Re: [PATCH] mktag: don't check SHA-1 object length under SHA-256

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

Ævar Arnfjörð Bjarmason  <avarab@gmail.com> writes:

> diff --git a/Documentation/git-mktag.txt b/Documentation/git-mktag.txt
> index fa6a7561236..bbcc0a086bf 100644
> --- a/Documentation/git-mktag.txt
> +++ b/Documentation/git-mktag.txt
> @@ -23,7 +23,7 @@ Tag Format
>  A tag signature file, to be fed to this command's standard input,
>  has a very simple fixed format: four lines of
>  
> -  object <sha1>
> +  object <sha>

Perhaps <hash>? or <objectname>.

> diff --git a/builtin/mktag.c b/builtin/mktag.c
> index 4982d3a93ef..3fa17243e34 100644
> --- a/builtin/mktag.c
> +++ b/builtin/mktag.c
> @@ -5,13 +5,15 @@
>  
>  /*
>   * A signature file has a very simple fixed format: four lines
> - * of "object <sha1>" + "type <typename>" + "tag <tagname>" +
> + * of "object <sha>" + "type <typename>" + "tag <tagname>" +

Ditto.

>   * "tagger <committer>", followed by a blank line, a free-form tag
>   * message and a signature block that git itself doesn't care about,
>   * but that can be verified with gpg or similar.
>   *
> + * The first four lines are guaranteed to be either 83 or 107 bytes;
> + * depending on whether we're referencing a SHA-1 or SHA-256 tag.
> + *
> + * "object <sha1>\n" is 48 or a 72 bytes, "type tag\n" at 9 bytes is the

At least <sha> to be consistent with the above, or <hash>.

>   * shortest possible type-line, "tag .\n" at 6 bytes is the shortest
>   * single-character-tag line, and "tagger . <> 0 +0000\n" at 20 bytes is
>   * the shortest possible tagger-line.
> @@ -46,9 +48,17 @@ static int verify_tag(char *buffer, unsigned long size)
>  	struct object_id oid;
>  	const char *object, *type_line, *tag_line, *tagger_line, *lb, *rb, *p;
>  	size_t len;
> -
> -	if (size < 84)
> -		return error("wanna fool me ? you obviously got the size wrong !");
> +	int minimum_size =
> +		/* Minimum possible input, see t/t3800-mktag.sh */
> +		strlen("object ") + the_hash_algo->hexsz + strlen("\n") +
> +		strlen("type tag\n") +
> +		strlen("tag x\n") +
> +		strlen("tagger . <> 0 +0000\n") +
> +		strlen("\n");
> +
> +	if (size < minimum_size)
> +		return error("got %"PRIuMAX" bytes of input, need at least %d bytes",
> +			     size, minimum_size);

I agree with the patch that this message is not _("marked for
translation"), as it is output from a plumbing.

You need to cast "size" as "(uintmax_t)size" here, probably.  


> @@ -58,7 +68,7 @@ static int verify_tag(char *buffer, unsigned long size)
>  		return error("char%d: does not start with \"object \"", 0);
>  
>  	if (parse_oid_hex(object + 7, &oid, &p))
> -		return error("char%d: could not get SHA1 hash", 7);
> +		return error("char%d: expected object ID, got garbage", 7);

Here you say object ID, which is better than <hash> or <sha>.  Let's
be consistent (I'd say "object name" if I were choosing which to
use).

>  	/* Verify it for some basic sanity: it needs to start with
> -	   "object <sha1>\ntype\ntagger " */
> +	   "object <sha>\ntype\ntagger " */

Here it is <sha>.

> diff --git a/t/t3800-mktag.sh b/t/t3800-mktag.sh
> index d696aa4e52e..93a19bb8df9 100755
> --- a/t/t3800-mktag.sh
> +++ b/t/t3800-mktag.sh
> @@ -26,24 +26,42 @@ test_expect_success 'setup' '
>  	echo Hello >A &&
>  	git update-index --add A &&
>  	git commit -m "Initial commit" &&
> -	head=$(git rev-parse --verify HEAD)
> +	head=$(git rev-parse --verify HEAD) &&
> +
> +	git tag -m"some tag" annotated &&

A SP between -m and its argument, like four lines above?

> +	annotated=$(git rev-parse --verify annotated)
>  '
>  
>  ############################################################
>  #  1. length check
> +for objType in t ta tag

cute ;-)

> +do
> +	cat >tag.sig <<-EOF
> +	object $annotated
> +	type $objType
> +	tag x
> +	tagger . <> 0 +0000
> +
> +	EOF
> +	len=$(wc -c tag.sig)
> +
> +	if test $objType = "tag"
> +	then
> +		test_expect_success "Tag object length check $len passed" '

Here $len may see excess leading whitespace depending on what "wc
-c" on the platform does, but the only effect of that is a bit
uglier test title, so it would be oK.

> +			git mktag <tag.sig >.git/refs/tags/x 2>message &&

Do we use "message" in any way?  

Let's not write directly into filesystem; instead do

			tagobj=$(git mktag <tag.sig) &&
			git update-ref refs/tags/x $tagobj &&

or something like that.

> +			git rev-parse refs/tags/x
> +		'
> +	else
> +		check_verify_failure "Tag object length check on length $len" \
> +			'^error: got .* bytes of input, need at least'

OK.  This is not an issue with this patch, but I do not know why we
want a subshell in check_verify_failure.  May want to clean it up
but not in this patch.

> +	fi
> +done
>  
>  ############################################################
>  #  2. object line label check
>  
>  cat >tag.sig <<EOF
> -xxxxxx 139e9b33986b1c2670fff52c5067603117b3e895
> +xxxxxx $head
>  type tag
>  tag mytag
>  tagger . <> 0 +0000
> @@ -53,17 +71,18 @@ EOF
>  check_verify_failure '"object" line label check' '^error: char0: .*"object "$'
>  
>  ############################################################
> -#  3. object line SHA1 check
> +#  3. object line SHA check

You say "object" or "object ID" to the tester below; let's use that
consistently instead of SHA.

> +invalid_sha=$(echo $head | tr A-Za-z N-ZA-Mn-za-m)
>  cat >tag.sig <<EOF
> -object zz9e9b33986b1c2670fff52c5067603117b3e895
> +object $invalid_sha
>  type tag
>  tag mytag
>  tagger . <> 0 +0000
>  
>  EOF
>  
> -check_verify_failure '"object" line SHA1 check' '^error: char7: .*SHA1 hash$'
> +check_verify_failure '"object" line object check' '^error: char7: .*expected object ID, got garbage'
>  
>  ############################################################
>  #  4. type line label check
> @@ -125,7 +144,7 @@ check_verify_failure '"type" line type-name length check' \
>  	'^error: char.*: type too long$'
>  
>  ############################################################
> -#  9. verify object (SHA1/type) check
> +#  9. verify object (SHA/type) check
>  
>  cat >tag.sig <<EOF
>  object $(test_oid deadbeef)
> @@ -135,7 +154,7 @@ tagger . <> 0 +0000
>  
>  EOF
>  
> -check_verify_failure 'verify object (SHA1/type) check' \
> +check_verify_failure 'verify object (SHA/type) check' \
>  	'^error: char7: could not verify object.*$'
>  
>  ############################################################

Ditto.

Thanks.