Re: [patch] checkpatch: warn on missing spaces in broken up quoted strings

[Dan Carpenter <dan.carpenter@oracle.com>]

On Fri, Jun 13, 2014 at 12:52:25PM -0700, Joe Perches wrote:
> On Fri, 2014-06-13 at 12:46 +0300, Dan Carpenter wrote:
> > On Fri, Jun 13, 2014 at 02:30:22AM -0700, Joe Perches wrote:
> > 
> > > > +# check for missing a space in a string concatination
> > > > +        if ($prevrawline =~ /[^\\][a-zA-Z]"$/ && $rawline =~ /^\+[\t ]+"[a-zA-Z]/) {
> > > > +            WARN('MISSING_SPACE',
> > > > +                 "break quoted strings at a space character\n" . $hereprev);
> > > > +        }
> > > 
> > > Probably want digits too so maybe \w instead of "[a-zA-Z]/
> 
> Couple nits:
> 
> The indentation isn't right here.
> 
> And this check is probably better placed immediately
> after the "SPLIT_STRING" test.
> 
> []
> 
> > ./drivers/scsi/pm8001/pm8001_ctl.c:297 			       "0x%08x 0x%08x\n",
> > ./drivers/scsi/pm8001/pm8001_ctl.c:432 			       "0x%08x 0x%08x\n",
> > ./drivers/scsi/qla2xxx/qla_nx2.c:1457 	    "0x%X 0x%X 0x%X 0x%X 0x%X 0x%X\n"
> > 	- hex false positives
> 
> These look more like defects to me.
> 
> > I thought about doing that when I wrote my original patch but I was
> > worried about more hex false positives.  If there are only those 3 then
> > it's probably not a big deal.  What do you think?
> 
> I know it works on files, but I'm not sure it works on patches.
> 
> What happens when checking a single line replacement patch?
> 
> Likely the \\[a-zA-Z] check should include
> all the tests that the multiple line string exceptions use.
> 
> (?:\\(?:[ntr]|[0-7]{1,3}|x[0-9a-fA-F]{1,2})|;\s*|\{\s*)"\s*$

That kind of complicate regex hurts my head.  How do I make it not
complain about:

		"foo\n\t"
		"bar";

regards,
dan carpenter