[libvirt] [PATCH 2/4] virt-result.m4: Align string more generously

[Michal Privoznik]

On 9/12/19 12:33 PM, Daniel P. Berrangé wrote:
> On Thu, Sep 12, 2019 at 12:18:10PM +0200, Michal Privoznik wrote:
>> On 9/12/19 12:30 AM, Cole Robinson wrote:
>>> On 9/9/19 3:49 AM, Michal Privoznik wrote:
>>>> The times, when we had small CRTs are long gone. Now, in the era
>>>> of wide screens we can be more generous when it comes to aligning
>>>> the output of configure. The longest string before the colon is
>>>> 'wireshark_dissector' which counts 19 characters.  Therefore,
>>>> align the strings at 20.
>>>>
>>>> At the same time, drop the useless result alignment. It behaves
>>>> oddly - it puts a space at the end of each "no" because of the
>>>> %-3s format we use.
>>>>
>>>> Signed-off-by: Michal Privoznik <mprivozn at redhat.com>
>>>> ---
>>>>    m4/virt-result.m4 | 4 ++--
>>>>    1 file changed, 2 insertions(+), 2 deletions(-)
>>>>
>>>> diff --git a/m4/virt-result.m4 b/m4/virt-result.m4
>>>> index cc622fe35b..36973ba0b5 100644
>>>> --- a/m4/virt-result.m4
>>>> +++ b/m4/virt-result.m4
>>>> @@ -33,9 +33,9 @@ dnl  LIBVIRT_RESULT([yajl], [yes], [-I/opt/yajl/include -lyajl])
>>>>    dnl
>>>>    AC_DEFUN([LIBVIRT_RESULT], [
>>>>      if test "$2" = "no" || test -z "$3" ; then
>>>> -    STR=`printf "%10s: %-3s" "$1" "$2"`
>>>> +    STR=`printf "%20s: %s" "$1" "$2"`
>>>>      else
>>>> -    STR=`printf "%10s: %-3s (%s)" "$1" "$2" "$3"`
>>>> +    STR=`printf "%20s: %s (%s)" "$1" "$2" "$3"`
>>>>      fi
>>>>      AC_MSG_NOTICE([$STR])
>>>>
>>>
>>> For the first 2:
>>>
>>> Reviewed-by: Cole Robinson <crobinso at redhat.com>
>>>
>>> I like the look of the colors and I agree it speeds up visually scanning
>>> the configure output. But I'm neutral on whether adding more m4 to the
>>> build system to facilitate it is worth it. So I'll abstain from giving
>>> ack or nack on those.
>>
>> Fair enough. When we switch to meson we'll get colours for free.
> 
> I've not checked how meson handles it, but in Travis with this series
> applied, the raw logs are now full of escape codes:
> 
>    https://api.travis-ci.org/v3/job/584045926/log.txt
> 
>    checking for library containing forkpty... -lutil
>    checking whether strerror(0) succeeds... yes
>    checking for strerror_r with POSIX signature... no
>    checking whether __xpg_strerror_r works... yes
>    checking whether strerror_r is declared... yes
>    checking for external symbol _system_configuration... no
>    checking for pthread_t... yes
>    checking for pthread_spinlock_t... yes
>    checking for PTHREAD_CREATE_DETACHED... yes
>    checking for PTHREAD_MUTEX_RECURSIVE... yes
> 
> which makes the raw log much less readable.
> 
> This is also the case already for the automake 'make check' output
> in fact >
> PASS: test-fnmatch-h
> PASS: test-fnmatch
> PASS: test-fpurge
> PASS: test-fputc
> PASS: test-fread
> PASS: test-freading
> PASS: test-fseek2.sh
> 
> 
> On the flip side, the default log view honours the colors which is
> nice:
> 
>    https://travis-ci.org/berrange/libvirt/jobs/584045926
> 

This is a travis bug. I've noticed this when doing my own builds and the 
problem is that somehow, travis reads TTY output instead of doing 
something like ./configure | tee. Travis runs ./configure from a TTY and 
thus my code (and obviously 'make check' too) sees FD 1 opened and being 
a TTY (`test -t 1') and thus colours are enabled. Fortunately, the 
default log view is not affected.

Michal