Re: [Gcl-devel] DO-SYMBOLS and DO-ALL-SYMBOLS don't have implicit tagbodies

[Paul F. Dietz]

Camm Maguire wrote:
> Greetings!  I've just committed a fix (I hope) to the setf values
> issue.  There are two remaining failues now in places.lsp, both
> involving psetq/psetf.  I don't understand what is going wrong, and am
> not sure if the test is correct.  I'd greatly appreciate some
> elaboration here.

This is a test of proper evaluation order.

First, note that PSETQ, like SETQ, behaved like PSETF (SETF) when
assigning to a symbol macro:

   " If any var refers to a binding made by symbol-macrolet, then
     that var is treated as if psetf (not psetq) had been used. "

Now, PSETF says:

   "  For psetf, if more than one pair is supplied then the assignments
    of new values to places are done in parallel. More precisely, all
    subforms (in both the place and newvalue forms ) that are to be
    evaluated are evaluated from left to right; after all evaluations
    have been performed, all of the assignments are performed in an
    unpredictable order. "

So, this form:

(symbol-macrolet ((x (aref a (incf i)))
                    (y (aref a (incf i))))
    (let ((a (copy-seq #(0 1 2 3 4 5 6 7 8 9)))
          (i 0))
      (psetq x (aref a (incf i))
             y (aref a (incf i)))
      (values a i)))

is equivalent to

    (let ((a (copy-seq #(0 1 2 3 4 5 6 7 8 9)))
          (i 0))
      (psetf (aref a (incf i)) (aref a (incf i))
             (aref a (incf i)) (aref a (incf i)))
      (values a i))

which is equivalent to

    (let ((a (copy-seq #(0 1 2 3 4 5 6 7 8 9))))
      (psetf (aref a 1) (aref a 2)
             (aref a 3) (aref a 4))
      (values a 4))

which should return

      #(0 2 2 4 4 5 6 7 8 9)
      4

        Paul