Email list hosting service & mailing list manager

set-map and bag-map argument order Jeremy Steward (02 Nov 2016 21:59 UTC)
Re: set-map and bag-map argument order Arthur A. Gleckler (05 Nov 2016 03:53 UTC)
Re: set-map and bag-map argument order Arthur A. Gleckler (07 Nov 2016 18:08 UTC)

set-map and bag-map argument order Jeremy Steward 02 Nov 2016 21:59 UTC

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Hello all,

Today I was perusing through the SRFI docs in an attempt to use `set-map`. To
my surprise the argument order specified by the SRFI document were slightly
backwards to the code. The problem is on lines 600 to 618 in srfi-impl.scm.
You can find this online at
https://github.com/scheme-requests-for-implementation/srfi-113/blob/master/sets
/sets-impl.scm.

The offending parts are as follows:

    (define (sob-map proc comparator sob)
      (let ((result (make-sob comparator (sob-multi? sob))))
        (hash-table-for-each
          (lambda (key value) (sob-increment! result (proc key) value))
          (sob-hash-table sob))
        result))

    (define (set-map comparator proc set)
      (check-set set)
      (sob-map comparator proc set))

    (define (bag-map comparator proc bag)
      (check-bag bag)
      (sob-map comparator proc bag))

Basically, sob-map, which implements the actual mapping process over a set or
a bag, has a different argument order than what set-map or bag-map uses. I
think the intent (and IMO, the more pleasant) of the two argument orders is
that it goes (set-map comparator proc set). Thus I propose that the argument
order in sob-map be changed so that proc and comparator are swapped. This
will put the implementation in line with the SRFI document, and shouldn't
require publishing any errata (outside of documenting the fact that this bug
existed, perhaps).

I will submit a pull-request shortly on Github to reflect this change. For
those who are using the CHICKEN Scheme egg for this SRFI, I have already
applied and uploaded this change (can be reverted if a different solution is
preferred).

Regards,
- --
Jeremy Steward
-----BEGIN PGP SIGNATURE-----
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=HCHD
-----END PGP SIGNATURE-----