conda bug

[rajeevn1]

The following should return (a1) but it returns (_0)

(reazon-run* x
  (reazon-conde
   ((reazon-== x 'a1))
   ((reazon-== x 'a2)))
  (reazon-conda
   ((reazon-== x 'a2) #'reazon-!U)
   (#'reazon-!S)))

[rajeevn1]

I get the correct answer, (a1), using minikanren under racket.

(run* (x) (conde ((== x 'a1)) ((== x 'a2))) (conda ((== x 'a2) (== x 'c)) ((== x x))))

[nickdrozd]

Congratualations on being the first external contributor to Reazon! I edited your comments for formatting.

Could you explain what is going on in this example? I don't really understand how conda works (or condu for that matter). Is it possible that this is undefined behavior or something like that, and that different implementations might yield different results?

If the behavior here really is wrong, do you have any idea about what might fix it on the Elisp side? The implementation here is pretty close to the one given in The Reasoned Schemer (2nd ed).

[rajeevn1]

I do not know the implementation details but according to https://scholarworks.iu.edu/dspace/bitstream/handle/2022/8777/Byrd_indiana_0093A_10344.pdf?sequence=1

QUOTE
conda and condu differ from conde in that at most one clause can succeed. Furthermore, the clauses are tried in order, from top to bottom. Also, the first goal in each clause is treated specially, as a “test” goal that determines whether to commit to that clause; in this way,conda and condu are reminiscent of cond.
UNQUOTE