Many source languages provide a way of concisely declaring a dictionary type object. However, whilst some have semantics that insist the key is a
string (or maybe even a
number), others allow the key to be any string-like type.
If you have a standard dictionary with a
string based key then you should continue to use the standard syntax for declaring it in the source language. For example, in TypeScript:
dict type enables us to declare a string-like key type in a source language that does not inherently allow it.
dict is aliased to a standard dictionary structure, which means we still work with it in code as if it were a
string based key.
In the emission
K will be used in place of
string for declarations. This is simply type name substitution, similar to how intrinsics are implemented.
NOTE This workaround is primarily to satisfy native contracts in languages that have more liberal dictionary constraints like Swift. If you are not exchanging dictionary instances with native third party code it is recommended you stick with the more common
string based key paradigm.