linux-kernel - Re: [PATCH] git-send-email: Add ability to cc: any "trailers" from commit message

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 31 Aug 2016 10:54:39 -0700
From:   Junio C Hamano <gitster@...ox.com>
To:     Joe Perches <joe@...ches.com>
Cc:     git@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] git-send-email: Add ability to cc: any "trailers" from commit message

Joe Perches <joe@...ches.com> writes:

> Many commits have various forms of trailers similar to
>      "Acked-by: Name <address>" and "Reported-by: Name <address>"
>
> Add the ability to cc these trailers when using git send-email.

I thought you were asking what we call these "<token> followed by
<colon>" at the end of the log message, and "footers or trailers"
was the answer.

I do not have a strong objection against limiting to "-by:" lines;
for one thing, it would automatically avoid having to worry about
"Bug-ID:" and other trailers that won't have e-mail address at all.

But if you are _only_ picking up "-by:" lines, then calling this
option "trailers" is way too wide and confusing.  I do not think
there is any specific name for "-by:" lines, though.  Perhaps you
would need to invent some name that has "-by" as a substring.

"any-by"?  or just "by"?  I dunno.

>  if ($suppress_cc{'all'}) {
> -	foreach my $entry (qw (cccmd cc author self sob body bodycc)) {
> +	foreach my $entry (qw (cccmd cc author self sob body bodycc trailers)) {
>  		$suppress_cc{$entry} = 1;
>  	}

OK.

> @@ -448,7 +448,7 @@ $suppress_cc{'self'} = $suppress_from if defined $suppress_from;
>  $suppress_cc{'sob'} = !$signed_off_by_cc if defined $signed_off_by_cc;
>  
>  if ($suppress_cc{'body'}) {
> -	foreach my $entry (qw (sob bodycc)) {
> +	foreach my $entry (qw (sob bodycc trailers)) {
>  		$suppress_cc{$entry} = 1;
>  	}
>  	delete $suppress_cc{'body'};

OK.

> @@ -1545,7 +1545,7 @@ foreach my $t (@files) {
>  	# Now parse the message body
>  	while(<$fh>) {
>  		$message .=  $_;
> -		if (/^(Signed-off-by|Cc): (.*)$/i) {
> +		if (/^(Signed-off-by|Cc|[^\s]+[_-]by): (.*)$/i) {

Micronits:

 (1) do you really want to grab a run of any non-blanks?  Don't
     you want to exclude at least a colon?
 (2) allowing an underscore looks a bit unusual.  

> @@ -1555,6 +1555,12 @@ foreach my $t (@files) {
>  			} else {
>  				next if $suppress_cc{'sob'} and $what =~ /Signed-off-by/i;
>  				next if $suppress_cc{'bodycc'} and $what =~ /Cc/i;
> +				next if $suppress_cc{'trailers'} and $what !~ /Signed-off-by/i && $what =~ /by$/i;
> +			}

It is a bit unfortunate that S-o-b is a subset of any-by that forces
you to do this.

> +			if ($c !~ /.+@.+/) {
> +				printf("(body) Ignoring %s from line '%s'\n",
> +				       $what, $_) unless $quiet;
> +				next;
>  			}

This check is new and applies to sob/cc, too.

I am aware of the fact that people sometimes write only a name with
no e-mail address when giving credit to a third-party and we want to
avoid upsetting the underlying MTA by feeding it a non-address.

Looking at existing helper subs like extract_valid_address and
sanitize_address that all addresses we pass to the MTA go through,
it appears to me that we try to support an addr-spec with only
local-part without @domain, so this new check might turn out to be
too strict from that point of view, but on the other hand I suspect
it won't be a huge issue because the addresses in the footers are
for public consumption and it may not make much sense to have a
local-only address there.  I dunno.

>  			push @cc, $c;
>  			printf("(body) Adding cc: %s from line '%s'\n",