User defined precedence for user defined operators.

Glen,


  > As far as I know, the languages that allow operator overloading
  > use the same precedence in all cases for each operator.


It depends what you mean by an operator. Algol 68 allowed people to
define an identifier to act like an operator. The following example
is copied from page 188 of "Informal introduction to Algol 68" by
Lindsey et al.


op min = (real a, b) real: (a < b | a | b),
        min = (int a, b) int: (a < b | a | b);


a := x min y;


Now the priority (what Algol 68 called precedence) of one of these
'operators' could be specified.


prio min = 9;


All of this could occur in nested blocks potentially resulting in the
case about which our moderator has lost neural connection to the
appropriate memories.


  > [I don't think I've ever seen a language where the operator precedence
  > depended on the types of the operands, and I hope I never do. -John]


What's the problem? Developer knowledge of operator precedence is so
bad it probably wouldn't make much difference.


http://www.knosof.co.uk/cbook/accu06.html
--
Derek M. Jones tel: +44 (0) 1252 520 667
Knowledge Software Ltd mailto:derek@knosof.co.uk
Applications Standards Conformance Testing http://www.knosof.co.uk
[Hey, that doesn't depend on the type, it's redefined from one block to
the next. But it still strikes me as a swell way to introduce bugs
that nobody can find. -John]

Return to the comp.compilers page.
Search the comp.compilers archives again.