string manipulation in BASH

ovince · 04-16-2007, 05:46 AM

Hi All,

I would like to extract numbers 5, 15, 35 from filenames like (the second number in name):

join5x_angle_box5_tri.dat
join5x_angle_box15_tri.dat
join5x_angle_box35_tri.dat
....

How to do this?

I was thinking to apply command like

`expr index "$fileName" '.*box.*'`

in a loop, but it gives me back wrong position of the first number (2). What I do wrong here?

Thanks

acid_kewpie · 04-16-2007, 06:23 AM

i would do something like:

Code:

# echo join5x_angle_box35_tri.dat | sed -e "s/^.*box\(.*\)_.*$/\1/g"
35

_john_i_ · 04-16-2007, 08:02 AM

Here's another way:

Code:

echo 'join5x_angle_box35_tri.dat' | tr "nx_" "   " | awk '{print $2" "$6}'

5 35

ghostdog74 · 04-16-2007, 11:07 AM

Code:

echo "join5x_angle_box35_tri.dat" | awk '{gsub(/[a-z_.]/," "); print $1,$2}'

cfaj · 04-16-2007, 07:15 PM

Quote:

Originally Posted by ovince

I would like to extract numbers 5, 15, 35 from filenames like (the second number in name):

join5x_angle_box5_tri.dat
join5x_angle_box15_tri.dat
join5x_angle_box35_tri.dat
....

Code:

file=join5x_angle_box5_tri.dat

temp=${file#*_box}
num=${temp%%_*}