Charles Levert <[EMAIL PROTECTED]> writes:
> I have checked that this will hold for CHAR_BIT
> up to 1023. Custom tools used for checking
> this available upon request (part analytical,
> part experimental).
Can you please put a comment to this effect in the source code, where
it's less likely to get lost? It would help to document as much as
you can, as long as it's concise.
> +#if CHAR_BIT < 4
> +#error "CHAR_BIT is smaller than 4; yeah, right!"
> +#else
> +#define DEPTH_SIZE (CHAR_BIT + CHAR_BIT/2)
> +#endif
Please change the name from DEPTH_SIZE to DEPTH_BOUND (since it's
an upper bound), and please avoid the #else. Also, please indent
the preprocessing directive, and please use a clearer diagnostic.
Finally, please use the GNU coding style for spacing within expressions.
Something like this, perhaps:
/* Upper bound for depth of trie. This bound is valid for CHAR_BIT >= 4 and
exact for CHAR_BIT in { 4..11, 13, 15, 17, 19 }. For details, please
see grep-internals.texi. */
#if CHAR_BIT < 4
# error "configuration error: impossible CHAR_BIT"
#endif
#define DEPTH_BOUND (CHAR_BIT + CHAR_BIT / 2)
