Merge pull request #1253 from wangkuiyi/python_learning_and_refactor

Rename Python function DataBase into create_data_config_proto

python/paddle/trainer/config_parser.py

@@ -893,11 +893,11 @@ class MaxOut(Cfg):
         self.add_keys(locals())
 
 
-def DataBase(async_load_data=False,
-             constant_slots=None,
-             data_ratio=1,
-             is_main_data=True,
-             usage_ratio=None):
+def create_data_config_proto(async_load_data=False,
+                             constant_slots=None,
+                             data_ratio=1,
+                             is_main_data=True,
+                             usage_ratio=None):
     # default: all sub dataproviders are treat as "main data".
     # see proto/DataConfig.proto for is_main_data
     data_config = DataConfig()
@@ -923,7 +923,7 @@ def SimpleData(files=None,
                context_len=None,
                buffer_capacity=None,
                **xargs):
-    data_config = DataBase(**xargs)
+    data_config = create_data_config_proto(**xargs)
     data_config.type = 'simple'
     data_config.files = files
     data_config.feat_dim = feat_dim
@@ -945,7 +945,7 @@ def PyData(files=None,
            constant_slots=None,
            load_thread_num=None,
            **xargs):
-    data_config = DataBase(**xargs)
+    data_config = create_data_config_proto(**xargs)
     data_config.type = 'py'
     if load_data_module in g_py_module_name_list:
 
@@ -996,7 +996,7 @@ def ProtoData(files=None,
               constant_slots=None,
               load_thread_num=None,
               **xargs):
-    data_config = DataBase(**xargs)
+    data_config = create_data_config_proto(**xargs)
     if type is None:
         data_config.type = 'proto'
     else:
@@ -1035,7 +1035,7 @@ def Data(type,
          buffer_capacity=None,
          **xargs):
 
-    data_config = DataBase(**xargs)
+    data_config = create_data_config_proto(**xargs)
     data_config.type = type
     data_config.files = files
     data_config.feat_dim = feat_dim

python/paddle/trainer_config_helpers/data_sources.py

@@ -58,8 +58,8 @@ def define_py_data_source(file_list,
     :param obj: python object name. May be a function name if using
                 PyDataProviderWrapper.
     :type obj: basestring
-    :param args: The best practice is using dict to pass arguments into 
-                 DataProvider, and use :code:`@init_hook_wrapper` to 
+    :param args: The best practice is using dict to pass arguments into
+                 DataProvider, and use :code:`@init_hook_wrapper` to
                  receive arguments.
     :type args: string or picklable object
     :param async: Load Data asynchronously or not.
@@ -98,7 +98,7 @@ def define_py_data_sources(train_list,
     The annotation is almost the same as define_py_data_sources2, except that
     it can specific train_async and data_cls.
 
-    :param data_cls: 
+    :param data_cls:
     :param train_list: Train list name.
     :type train_list: basestring
     :param test_list: Test list name.
@@ -111,8 +111,8 @@ def define_py_data_sources(train_list,
                 a tuple or list to this argument.
     :type obj: basestring or tuple or list
     :param args: The best practice is using dict() to pass arguments into
-                 DataProvider, and use :code:`@init_hook_wrapper` to receive 
-                 arguments. If train and test is different, then pass a tuple 
+                 DataProvider, and use :code:`@init_hook_wrapper` to receive
+                 arguments. If train and test is different, then pass a tuple
                  or list to this argument.
     :type args: string or picklable object or list or tuple.
     :param train_async: Is training data load asynchronously or not.
@@ -163,12 +163,12 @@ def define_py_data_sources2(train_list, test_list, module, obj, args=None):
 
     ..  code-block:: python
 
-        define_py_data_sources2(train_list="train.list", 
-                                test_list="test.list", 
+        define_py_data_sources2(train_list="train.list",
+                                test_list="test.list",
                                 module="data_provider"
                                 # if train/test use different configurations,
                                 # obj=["process_train", "process_test"]
-                                obj="process", 
+                                obj="process",
                                 args={"dictionary": dict_name})
 
     The related data provider can refer to :ref:`api_pydataprovider2_sequential_model` .
@@ -185,8 +185,8 @@ def define_py_data_sources2(train_list, test_list, module, obj, args=None):
                 a tuple or list to this argument.
     :type obj: basestring or tuple or list
     :param args: The best practice is using dict() to pass arguments into
-                 DataProvider, and use :code:`@init_hook_wrapper` to receive 
-                 arguments. If train and test is different, then pass a tuple 
+                 DataProvider, and use :code:`@init_hook_wrapper` to receive
+                 arguments. If train and test is different, then pass a tuple
                  or list to this argument.
     :type args: string or picklable object or list or tuple.
     :return: None
@@ -195,7 +195,7 @@ def define_py_data_sources2(train_list, test_list, module, obj, args=None):
 
     def py_data2(files, load_data_module, load_data_object, load_data_args,
                  **kwargs):
-        data = DataBase()
+        data = create_data_config_proto()
         data.type = 'py2'
         data.files = files
         data.load_data_module = load_data_module