Multiple Return Values about latexify_py HOT 5 CLOSED

I think math formula is basically not designed for multiple return values, but sometimes they are represented as a tuple, i.e., direct product of some single values.

In Python, returning multiple values like return a, b, c is just the same as returning a tuple return (a, b, c) and it reasonably corresponds to the math formula. In this case, we need not to add any special treatment because both looks already the same.

Python also supports decomposition of lists (e.g., x, y, z = [a, b, c] works fine), but this form can not be reasonably represented by type annotations when the inner values have different types, except using Union or Any.
So we should assume that such functions are just functions returning a list of something, not multiple values.

from latexify_py.

For lists of the same type values, e.g., int, it is represented as List[int] by the type annotation. In this case it is also better to assume that the function is returning a list.

from latexify_py.

Yes I know it returns tuple but the latex output writes something like <_ast.Tuple object at 0x********(or _ast.List, depending on which type is transferred to the return statement).
Now, latexify works like this:

def func(a,b):
  return (a+b,a-b)

print(get_latex(func))
#\operatorname{func}(a, b)\triangleq <_ast.Tuple object at 0x000002B53785A388>

What I want is something like:

def func(a, b):
  return (a + b, a - b)

print(get_latex(func))
#\operatorname{func}(a, b)\triangleq \left( a + b\space,\space a - b\right)

I'm thinking of implementing the logic of this part.

from latexify_py.

The reason is our parser does not implement visit_Tuple currently so all tuple expressions are not parsed for now (and it is the same for list and dict). As I am stopping to push everything (#17) it is welcome if you add it. I think implementing tuple and list are straightforward.

from latexify_py.

That's right, I'm just looking for an agreement of opinion for this. It will be very embarrassing if I implement it, then other contributors think it useless. Good luck!

from latexify_py.