gh-114920: Make iterable and subscriptable the Namespace object in argparse lib

[milanbalazs]

With this change the Namespace will be iterable and subscriptable.

Issue reference: #114920

For example:

import argparse

parser: argparse.ArgumentParser = argparse.ArgumentParser(description=__doc__)

parser.add_argument(
    "--test_1",
    action="store_true",
)

parser.add_argument(
    "--test_2",
    type=str,
)

parser.add_argument(
    "--test_3",
    type=int,
    default=0
)

args = parser.parse_args()
for x in args:
    print(x)

Test:

>>> python3 test_2.py
{'test_1': False}
{'test_2': None}
{'test_3': 0}

>>> python3 test_2.py --test_3 5 --test_1
{'test_1': True}
{'test_2': None}
{'test_3': 5}

Without my change the same behavior can be done only with tricky and ugly solutions, like:

args = parser.parse_args()

for x in vars(args):
    print({x: getattr(args, x)})

Furthermore, with the __getitem__ method implementation, the Namespace object is subscriptable, so it's possible to get the value directly from the object, like:

args = parser.parse_args()

print(args["test_1"])

• Issue: Make iterable (and subscriptable) the Namespace object in argparse lib #114920

[encukou]

The issue was closed, so I'll close the pull request as well.