Solved script doesnt give dsired results

nails October 23, 2011 at 08:48:58

✔ Best Answer

Your most glaring error is your placement of the echo and if statements after the for loop and between the do. It should be like this: for file_name in .... do . . Since you asked, I would like to discuss coding philosophy; first, I would also include the shell invocation at line 1, column 1: #!/bin/sh or whatever shell you are using. Next, your use of the for loop is considered a useless use of cat - or UUOC. you can read about it here: http://partmaps.org/era/unix/award.... Which means you could do something like this: cd /home/jessie for file_name in * do . . Next, keep in mind that using `ls /home/jessie` does not guarantee that all objects are listed one per line. Consider using the -1: ls -1 /home/jessie # guarantees one object per line One of the biggest problems with your original for loop is that: for file_name in `ls ....` do breaks if the output of ls has more that one column. You might consider using a while loop instead of a for loop: ls /home/jessie| while read file_name do . . Let me know if you have any questions or comments. Good luck with your script.

Greeaaat thankss i have took some notes and came up with the following which works perfectly however began to notice another problem, the contents of my path /home/jessie are files in the form ABC_25102011.DAT when we cat the file i have records with two types, one starts with "O" and other starts with "I" i need only those that start with O as below cd /home/jessie/ for file_name in `*.DAT` do {TODAY_DATE}=`date +"%Y%m%d"`; echo "" >>/home/jessie/output/CountCheck_${TODAY_DATE}.txt echo "Script Started at: `date`" >> /home/jessie/output/CountCheck_${TODAY_DATE}.txt echo "" >> /home/jessie/output/CountCheck.txt echo "opening /home/jessie" >> /home/jessie/output/CountCheck_${TODAY_DATE}.txt if [ $file_name -eq 0 ];then echo "no files found...." >> /home/jessie/output/CountCheck_${TODAY_DATE}.txt else basic_name=`echo $file_name | cut -f1 -d "."` echo "spliting files....." >> /home/jessie/output/CountCheck_${TODAY_DATE}.txt egrep -v " 00122 | 00123 | 00180 " $file_name >$basic_name"_1.DAT" echo "file 1 done...." >> /home/jessie/output/CountCheck_${TODAY_DATE}.txt egrep " 00122 | 00123 | 00180 " $file_name && grep -s ^O $file_name >$basic_name"_shrtcode.DAT.tmp" <== i tried this but it doesnt work...? echo "file 2 done...." >> /home/jessie/output/CountCheck_${TODAY_DATE}.txt echo "preparing to replace text....." >> /home/jessie/output/CountCheck_${TODAY_DATE}.txt sed -e "s/\ 00122\ /\ 00200122\ /g" -e "s/\ 00123\ /\ 00200123\ /g" -e "s/\ 00180\ /\ 00200180\ /g" $basic_name"_shrtcode.DAT.tmp" >$basic_name"_shrtcode.DAT" echo "records updated successfully ..........." >>/home/jessie/output/CountCheck_${TODAY_DATE}.txt fi;

nails October 25, 2011 at 09:46:10

Try something like this: egrep "^O" $file_name |egrep " 00122 | 00123 | 00180 " Get all the lines starting with uppercase "O" before searching for the other text.

yea actually it works fine, so the problem was the && ??

nails October 27, 2011 at 13:58:16

This seems to work, but I have had portability problems with & and egrep: egrep "^O&00122|00123|00180" test.txt This is more efficient than piping a grep to grep as it spawns one less shell.