threads / rfc / 17953

RFC patch, 6 partstag: read signature

Subject: [PATCH/RFC 1/6] tag: read signature

## tl;dr

3 messages between Feb 22, 2009 and Feb 22, 2009. Diffs are folded; open one to read it.

replies: 2people: 2as markdown or json

Marc-André Lureau· Feb 22, 2009, 18:06 UTC · lore
Signed-off-by: Marc-Andre Lureau <marcandre.lureau@gmail.com>
---
 tag.c |   11 ++++++++---
 tag.h |    2 +-
 2 files changed, 9 insertions(+), 4 deletions(-)
Show changes to 2 files +9 −4

tag.c, tag.h

diff --git a/tag.c b/tag.c
index 4470d2b..42e5b22 100644
--- a/tag.c
+++ b/tag.c
@@ -38,9 +38,9 @@ struct tag *lookup_tag(const unsigned char *sha1)

 int parse_tag_buffer(struct tag *item, void *data, unsigned long size)
 {
-	int typelen, taglen;
+	int typelen, taglen, siglen;
 	unsigned char sha1[20];
-	const char *type_line, *tag_line, *sig_line;
+	const char *type_line, *tag_line, *sig_line, *msg_line;
 	char type[20];
 	const char *start = data;

@@ -62,9 +62,12 @@ int parse_tag_buffer(struct tag *item, void *data,
unsigned long size)
 		return -1;

 	sig_line = memchr(tag_line, '\n', size - (tag_line - start));
+	if (!sig_line || memcmp("tagger ", ++sig_line, 7))
+		return -1;
+
+	msg_line = memchr(sig_line, '\n', size - (sig_line - start));
 	if (!sig_line)
 		return -1;
-	sig_line++;

 	typelen = tag_line - type_line - strlen("type \n");
 	if (typelen >= 20)
@@ -73,6 +76,8 @@ int parse_tag_buffer(struct tag *item, void *data,
unsigned long size)
 	type[typelen] = '\0';
 	taglen = sig_line - tag_line - strlen("tag \n");
 	item->tag = xmemdupz(tag_line + 4, taglen);
+	siglen = msg_line - sig_line - strlen("tagger \n");
+	item->signature = xmemdupz(sig_line + 7, siglen);

 	if (!strcmp(type, blob_type)) {
 		item->tagged = &lookup_blob(sha1)->object;
diff --git a/tag.h b/tag.h
index 7a0cb00..bc2cab3 100644
--- a/tag.h
+++ b/tag.h
@@ -9,7 +9,7 @@ struct tag {
 	struct object object;
 	struct object *tagged;
 	char *tag;
-	char *signature; /* not actually implemented */
+	char *signature;
 };

 extern struct tag *lookup_tag(const unsigned char *sha1);

-- 
1.6.2.rc1.28.g05ef4.dirty
Junio C Hamano· Feb 22, 2009, 18:18 UTC · re: Marc-André Lureau · lore

Re: [PATCH/RFC 1/6] tag: read signature

Marc-André Lureau <marcandre.lureau@gmail.com> writes:
Show 11 quoted lines
> diff --git a/tag.h b/tag.h
> index 7a0cb00..bc2cab3 100644
> --- a/tag.h
> +++ b/tag.h
> @@ -9,7 +9,7 @@ struct tag {
>  	struct object object;
>  	struct object *tagged;
>  	char *tag;
> -	char *signature; /* not actually implemented */
> +	char *signature;
>  };

I do not speak for Daniel, but I think the original intent of "signature" is about the GnuPG signature, not the "tagger" field.

And this is not an objection. The use of GnuPG is accidental and at the low level of the object layer like this codepath we would not necessarily want to be married to it. Grabbing and parsing the tagger field like your patches 1/6 and 2/6 did would be more appropriate.

But then we would probably want to rename this field "tagger" (and then the timestamp field you add in the next patch "tagger_date").

Marc-André Lureau· Feb 22, 2009, 18:25 UTC · re: Junio C Hamano · lore

Re: [PATCH/RFC 1/6] tag: read signature

Hi
On Sun, Feb 22, 2009 at 8:18 PM, Junio C Hamano <gitster@pobox.com> wrote:
Show 25 quoted lines
> Marc-André Lureau <marcandre.lureau@gmail.com> writes:
>
>> diff --git a/tag.h b/tag.h
>> index 7a0cb00..bc2cab3 100644
>> --- a/tag.h
>> +++ b/tag.h
>> @@ -9,7 +9,7 @@ struct tag {
>>       struct object object;
>>       struct object *tagged;
>>       char *tag;
>> -     char *signature; /* not actually implemented */
>> +     char *signature;
>>  };
>
> I do not speak for Daniel, but I think the original intent of "signature"
> is about the GnuPG signature, not the "tagger" field.
>
> And this is not an objection.  The use of GnuPG is accidental and at the
> low level of the object layer like this codepath we would not necessarily
> want to be married to it.  Grabbing and parsing the tagger field like your
> patches 1/6 and 2/6 did would be more appropriate.
>
> But then we would probably want to rename this field "tagger" (and then
> the timestamp field you add in the next patch "tagger_date").
>

Yes, I though the same. But when I saw the way it was parsed before (the value of sig_line), I was not sure about naming. I agree it would be better to name it "tagger".

regards,
-- 
Marc-André Lureau
Sent from: Helsinki Southern Finland Finland.

← back to recent threads