Created by: liym27

PR types:

New features and Bug fixes

PR changes:

dygraph to static graph

Describe:

1. Add function convert_ifelse to run the transformed code dynamically:

• All if/else and ifExpr stmts are transformed whether it depends on Tensor or not;
• But in run-time, the code will be run in different ways (1) run paddle op like cond; (2) run python if/else. How it runs depends on whether it depends on Tensor.

2. Fix bug:

cast var dtype to bool in function _run_paddle_logical_XX

3. Code optimization:

3.1 Add LogicalTransformer : move LogicalTransformer out from LoopTransformer and IfElseTransformer, and run LogicalTransformer separately.

3.2 Remove unnecessary class: MergeAssignTransformer, NodeTestTransformer and IfConditionVisitor in ifelse_transformer.py

4. Comment and modify code that doesn't support return statement in unittest.