[x-goose-x]

Implement sequence as per #502.

[thomasmulvaney]

Thanks for working on this. Would you be able to add some tests please?

[x-goose-x]

Sure thing. While I was testing I noticed that eductions break the seqable contract in the empty case.

We'd expect:
(seq (eduction (map inc) '[])) to produce nil

but it produces []

Calling seq on the result of the reduce seems to resolve this.

[thomasmulvaney]

I think the problem is that the -seq implementation on Eduction just does a reduce rather than returning a LazySeq. I think the best thing to do would be to fix that by making it lazy.

[x-goose-x]

Agreed, I'll take a crack at it.

On Wed, Mar 9, 2016 at 11:17 AM Thomas Mulvaney notifications@github.com
wrote:

I think the problem is that the -seq implementation on Eduction just does
a reduce rather than returning a LazySeq. I think the best thing to do
would be to fix that by making it lazy.

—
Reply to this email directly or view it on GitHub
#509 (comment).

[x-goose-x]

So I inverted things a little bit by making sequence work lazily and eduction make use of sequence.

There wasn't a clear way to lazily iterate over transducers and as far as I can tell there's no equivalent to the iterators that clojure implements, so I had to write lazy transducer application.

[thomasmulvaney]

Thanks for working on this!