While easy to create, it is crazy that python doesn't provide mutable primitives by default (AFAIK).
This package provides some simple python primitive types in a mutable shell:
BoolFloatIntStr
The safest usage is to always use .set() and .get():
from mutable_primitives import Int
x = Int(5)
def make_x_seven():
x.set(7)
make_x_seven()
print(x.get()) # should print 7
However if you understand the limitations, you can do some normal operations:
from mutable_primitives import Int
x = Int(5)
print(x + 4) # prints 9 (technically Int(9))
print(4 + x) # prints 9 (technically int(9))
assert x == 5
assert 5 == x
TODO some invalid/bad/dangerous use cases
Q: This whole library is unnecessary.
A: That's a statement.
Q: Why make a library when you can do this in a few lines when needed?
A: Having a library just makes it more uniform and clear what's happening.
Q: There are 4 competing libraries for this functionality, why add another?
A: There are now 5 competing libraries.
Q: Why write out so many repetitive functions when you could just inherit?
A: So that test coverage can be sure that it's run. Also the code is generated and committed, so it's not extra dev effort.
Q: Why not subclass int/float/etc?
A: You can't subclass bool, and subclassing the other primitives brings complexity.
In rough order of preference:
- Better README.md
- Ensure correct type used with
.set(), e.g.x = Int(5); x.set('bad')should error - Add thread-safe mutables classes