>> variable names unique, while correctly respecting the language scoping 
>> rules.
> 
> Only surprising if you haven't written LISP code. :-)

...or a lambda calculus interpretter. :-P

> Still sounds easier to debug than crashing kernel code on a machine with 
> insufficient memory to host a debugger or compiler. :-)

Probably, yes.


Haskell really is a PITA to parse. I mean, think about it:
- Whitespace is significant. (No context-free parsing here!)
- There are arbitrary user-defined operators, which arbitrary precedence 
and associativity.

Mercifully, there are few language keywords. (There's a lot more than 
you think, when you sit down and catelogue them, but it's still nothing 
compared to, say, Pascal or C.)

Even so, when a user can define an operator like *&*#*|* and set it to 
any precedence or associativity, it's not so easy to produce a parse tree...

Post a reply to this message