----- On May 4, 2022, at 11:24 AM, Mathieu Desnoyers
mathieu.desnoy...@efficios.com wrote:
> ----- On May 4, 2022, at 10:43 AM, Robbie Harwood rharw...@redhat.com wrote:
>
>> Mathieu Desnoyers <mathieu.desnoy...@efficios.com> writes:
>>
>>> +version_reverse_sort ()
>>> +{
>>> + case $version_reverse_sort_sort_has_v in
>>
>> Does this need to be its own variable, or could it just reuse
>> version_sort_sort_has_v?
>
> We could do that, assuming that both sort -r -V and sort -V always work (or
> don't)
> in the same way.
>
> Then I would change the test in version_reverse_sort to:
>
> if sort -V </dev/null > /dev/null 2>&1; then
>
> There is no point in calling this check with "-r" if it sets
> version_sort_sort_has_v.
>
>>
>>> +reverse_sorted_list=$(echo $list | tr ' ' '\n' | sed 's/$/ 2/' | sed
>>> 's/\.old
>>> 2/ 1/' | version_reverse_sort | sed 's/ 1$/.old/' | sed 's/ 2$//')
>>
>> Could the `sed | sed` pattern be collapsed into a single invocation
>> using multiple -e switches?1
>
> I think we can straightforwardly turn "| sed 's/ 1$/.old/' | sed 's/ 2$//'"
> into "| sed -e 's/ 1$/.old/' -e 's/ 2$//'".
>
> How to match that all lines ending with ".old" need to be replaced with a
> " 1" suffix, and all _other_ lines need to be suffixed with " 2" in a single
> sed invocation is unclear to me. Any idea ?
Actually with sed -e 's/$/ 2/' -e 's/\.old 2$/ 1/' the second -e applies on the
result of the first, so I can just do that.
Thanks,
Mathieu
>
> Thanks,
>
> Mathieu
>
>>
>> Be well,
>> --Robbie
>
> --
> Mathieu Desnoyers
> EfficiOS Inc.
> http://www.efficios.com
--
Mathieu Desnoyers
EfficiOS Inc.
http://www.efficios.com
_______________________________________________
Grub-devel mailing list
Grub-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/grub-devel
