From: Junio C Hamano <gitster@pobox.com>
Date: Wed, 23 Sep 2020 19:23:46 GMT
Subject: Re: [PATCH 1/2] diff: Fix modified lines stats with --stat and --numstat
Message-ID: <xmqqk0wki9fh.fsf@gitster.c.googlers.com>
In-Reply-To: <nycvar.QRO.7.76.6.2009231709340.5061@tvgsbejvaqbjf.bet>

Johannes Schindelin <Johannes.Schindelin@gmx.de> writes:

> I believe that that is exactly the reason why we want this:
>
> -	same_contents = oideq(&one->oid, &two->oid);
> +	same_contents = one->oid_valid && two->oid_valid ?
> 		oideq(&one->oid, &two->oid) : !strcmp(one->data, two->data);

Not quite.  The other side should either be

	one->size == two->size && !memcmp(...)

or just left to false, as the downstream code must be prepared for
same_contents being false even when one and two turns out to be
not-byte-for-byte-same but equivalent anyway.