On Tuesday 21 June 2011, Ralf Wildenhues wrote: > * Stefano Lattarini wrote on Mon, Jun 20, 2011 at 11:59:57PM CEST: > > On Monday 20 June 2011, Ralf Wildenhues wrote: > > > * Stefano Lattarini wrote on Mon, Jun 20, 2011 at 05:05:45PM CEST: > > > > sc_tests_plain_automake: > > > > - @if grep -v '^#' $(srcdir)/tests/*.test | grep -E ':[ > > > > ]*automake([^:]|$$)'; then \ > > > > + @if grep -v '^#' $(srcdir)/tests/*.test | grep -E ':[ > > > > ]*automake\>([^:]|$$)'; then \ > > > > > > The RE that was there before was there specifically to emulate the > > > nonportable '\>' construct. Now, I'm not sure I should fight for using > > > Posix compatible regular expressions in maintainer-check rules (seems I > > > lost that battle earlier already), > > > > > Well, notice that I've just followed the existing practice in using GNU > > grep extensions in the maintcheck rules; > > Yes; notice that I had noticed that. > OK :-)
> > > but if you require GNU grep, please be consistent and remove the > > > now-unneeded stuff afterwards and the -E. > > > > > OK, I will push the attached patch if that's OK with you. > > Ugh; 1/2 seems to be going into the wrong direction of changing test > code to work around suboptimal maintainer checks, so sorry for leading > you on a wrong path there again. 2/2 seems fine, thanks. > So should I drop the first patch and keep only the second one? Regards, Stefano
