Making s/spec symbolic (spec2)

[Ambrose Bonnaire-Sergeant]

Hi,

I've done some preliminary work on making s/spec symbolic

in CLJ-2562. I have a feeling I might be missing some deeper

issues or tradeoffs. Is this something that's been attempted before?

Apologies for jumping to solutions before understanding all the problems.

Sleeping on it reminded me s/spec needs to roundtrip on regex specs.

Relatedly, CLJ-2561 fixes a spec2 bug that is (hopefully) more straightforward.

Thanks,

Ambrose