Re: [PATCH 5/5] treewide: always have a valid "index_state.repo" member

From: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Date: Jan 10, 2023, 12:24 UTC
Message-ID: <230110.86o7r6zibu.gmgdl@evledraar.gmail.com>
In-Reply-To: <patch-5.5-b3b0e592101-20230110T060340Z-avarab@gmail.com>

On Tue, Jan 10 2023, Ævar Arnfjörð Bjarmason wrote:

Show 12 quoted lines

> diff --git a/repository.c b/repository.c
> index 3427085fd6d..fdebc8c597c 100644
> --- a/repository.c
> +++ b/repository.c
> @@ -28,6 +28,8 @@ void initialize_the_repository(void)
>  	the_repo.remote_state = remote_state_new();
>  	the_repo.parsed_objects = parsed_object_pool_new();
>  
> +	the_index.repo = the_repository;
> +
>  	repo_set_hash_algo(&the_repo, GIT_HASH_SHA1);
>  }

It's slightly odd to have initialize_the_repository() initialize this other variable, I did consider adding a initialize_the_variables(), and have that call an initialize_the_repository() and an initialize_the_index(), but deemed it not worth it for now. Perhaps something to revisit once we're initializing more variables here.

Arguably such a split doesn't even conceptually make sense, as we're now declaring that "the_repository" isn't fully initalized unless the corresponding "the_index.repo" is also set, i.e. this really is part of "the_repository" initialization.