On 11/26/20 4:23 PM, Larry McVoy via cctalk wrote:
I'm sort of wondering why the solution for the
problem isn't good
enough.
As indicated elsewhere the code generating the output (via print(f)
statements) is written in (m)awk.
I'm not aware of a good / convenient / nice / pretty / shiny interface
between awk and tput, save for shelling out via system(...), which seems
less than graceful.
I will evaluate the possibility of shelling out via system(...) to tput
to get the proper string sequences. There are only about 10 different
strings, so a BEGIN{...} could conceptually retrieve the output of tput,
store it in variables that are subsequently used in the main body and
the END{...}.
Shelling / system(...) thousands of times seems less than ideal, no
matter ho fast the CPU is. It's just wasteful.
--
Grant. . . .
unix || die