- Created
- 2 years ago
- Modified
- 2 years ago
- Author
~Luyten-Orion
- Status
- Todo
- Labels
#64Separate operator for concatenation
- Author
Issue title is basically it, but my suggestion is to use
+for arithmetic and&for concatenation operations, such as adding an item to a list or joining strings together, as in my opinion, it can reduce logic errors where someone is trying to treat user input as a number without first converting it
Milestone: 0.3.0.
note & is currently bitwise AND, so the same mistake can still be made. & is likely to become a prefix and postfix operator for reference related stuff (no issue open)
other suggestions are
++(alike zig) and..(lua and php iirc). adding a new operator might be a good idea.when this is implemented in sema, the error for failed binary type resolution should add a note for the correct operator if both ends are strings.
I don't love a separate operator for concatenation. I'm pretty strongly in favor of just keeping it as
+. If we had to have a concat operator,++would be my vote. I'm a little confused what the argument against+as concat is.Sidenote, we should avoid prefix operators if it all possible.
-is fine, but I'd give any others a strong 🤨.just for reference the current list of prefix operators / planned prefix operators
- -x (negation)
- !x (bool not)
- ~x (bitwise not)
- []x (array type, in dev)
- &x (reference type, planned)
- ?x (optional type, planned)
as it stands i don't want to introduce any more non-type prefix operators.
Unrelated to issue but:
&xas a type is a +1 for me.x.&would be my preferred syntax for referencing a value. Given that we're discussing referencing, should we maybe move that discussion to a separate issue? I know you were initially against it but have potentially changed positions at this point.