Re: [PATCH v2 2/3] merge-ort: add comment to avoid surprise with new sub_flag variable

[Junio C Hamano <gitster@pobox.com>]

"Elijah Newren via GitGitGadget" <gitgitgadget@gmail.com> writes:

> From: Elijah Newren <newren@gmail.com>
>
> Commit 4057523a40 ("submodule merge: update conflict error message",
> 2022-08-04) added a sub_flag variable that is used to store a value from
> enum conflict_and_info_types, but initializes it with an invalid value
> of -1.  The code may never set it to a valid value, and use the invalid
> one.  This can be surprising when reading over the code at first, but it
> was intentional.  Add a comment making it clear that it is okay to be
> using an invalid value, due to how it is used later.

The current code uses -1 as the "suggest the default course of
action", so -1 is very much a "valid value" from the viewpoint of
the code that suggests how to resolve.  It indeed is an invalid
value from the viewpoint of those who maintain conflict_and_info_types
enum.

> Signed-off-by: Elijah Newren <newren@gmail.com>
> ---
>  merge-ort.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/merge-ort.c b/merge-ort.c
> index 67159fc6ef9..0a935a8135f 100644
> --- a/merge-ort.c
> +++ b/merge-ort.c
> @@ -1886,7 +1886,7 @@ cleanup:
>  		const char *abbrev;
>  
>  		util = xmalloc(sizeof(*util));
> -		util->flag = sub_flag;
> +		util->flag = sub_flag; /* May still be -1 */
>  		util->abbrev = NULL;
>  		if (!sub_not_initialized) {
>  			abbrev = repo_find_unique_abbrev(&subrepo, b, DEFAULT_ABBREV);

This new comment may be a slight improvement, but a valid value of
sub_flag is used only to signal the situation where the code does
not know what to suggest, which feels backwards for longer-term code
evolution.  Presumably, we would use the util->flag field to store
which of the known cases we know what to suggest as we know better.

I wonder if we should initialize the variable to the most generic
CONFLICT_SUBMODULE_FAILED_TO_MERGE instead of -1.  The value would
mean "use the default suggestion", and the two known unworkable
values (not-initialized and history-not-available) are currently
handled according to what these two values mean.  We may later add
more specialization based on other CONFLICT_SUBMODULE_* values.

Thanks.