[PATCH] checkpatch: simplify the output of --list-types

[PATCH] checkpatch: simplify the output of --list-types

[Jean Delvare]

Drop the header and numbering of types. This format was confusing as
it suggested one could pass the number instead of the type name,
however it did not actually work, and numbering wasn't stable anyway.

Signed-off-by: Jean Delvare <jdelvare@suse.de>
Cc: Andy Whitcroft <apw@canonical.com>
Cc: Joe Perches <joe@perches.com>
---
 scripts/checkpatch.pl |    5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

--- linux-4.13-rc7.orig/scripts/checkpatch.pl	2017-09-04 08:51:13.234841824 +0200
+++ linux-4.13-rc7/scripts/checkpatch.pl	2017-09-04 10:07:11.994427535 +0200
@@ -134,8 +134,6 @@ sub uniq {
 sub list_types {
 	my ($exitcode) = @_;
 
-	my $count = 0;
-
 	local $/ = undef;
 
 	open(my $script, '<', abs_path($P)) or
@@ -150,9 +148,8 @@ sub list_types {
 		push (@types, $_);
 	}
 	@types = sort(uniq(@types));
-	print("#\tMessage type\n\n");
 	foreach my $type (@types) {
-		print(++$count . "\t" . $type . "\n");
+		print "$type\n";
 	}
 
 	exit($exitcode);

-- 
Jean Delvare
SUSE L3 Support

Re: [PATCH] checkpatch: simplify the output of --list-types

[Joe Perches]

On Mon, 2017-09-04 at 10:08 +0200, Jean Delvare wrote:
> Drop the header and numbering of types. This format was confusing as
> it suggested one could pass the number instead of the type name,
> however it did not actually work, and numbering wasn't stable anyway.

<shrug>  If it makes you happy.

I suppose a bigger improvement would be to detail
the purpose of each type.

Most are fairly obvious, some like MEMSET are
a bit obscure.

Another defect:

The list also includes types that are commented out.

Re: [PATCH] checkpatch: simplify the output of --list-types

[Jean Delvare]

On Mon, 04 Sep 2017 08:48:36 -0700, Joe Perches wrote:
> On Mon, 2017-09-04 at 10:08 +0200, Jean Delvare wrote:
> > Drop the header and numbering of types. This format was confusing as
> > it suggested one could pass the number instead of the type name,
> > however it did not actually work, and numbering wasn't stable anyway.  
> 
> <shrug>  If it makes you happy.
> 
> I suppose a bigger improvement would be to detail
> the purpose of each type.

That would be great indeed, however this would require a completely
different approach, as I can't think of a way to automatically extract
this information by just parsing the script.

> Most are fairly obvious, some like MEMSET are
> a bit obscure.
> 
> Another defect:
> 
> The list also includes types that are commented out.

This could be considered a feature: if they were not listed, the user
would have no idea that they can be commented out.

OTOH I'm not sure why some tests are commented out in the first place,
it doesn't make much sense to me. Either they serve a purpose and they
should be enabled, or there is some kind of issue with them and they
should be either fixed or deleted altogether.

-- 
Jean Delvare
SUSE L3 Support