Re: Implementation Language Choice

Lex Spoon wrote:
> The feature that is especially nice is the pattern-matching switch
> statement that most functional languages include.


Agreed, pattern matching is nice. ("Pattern matching" as in "use a
typesafe union type, and recognize the variant at hand and disassemble
the record elements, all in a single construct"; the pattern matching
we're talking about here is unrelated to regular expressions or image
processing.)


Though it's beyond me why this never made it into mainstream imperative
languages; it's exceedingly useful.


> But also garbage collection is a key advantage.


Agreed, and, again, useful beyond functional programming.


> I sometimes wonder if functional language designers know about any
> programs *other* than compilers. :)


There's indeed a bias in that direction.
However, I think that's normal for languages that haven't broken out
into the mainstream.


> However you take this, it is clear that the ML family has proven a
> very effective implementation language for compilers.


Not just ML languages. There's the Haskell family, and there are less
well-known languages like Oz, Alice, or Scala.
All are good at implementing languages. It's a consequence of having
higher-order functions - it makes it so easy to implement languages that
the language designers often go the extra step to make it *really* easy.
This has been creating a bias for languages that are used for compiling
themselves and not much else, but ML and Haskell are definitely beyond
that stage now.


Regards,
Jo
--
Currently looking for a new job.

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