As the title, add Python helper function to get block Inputs/Outputs. This function is helpful to distributed computing for sharding.

* use PYTHON_C_API in dygraph, test=develop

* set static shape for shape tensor with constant

* remove debug code

* fix typo

* add ut

* refine code

* refine example

[Dy2Static]Convert var.shape stmt and Convert the return variables of Tensor-dependent 'if' staments to Tensor if it not  (#24911)

* Support int and long: int or long -> six.integer_types. 

* Modify test_tensor_shape: fix bug and modify comment. 

* Support convert_var_shape to convert var.shape stmt

* Modify code in ifelse_simple_func.py because don't support return non-Tensor in Tensor-dependent 'if' stament currently. 

* Convert the return variables of Tensor-dependent 'if' staments to Tensor if it not. test=develop

* Compatible int32 and int64 for attr in op slice/strided_slice. test=develop

* Polish code in nn.py  test=develop

* Fix bug: set the same dtype for the inputs of elementwise_add. test=develop

* Convert int32 to int64 in slice op to avoid data overflow. test=develop

* Convert int32 to int64 in strided_slice_op to avoid data overflow. test=develop

* add support for value tensor support of fill_constant Op

* support to modify dict and list in while_loop test=develop

* polish code test=develop

* refine code, test=develop

* reuse contain_var, test=develop