ASR -> CPython: Add visitors for ListConstant, TupleConstant & SetConstant

[kmr-srbh]

from lpython import i32, f64

my_first_list: list[i32] = [1, 2, 3 , 4]
print(my_first_list)

my_second_list: list[str] = ["a", "b", "c", "d"]
print(my_second_list)

my_third_list: list[list[i32]] = [[1, 2], [3, 4], [5, 6]]
print(my_third_list)

my_fourth_list: list[list[f64]] = [[1.0, 2.2], [3.6, 4.9], [5.1, 6.3]]
print(my_fourth_list)

my_first_tuple: tuple[i32, str, f64] = (1, "hello", 2.4)
print(my_first_tuple)

my_second_tuple: tuple[tuple[i32, str], str] = ((1, "hello"), "world")
print(my_second_tuple)

my_first_set: set[i32] = {1, 2, 3, 2, 4}
print(my_first_set)

my_second_set: set[f64] = {1.1, 2.5, 6.8}
print(my_second_set)

my_third_set: set[str] = {"a", "b", "a", "c"}
print(my_third_set)

(base) saurabh-kumar@Awadh:~/Projects/System/lpython$ ./src/bin/lpython ./examples/example.py --show-python
def __main__global_init():
    my_first_list = [1, 2, 3, 4]
    my_second_list = ["a", "b", "c", "d"]
    my_third_list = [[1, 2], [3, 4], [5, 6]]
    my_fourth_list = [[1.000000, 2.200000], [3.600000, 4.900000], [5.100000, 6.300000]]
    my_first_tuple = (1, "hello", 2.400000)
    my_second_tuple = ((1, "hello"), "world")
    my_first_set = {1, 2, 3, 2, 4}
    my_second_set = {1.100000, 2.500000, 6.800000}
    my_third_set = {"a", "b", "a", "c"}
def __main__global_stmts():
    print(my_first_list)
    print(my_second_list)
    print(my_third_list)
    print(my_fourth_list)
    print(my_first_tuple)
    print(my_second_tuple)
    print(my_first_set)
    print(my_second_set)
    print(my_third_set)

TODO

• Add tests

[kmr-srbh]

@Shaikh-Ubaid what are your thoughts on this, is this the right way to do it? Could you please guide me on adding tests for this?

[nikabot]

Regarding tests, I think you enable tests for python in tests.toml and search for list1.py and add python=true.

[nikabot]

And please make sure the generated python code and tests matches the same line by line. I mean, manually verify the output is the same as the test code.

[Shaikh-Ubaid]

Regarding tests, I think you enable tests for python in tests.toml and search for list1.py and add python=true.

Yes, let's just add one reference test that includes all three list, tuple, set. In the long term, we need an integration_tests approach for testing this.

[kmr-srbh]

@Shaikh-Ubaid @nikabot since the visitors will be similar for 3 of 4 data-structures, shall I implement them and test them together? I think it would be fast this way. For DictConstant, we can have the separate PR.

[Shaikh-Ubaid]

@Shaikh-Ubaid @nikabot since the visitors will be similar for 3 of 4 data-structures, shall I implement them and test them together? I think it would be fast this way. For DictConstant, we can have the separate PR.

Yes.

[kmr-srbh]

@Shaikh-Ubaid @nikabot what can be a better name for the test file?

[Shaikh-Ubaid]

what can be a better name for the test file?

test_aggregate_constants or something similar.

[kmr-srbh]

Generated code verified using CPython interpreter v3.10.13.

[Shaikh-Ubaid]

Generated code verified using CPython interpreter v3.10.13.

@kmr-srbh the generated code does not work with CPython. There are couple things we will need to fix for it to work.

• call to __main__global_init() , __main__global_stmts() needs to be added at the end.
• ensure the variables defined in __main__global_init() are accessbile to __main__global_stmts().
• and any other issues that we come across while fixing the above two

The above should be tackled in different PR and should not be part of this PR.

[kmr-srbh]

Generated code verified using CPython interpreter v3.10.13.

@kmr-srbh the generated code does not work with CPython. There are couple things we will need to fix for it to work.

* call to `__main__global_init()` , `__main__global_stmts()` needs to be added at the end.

* ensure the variables defined in `__main__global_init()` are accessbile to `__main__global_stmts()`.

* and any other issues that we come across while fixing the above two

The above should be tackled in different PR and should not be part of this PR.

You are right @Shaikh-Ubaid. I had planned to look into this to get the CPython back-end working. A separate PR will be good! 👍

[kmr-srbh]

Generated code verified using CPython interpreter v3.10.13.

I tested the code for syntax errors using this.