Add firstArg to spy calls and fakes.

[GreeKatrina]

Purpose - Add firstArg convenience property to fakes and spy calls.

Currently, we allow getting the last argument for spy calls and fakes. It would be nice to be able to get the first arg without doing args[0]. I couldn't find any open issues or feature requests for this.

Solution

spyCall.firstArg

This property is a convenience for the first argument of the call.

var spy = sinon.spy();
var date = new Date();

spy(date, 1, 2);

spy.lastCall.firstArg === date;
// true

fake.firstArg

This property is a convenient way to get a reference to the first argument passed in the last call to the fake.

var f = sinon.fake();
var date1 = new Date();
var date2 = new Date();

f(date1, 1, 2);
f(date2, 1, 2);

f.firstArg === date2;
// true

How to verify

1. Check out this branch
2. npm install
3. npm test

Checklist for author

• npm run lint passes
• References to standard library functions are cached.

[mroderick]

Thank you for your very first pull request to Sinon 🎉

[mantoni]

Very nice pull request. Thank you! The changes to the documentation should be made in docs/release-source though.

[GreeKatrina]

@mantoni Ah, gotcha. That makes sense. There are a lot of files in that folder so I probably missed it haha. I'll update it and let you know.

[mantoni]

You could use assert.isUndefined here, as you did in fake-test.js.

[GreeKatrina]

I just copied the tests for lastArg. I can change it though if you'd like.

[mantoni]

Maybe assert.isTrue?

[GreeKatrina]

Also copied the tests for lastArg here as well. Happy to change them though.

[fearphage]

It's a bit weird/inconsistent that the values start out as undefined, but become null after being reset().

Not sure this needs to be fixed, but just making note of it.

[GreeKatrina]

I noticed this too. I was going to change it, but didn't want to assume whether the values should be undefined or null. (My vote is undefined.)

[mroderick]

I agree with the idea of returning to undefined. Do you want to make the change, before we merge this in?

[GreeKatrina]

Sure. Did you just want me to change this line to undefined? I feel like it'd be weird to have firstArg and lastArg default to different values, but changing lastArg's default value is a big change. Maybe we could do that in a separate PR and just change firstArg to default to undefined for now?

[mroderick]

I agree. Let's leave lastArg unchanged in this PR.

A separate PR to tidy up the initial and reset values would be most welcome, and will trigger a new major release.

[mantoni]

Nice work. The minor suggestions to change the assertions shouldn't block merging.

[fearphage]

Nice work. The minor suggestions to change the assertions shouldn't block merging.

Should we merge this and fix the minor assertion issues in another PR?

[mroderick]

This has been published as part of sinon@9.0.0