "validator.p.gz"
         if p.exists():
             with gzip.open(p, "r") as fin:
-                self.validator = cloudpickle.loads(fin.read())
+                self.validator = cloudpickle.load(fin)
 
     def reset_gen_data(self):
         """Reset all objects that accumulate or track synthetic
@@ -180,7 +180,8 @@ def _create_batch_from_dir(batch_dir: str):
 
     if not (path / HEADER_FILE).is_file():  # pragma: no cover
         raise ValueError("missing headers")
-    headers = json.loads(open(path / HEADER_FILE).read())
+    with open(path / HEADER_FILE) as f:
+        headers = json.load(f)
 
     if not (path / CONFIG_FILE).is_file():  # pragma: no cover
         raise ValueError("missing model param file")

src/gretel_synthetics/config.py

@@ -349,7 +349,8 @@ def config_from_model_dir(model_dir: str) -> BaseConfig:
     package and will instantiate a TensorFlowConfig
     """
     params_file = Path(model_dir) / const.MODEL_PARAMS
-    params_dict = json.loads(open(params_file).read())
+    with open(params_file) as f:
+        params_dict = json.load(f)
     model_type = params_dict.pop(const.MODEL_TYPE, None)
 
     # swap out the checkpoint dir location for the currently

src/gretel_synthetics/tokenizers.py

@@ -232,13 +232,13 @@ def _load_delimiter_data(self):
         params_file = Path(self._model_dir) / self.settings_fname
 
         if not params_file.is_file():
-            model_params_dict = json.loads(
-                open(Path(self._model_dir) / const.MODEL_PARAMS).read()
-            )
+            with open(Path(self._model_dir) / const.MODEL_PARAMS) as f:
+                model_params_dict = json.load(f)
             self.field_delimiter = model_params_dict[FIELD_DELIM]
             self.field_delimiter_token = model_params_dict[FIELD_DELIM_TOKEN]
         else:
-            params_dict = json.loads(open(params_file).read())
+            with open(params_file) as f:
+                params_dict = json.load(f)
             self.field_delimiter = params_dict.get(FIELD_DELIM, None)
             self.field_delimiter_token = params_dict.get(FIELD_DELIM_TOKEN, None)