SeLFiE: Should I have only one `EVAL_PATH` instead of `EVAL_INNER_PATH` and `EVAL_OUTER_PATH`

[yutakang]

signature TEST = sig type test_type; end; structure Test = struct datatype test_type = One | Two; end; Test.One; structure Test2 = Test:TEST; (Test2.One;)

functor mk_Test (T:TEST) = T; structure Test3 = mk_Test(Test); (Test3.One;)

[yutakang]

Probably I should have only one signature for Eval_Path and use it for both Eval_Inner_Path and Eval_Outer_Path.

The good news is that we do not have to expose the tables (path-to-(inner/outer)node-table and print-to-(inner/outer)-nodes-table).

So, we can have an ML functor from_Table_to_Eval_Path: PATH_TO_UNODE_TABLE -> EVAL_PATH.

[yutakang]

So, it is going to be something like

• structure Eval_Inner_Path = from_Table_to_Eval_Path (Inner_Table)
• structure Eval_Inner_Parameter = from_Path_To_Parameter (Eval_Inner_Path)`

[yutakang]

This means I have to have constructs for outer-nodes even when I deal with inner-nodes. Use Utils.try'.

[yutakang]

Essentially, the only difference between Inner and Outer should be

• the path-to-node-table used in from_Table_to_Eval_Path and from_Variable_To_Quantifier,
• the print-to-paths-table used in from_Variable_to_Eval_Quantifier and from_Variable_To_Quantifier.

I think we can produce Quantifier_Domain from a path-to-node-table. i.e.

• structure Quantifier_Inner_Domain = from_Table_To_Quantifier_Domain (Inner_Table)
• structure Quantifier_Outer_Domain = from_Table_To_Quantifier_Domain (Outer_Table)

which means we probably do not have to have these functors independently.

[yutakang]

Done.