bash and awk performance with clear and cursor up command -

i'm testing bash , awk script performance clear vs tput clear , tput cuu1 (cursor up) commands. implemented similar scripts in bash , in awk.

bash: http://pastebin.com/0dsc0a71

awk: http://pastebin.com/waj9inrx

admitting have written them in similar way, analyze different execution times.

in bash script:

• clear bash command fast tput clear command
• and tput cuu1 expensive

in awk script:

• tput cuu1 not expensive

@ system( "clear" ); @

• and clear bash command slower tput clear command

@ "clear" | getline clear ( http://pastebin.com/afh3wfgr ) @

• and clear bash command fast tput clear command

so seems awk performs better tput cuu1 command bash , awk script, system() function slower other direct recall.

@ adding cpu information @

the awk script uses less cpu bash script. bash script uses 4 times more cpu awk script. possible perform bash script?

why tput cuu1 expensive in bash script?

both awk , bash calling same external clear , tput utilities. suspect of difference may in exec overhead of 1 versus other , difference in way 2 scripts implemented. on system clear outputs same sequence tput clear , docs that's case. note clear , tput external utilities , not bash or awk commands.

why want this? use language that's needed other reasons. save sequence in variable @ script's startup don't have make repeated calls tput. use clear convenience since that's - tput more versatile.