Is there any difference in behavior when reusing the same Arbitrary instance vs. using distinct instances?

[jmartinezmaes]

💬 Question and Help

Does this:

import * as fc from "fast-check";

const strUpper = fc.string().map((s) => s.toUpperCase());

fc.assert(fc.property(strUpper, strUpper, (x, y) => { /* asserts here */ }));

and this:

import * as fc from "fast-check";

function strUpper(): fc.Arbitrary<string> {
    return fc.string().map((s) => s.toUpperCase());
}

fc.assert(fc.property(strUpper(), strUpper(), (x, y) => { /* asserts here */ }));

have any difference in functionality?

[dubzzz]

I'd say there is no difference at all. All arbitraries are stateless and re-using or not the instance has no impact on fast-check's behaviour, it's supposed to behave fully the same.

The function approach is closer to the APIs of fast-check, but while it makes sense in the library to offer configurable arbitraries it's mostly up to the user when it's for user specific arbitraries.