RE:  407 assignment operators

 I have been looking at removing the "shift-reduce conflicts" from the
 grammar for the language accepted by the 407 expression evaluator.

 The reason for the conflicts is that an assignment is treated like a
 term that you can then perform other calculations on.  For example,
 this is valid syntax for the 407.

    spec ... if -#0:=...

 But such an expression is not valid C; were the 407 to become more
 "C-aligned", the expression would have to change so that the left-hand
 side of the assignment is a proper "l-value", which for a 407 means a
 counter.

    spec ... if -(#0:=...)

 I'm toying with the idea of introducing a nuisance message for this
 condition.  I also want to issue the message about the lack of side
 effect in this:

    spec ... set #0=17

 Comments?  Thoughts?

     j.                                          [EMAIL PROTECTED]

Reply via email to