* replace part of the old implementation, test=develop

* restore concat op, test=develop

* update all ops implemention & delete GetDataTypeOfVar func, test=develop

* refine ctc align op with padding 
* refine api sample code

* test=develop support Tensor input for ctc_align_op

* test=develop add some comment

test=develop

Do not use ctor

* Reduce line of codes.
* We can use virtual function for Maker now.
* The implementation does not care what maker holds, it is easier to
refactor later.

1. Allocate memory for output before compute.
2. Rename 'ctc_decode' to 'ctc_align'

1. Remove 'top 1'(or argmax) from CPU and GPU kernel
2. Add a new test case
3. Refine doc