On Wed, Jan 02, 2019 at 01:20:47AM +0100, Johannes Sixt wrote:
> > diff --git a/t/test-lib.sh b/t/test-lib.sh
> > index 0f1faa24b2..f47a191e3b 100644
> > --- a/t/test-lib.sh
> > +++ b/t/test-lib.sh
> > @@ -324,9 +324,12 @@ do
> > # isn't executed with a suitable Bash version.
> > if test -z "$test_untraceable" || {
> > test -n "$BASH_VERSION" && {
> > - test ${BASH_VERSINFO[0]} -gt 4 || {
> > - test ${BASH_VERSINFO[0]} -eq 4 &&
> > - test ${BASH_VERSINFO[1]} -ge 1
> > + bash_major=${BASH_VERSION%%.*}
> > + bash_minor=${BASH_VERSION#*.}
> > + bash_minor=${bash_minor%%.*}
> > + test $bash_major -gt 4 || {
> > + test $bash_major -eq 4 &&
> > + test $bash_minor -ge 1
> > }
> > }
> > }
> >
>
> Would it perhaps be simpler to just hide the syntax behind eval? Like
>
> if test -z "$test_untraceable" || {
> test -n "$BASH_VERSION" && eval '
> test ${BASH_VERSINFO[0]} -gt 4 || {
> test ${BASH_VERSINFO[0]} -eq 4 &&
> test ${BASH_VERSINFO[1]} -ge 1
> }
> '
That was my first thought, too. :)
The parsing here is simple enough that I'd be fine either with the
original patch, or an eval-based version (and otherwise, the goal and
description seem quite good to me).
-Peff
