require initialization of structs and arrays at decl site #747
Conversation
g-r-a-n-t
changed the title
g-r-a-n-t
changed the title
g-r-a-n-t
force-pushed
the
var-init-check
branch
from
532cc32
to
6252820
Compare
g-r-a-n-t
force-pushed
the
var-init-check
branch
2 times, most recently
from
b909d29
to
363b32b
Compare
| @@ -264,7 +264,7 @@ note: | |||
| ┌─ aug_assign.fe:65:5 | |||
| │ | |||
| 65 │ ╭ pub fn add_from_mem(a: u256, b: u256) -> u256 { | |||
| 66 │ │ let my_array: Array<u256, 10> | |||
| 66 │ │ let my_array: Array<u256, 10> = [0; 10] | |||
There was a problem hiding this comment.
Adding this repeat expression feels a bit like a crutch to me. It doesn't seem overly useful for the user since it only allows a single value to be repeated. Did you consider to just initialize the array with default values on behalf of the user. Maybe that gets complicated if the array is over a complex type such as Array<Customer, 10>. However, in the future we could have a Default trait and if a user wants to initialize an array of Array<Customer, 10> then it will automatically use the values provided by Customer::default() or else error if Customer doesn't implement the Default trait. I'm just thinking out loud here...I'm not really against introducing the repeat expression but if in the future we find a way to get rid of it again or make it more useful I wouldn't be sad to see it go 😄
There was a problem hiding this comment.
Ok, I just learned that one can use the same approach to initialize fixed size arrays in Rust. So, I guess it's fine. I still think that in the future it could be nice syntactical sugar if one could just write let x: Array<Customer, 10> and it would de-sugar into something like let x: Array<Customer, 10> = [Customer::default(), 10]. And if T does not implement the Default trait it would cause a compile error.
What was wrong?
see #731
How was it fixed?
Added array repeat expressions and checks for initialization of structs and arrays.
To-Do