UNIX for Dummies Questions & Answers

Hi,

I've been trying to replace the numbers in the first column of my text file with all ones, unless the number is equal to 8. I have this:

1 1 11 123 258
2 1 9 135 175
1 1 15 143 274
8 1 13 153 172
8 1 13 154 166
8 1 13 154 167
3 1 15 237 255
4 1 15 243 202
1 1 13 133 166
1 1 13 133 167

And I would like this:

1 1 11 123 258
1 1 9 135 175
1 1 15 143 274
8 1 13 153 172
8 1 13 154 166
8 1 13 154 167
1 1 15 237 255
1 1 15 243 202
1 1 13 133 166
1 1 13 133 167

This can be done pretty easily with structures in Matlab, but only somewhat awkwardly in bash using awk and a while loop, e.g.:

Code:

cat file | while read line; 
   do echo $line;
   then some if loop/case statement here (when == 8, leave alone, else set = 1); 
   then a rewriting of the file here; 
done

....but this is so clunky and cumbersome. I'm wondering if someone out there has a sleeker, more streamlined solution with of awk or sed (or anything).

Any ideas are welcome. Thanks in advance.

Code:

awk '$1!=8 {$1=1} {print}' filename > newfilename

Try this,

Code:

#!/bin/sh
>new_file
while read line; do
        first_field=`echo ${line} | cut -d ' ' -f 1`
        if [[ $first_field -ne 8 ]]
        then
                change_line=`echo ${line} | sed -r 's/[0-9]{,} /1 /'`
                echo ${change_line} >>new_file
        else
                echo ${line} >>new_file
        fi

done < file

Code:

sed -ie 's/^[1-7,9] /1 /' filename

or

Code:

sed -e 's/^[^8] /1 /'  filename

HTH,
PL

Jim - once again, your code worked flawlessly. Thanks to all for the other solutions, too!