Print diagnostics in each interactive loop

[hankluo6]

Fixed #2807
Fixed #2757

In interactive mode, some errors or suggestions were added to the diagnostics but did not throw any errors. These should also be printed.

[Shaikh-Ubaid]

I think this works and does the job:

(lp) ubaid@DESKTOP-TEN39M0:~/projects/lpython$ ./src/bin/lpython
Interactive LPython. Experimental prototype, not ready for end users.
LPython version: 0.22.0-53-g1ef4675ae
  * Use Ctrl-D to exit
  * Use Enter to submit
  * Use Alt-Enter or Ctrl-N to make a new line
    - Editing (Keys: Left, Right, Home, End, Backspace, Delete)
    - History (Keys: Up, Down)
>>> i: i32 = 5                                                                                                                                 1,11  ]
>>> from math import sin                                                                                                                       1,21  ]
>>> sin(i)                                                                                                                                     1,7   ]
semantic error: `x` argument of `Sin` must be real or complex
 --> input:1:5
  |
1 | sin(i)
  |     ^ 


Note: Please report unclear or confusing messages as bugs at
https://github.com/lcompilers/lpython/issues.
>>> pow(2, 3)                                                                                                                                  1,10  ]
style suggestion: Could have used '**' instead of 'pow'
 --> input:1:1
  |
1 | pow(2, 3)
  | ^^^^^^^^^ '**' could be used instead


Note: Please report unclear or confusing messages as bugs at
https://github.com/lcompilers/lpython/issues.
8
>>>     

[Shaikh-Ubaid]

LGTM. Thanks for fixing this!

[Shaikh-Ubaid]

On second thought, I think this approach is good to print style info/warnings. But for printing errors (semantic errors like in #2757), we should fix the python_ast_to_asr to return the result as not ok (that is res.ok = False).

[Shaikh-Ubaid]

I created a new issue for this here #2810. I think this PR is good to merge.

[hankluo6]

I noticed this as well. The reason is that for the intrinsic functions, we used append_error instead of throwing a semantic error. This could potentially be resolved by throwing the error later from python_ast_to_asr if the intrinsic function returns null.