UNIX for Dummies Questions & Answers

Hi,

Yes, Bdiff is not behaving as i wanted. I am using Sun Solaris.

As you suggested, the best approach will be writing a script. As i am a novice to shell scripting, could you please help me out how to approach in solving this problem? Are there any sample scripts avaialble?

Thanks again

Sue

Assume we have a file called "big". Then we copy "big" to a file called "little" except that we delete some lines. In that case we can display the missing lines with:

Code:

#! /usr/bin/ksh
exec < little
exec 4< big
IFS=""
while read line1 ; do
        match=0
        while ((!match)) ; do
                read -u4 line2
                if [[ "$line1" = "$line2" ]] ; then
                        match=1
                else
                        echo "$line2"
                fi
        done
done
while read -u4 line2 ; do
        echo "$line2"
done
exit 0

Hi,

Your script seems to be working correctly. The problem now i am facing is that the output file kept on increasing its size as there is not end of file defined in the script.

How can we exit immediately after comparing those two files?

Thanks a ton

Sue

I don't understand. It does exit upon reaching end-of-file.

hi,

try the below code...hope it will work.

awk 'file_name=="file1" {arr[$0]++}
file_name=="file2"{ if($0 in arr) {continue}
else {print $0}}' file1 file2

Looks like your "little" file is not a proper subset of your "big" file. If you have lines in your "little" file which are not in your "big" file, Perderabo's script will never exit but will continue to increase the size of the diff file. The simple fix is to check the exit status after each read i.e.

Code:

      read -u4 line2
      if [[ $? > 0 ]]
      then
          # print $line1
          break
      fi

If the little file has any line at all which is not present in the big file, that condition will send the original script into an infinite loop. fpmurphy's fix stops the infinite loop, but the output is probably still wrong. After encountering such a line, my script will output the remainder of the big file. If the little file has a copy of the big file's final line, and all extra lines in the little file follow this final line, I guess it's ok.

BTW, if the little file has any extra lines, it is not a subset at all. A proper subset would have fewer lines. An "improper" or "non-proper" subset would be an exact copy of all lines.