Add tests for >=

[dgr]

Closes #123

[E-A-Griffin]

nitpick, I think test>= is more clear since to me this looks like we're testing some function/macro ->=

[dgr]

The template has a standard format: "test-" appended with function name. I'm going to leave this as is unless you have a lot of heartburn, just for consistency with every other test.

[E-A-Griffin]

indentation is weird here

[dgr]

Fixed.

[jeaye]

I think we'll want to scale back on these negative tests, since they're technically undefined behavior in Clojure. The docs string for >= says Returns non-nil if nums are in monotonically non-decreasing order, otherwise false. so there is an implication that all inputs must be nums.

Interestingly, the doc string also says non-nil, rather than true.

[E-A-Griffin]

Yeah "returns non-nil" is very oddly non specific

[dgr]

I'll switch the tests to check for non-nil and false as opposed to true and false.

[dgr]

Also removed all negative tests for anything other than numbers and nil.

[E-A-Griffin]

Small things but looks good to me!

[E-A-Griffin]

One thing that I think could be a good add would be binary numeric comparisons for applicable hosts, Clojure, ClojureScript, ClojureCLR, bb for example have

user=> (>= 2r0110 2r0100)
true

[E-A-Griffin]

Also would be good to test negative Infinity

[jeaye]

One thing that I think could be a good add would be binary numeric comparisons for applicable hosts, Clojure, ClojureScript, ClojureCLR, bb for example have

user=> (>= 2r0110 2r0100)
true

I like the out-of-the-box thinking, but in this case, 2r0110 is just parsed into a normal number and the behavior of >= is not different. We'd just be testing the parser, which is not the goal here.

[jeaye]

Also would be good to test negative Infinity

Good thinking.

[dgr]

Also would be good to test negative Infinity

There are multiple tests with ##-Inf, so I'm not sure what you're looking for here.

[dgr]

One thing that I think could be a good add would be binary numeric comparisons for applicable hosts, Clojure, ClojureScript, ClojureCLR, bb for example have

user=> (>= 2r0110 2r0100)
true

I like the out-of-the-box thinking, but in this case, 2r0110 is just parsed into a normal number and the behavior of >= is not different. We'd just be testing the parser, which is not the goal here.

I agree with this. There's no added benefit as the syntax is just specific to the reader and it's all converted to integers by the time it reaches >=.

[dgr]

Changing this to draft PR until I can change the tests to check for non-nil rather than true.

[dgr]

Alright, I think this is ready to go now.

[jeaye]

Thanks, Dave!