Re: [PATCH] checkpatch: allow URL >80 chars

[Andreas Brauchli]

On Fre, 2017-11-17 at 21:21 -0800, Joe Perches wrote:
> On Fri, 2017-11-17 at 16:56 +0100, Andreas Brauchli wrote:
> > Allow URL to exceed the 80 char limit for improved interaction in
> > adaption to ongoing but undocumented practice.
> > 
> > $ git grep -E '://\S{77}.*' -- '*.[ch]'
> > 
> > The patch checks that the URL is indeed on its own line in that it
> > allows a maximal prefix of 4 characters to account for a URL after a
> > comment (e.g. ' // https://...')
> > 
> > The URL format allows for up to 5 protocol characters before the
> > separator :// (e.g. https, git, ...). Additionally, an URL starting
> > with a "domain.tld" or "subdomain.domain" pattern (e.g.
> > 'www.kernel.org') is also accepted.
> > 
> > $rawline is used in the check as comments are removed from $line.
> > 
> > Signed-off-by: Andreas Brauchli <andreas.brauchli@xxxxxxxxxxxxx>
> > ---
> >  scripts/checkpatch.pl | 5 +++++
> >  1 file changed, 5 insertions(+)
> > 
> > diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
> > index 8b80bac055e4..aed447923354 100755
> > --- a/scripts/checkpatch.pl
> > +++ b/scripts/checkpatch.pl
> > @@ -2904,6 +2904,11 @@ sub process {
> >  			} elsif ($line =~ /^\+.*\bEFI_GUID\s*\(/) {
> >  				$msg_type = "";
> >  
> > +			# URL (w/ minimal padding e.g. "+ // ")
> > +			} elsif ($rawline =~ /^\+.*?\b((?:\w{1,5}:\/\/|\w+\.\w+)\S+).*$/ &&
> 
> nack.
> 
> this matches member dereference lines like
> 
> 	foo.bar = baz;
> 
> and include lines like
> 
> 	#include <foo.h>

Agreed, there are even a few false positives (all member derefs).
Apologies for failing to catch those.

Is the patch acceptable if it only catches URL with protocol specifiers
foo:// ? That would catch most cases and doesn't show false positives:
$ git grep -E '://\S{75}.*' -- '*.[ch]'

Cheers,
Andreas