intern used bignum unicode code points?

It would be nice to be able to compare characters with eq?. This would remove a 
constant cost per processed character from parsing, regexp matching, suffix 
arrays and many other things.

Sadly this can't be done because the characters can be bignums and thus not be 
eq?. This could be done if one automatically interned such code points during 
string construction. This might well be a worthwhile thing to do, because one 
would typically have to pay for this only at Unicode parsing phase, and most 
likely typically there wouldn't be millions of high code points in active use.

On the other hand adding another store and making bytes->string depend on 
threading, which might change a lot in the future, seem like bad ideas.

Better to think about this some more.

Original issue reported on code.google.com by aohelin on 27 Dec 2011 at 9:20

Merged into: #130

kleopatra999 / owl-lisp

intern used bignum unicode code points? #102