Stefan Beller <sbeller@xxxxxxxxxx> writes:
> Signed-off-by: Stefan Beller <sbeller@xxxxxxxxxx>
> ---
> sha1-array.c | 39 +++++++++++++++++++++++++++++++++++++++
> sha1-array.h | 3 +++
> 2 files changed, 42 insertions(+)
>
> diff --git a/sha1-array.c b/sha1-array.c
> index 265941fbf40..10eb08b425e 100644
> --- a/sha1-array.c
> +++ b/sha1-array.c
> @@ -77,3 +77,42 @@ int oid_array_for_each_unique(struct oid_array *array,
> }
> return 0;
> }
> +
> +int oid_array_remove_if(struct oid_array *array,
> + for_each_oid_fn fn,
> + void *data)
> +{
> + int i, j;
> + char *to_remove = xcalloc(array->nr, sizeof(char));
> +
> + /* No oid_array_sort() here! See the api-oid-array.txt docs! */
> +
> + for (i = 0; i < array->nr; i++) {
> + int ret = fn(array->oid + i, data);
> + if (ret)
> + to_remove[i] = 1;
> + }
Doing the same without this secondary array and loop, i.e.
for (src = dst = 0; src < array->nr; src++) {
if (!fn(&array->oid[src], cbdata)) {
if (dst < src)
oidcpy(&array->oid[dst], &array->oid[src]);
dst++;
}
}
array->nr = dst;
would be no less efficient. The only reason why you might want to
measure move-span by a secondary array and preliminary counting loop
like your version does is that moving contiguous area of memory may
be more efficient than moving only by a single oid sized chunks, but
as far as I can tell you are not doing that "optimization", either.
I doubt that remove_if() is particularly a good name. A version of
this function, for which polarity of fn() is reversed, can be called
"grep", or "filter", I think, and would be more understandable.
