Some fix to src/kwset.c [was: Use of 0 instead of NULL in src/kwset.c]

[Julian Foad]

Charles Levert wrote:
> I think the only pending patch that's affected
> is one you were working on to fix a bug.

Sorry to be so forgetful, but could you post the URL or tracker number of the 
particular bug that you are talking about?

>  I would now propose
[second revision]
> --- src/kwset.c 2005-07-04 01:14:37 -0400
> +++ src/kwset.c 2005-07-04 04:50:38 -0400
> @@ -404,7 +404,13 @@ kwsprep (kwset_t kws)
>         }
>        /* Build the Boyer Moore delta.  Boy that's easy compared to CW. */
>        for (i = 0; i < kwset->mind; ++i)
> -       delta[U(kwset->target[i])] = kwset->mind - (i + 1);
> +       {
> +         int d = kwset->mind - (i + 1);
> +
> +         if (UCHAR_MAX < d)
> +           d = UCHAR_MAX;
> +         delta[U(kwset->target[i])] = d;
> +       }
>        /* Find the minimal delta2 shift that we might make after
>          a backwards match has failed. */
>        c = kwset->target[kwset->mind - 1];
>
>
>
> along with tests that would be surrounded by a
>
>    for mode in F G E
>
> loop in "tests/foad1.sh".

Please could you write those tests?

>  I still have to fully
> understand how this delta[] array is used and
> may (or not) suggest additional individual tests
> later on.

In another email you wrote:
> (I still don't understand the second test case,
> but I think the fix is correct nevertheless.
> Please don't use a MIN() macro with a complex
> twice-evaluated argument that the optimizer
> might not pick up on, though; see a previous
> email for my proposed alternative.)

Please could you be more specific - what test are you looking at, and 
what/where is that alternative?

- Julian