Re: User defined precedence for user defined operators.

glen herrmannsfeldt wrote:


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


The same precedence, associativity, and other semantics of the
"original" operators.


>>Type(a) + Type(b) returns a Type(b)


IMO binary operands will expand to a.<op>(b), with a possible type cast
of b, into an type compatible with a.<op>. But a and b may be swapped as
well, according to the grouping of the operator <op>.




> [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]


At least not in a language, that allows for user defined and possibly
polymorphic types, whose exact type is unknown at compile time.


There may exist situations, like a multiplication of an vector by a
scalar, where special handling might make sense. Nontheless it also
would make sense, then, to introduce new operators for such operations.


DoDi
[I don't doubt that it's possible to write a parser with type-dependent
precedence, but if ever there were a feature guaranteed to produce
unreadable code with hard to find bugs, that's it. -John]

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