Re: GMP parser

[Kevin Ryde]

Hans Aberg <address@hidden> writes:
> 
> Note that in the calclex.l file, one do a table binary search, as it is
> sorted -- it is of course of no practical importance with such a small
> table.

Yes, and the program aims to be a demo, so it just tries to be as
simple as possible.

> The straightforward
>   "abs"    { return ABS; }
>   "bin"    { return BIN; }
>   ...
> is probably also much faster, as Flex should translate it into a one
> character look-ahead NFA.

I think I wanted to lex out a whole word, so for instance "hexx" would
be an error, rather than "hex" followed by "x".  Of course in a
different calculator perhaps the latter would be wanted.