From: joel de guzman (isis-tech_at_[hidden])
Date: 2001-06-02 10:27:07
Micro parser functionality as hinted by Vesa Karvonen
is already implemented in the Spirit parser framework.
Here's a working code snippet:
int a, b;
char const* str = "( -123456, 654321 )";
Scanner<> scanner(str, space);
bool ok = ('(' >> int_p(a) >> ',' >> int_p(b) >> ')').Parse(scanner);
/*... if successful, a and b will hold the parsed numbers */
/* if not, scanner will point to error in the input */
* int_p terminal is a const object that takes a templatized place-holder
this can be an int, a long, even a float, or possibly a big integer.
Doesn't matter; as long as it conforms to the semantics of an integer.
* There's a corresponding uint_p terminal for unsigned ints.
* I am working on the floating point counterpart of int_p: float_p
I welcome suggestions on how to make the code generic such that
big floats can be supported but without penalizing float and double
built in types (template partial specialisation?).
* The parser can of course be quite complicated (with many rules)
Comments and suggestions welcome.
I will upload the code tomorrow: http://isis-tech.n3.net
Joel de Guzman
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk