Shell Programming and Scripting

Please note the text marked in red above. You say you want every 4th line from your file. Every 4th line is marked in red. Then you say that you want to extract 4 values (one of which does not appear in the file at all) from those 12 lines.

mahi_mayu069 and RudiC have suggested ways to get the unique values from your output file; but that isn't what you requested. RudiC's code also strips off leading zeros; but you said you wanted to convert the 3rd and 4th characters of a 4 or more digit string treated as octal values to decimal.

I'm going to try once more before I give up on this thread:

1.Why do you want an array containing 3, 5, 4, and 7 when 4 never appears in your data?
sorry it's a typo.

2.If you want every 4th line from a file, why do you expect to get 4 values from a 12 line file?
my program generate four instance so i get four values.

3.Do you really want to extract data from every 4th file from your list of files, or do you want to extract the digits between the 1st and 2nd periods in all of the filenames and process all distinct values you find.

I want to extract the digit for example i want to extract 0003 and convert to decimal 3 from this (/tmp/grepd.0003.00049:12:server_read_cb:xxld died)

4.Why do you just want to look at the 3rd and 4th characters from the digits between the 1st and 2nd periods in your filenames?

In 0003 i need decimal coverted 3.

Code:

cut -c 3,4

gives 03 and i convert to decimal using

Code:

val=03
echo "$0 + $val"|bc

don't confuse with code above code is just example.


5.Why do you want an array? Why not just process the values as you find them?

I've a for loop in my script and further processing of restarting of process based on these values. so i need array for temp storage of values and process it in loop.

Here I've two things
1. After extracting the value from the filename , covert to decimal

2. Store those values in array and process array in loop further restarting of process.

I hope u r clear now with above ans. plz ans how i can i store values in array and process them later.

Did you try the proposal in post#7?

If I correctly understand what you're trying to do (and I'm not convinced that I do), you can combine your script that produces your data file:

Code:

grep -rn "died" /tmp/grepd.0*|cut -d':' -f1|cut -d'.' -f2|cut -c 3,4

and RudiC's suggestion:

Code:

ARR=($(awk '/died/ {T[$2]} END {for (i in T) print i+0}' FS="." file))

into:

Code:

array=($(grep -rl "died" /tmp/grepd.0*|awk -F'[.]' '{T[$2]} END{for(i in T) print i+0}'))

or:

Code:

array=($(find /tmp/grepd.0* -type f|awk -F'[.]' '/died/{T[$2]} END{for(i in T) print i+0}'))

This should work both with ksh and bash. If you want to try either of these on a Solaris/SunOS system, change awk to /usr/xpg4/bin/awk, /usr/xpg6/bin/awk, or nawk.

What you do with array is up to you...

yes this works fine.. But values are not coming in sorted order

You didn't say anything about needing sorted output before (and you showed us one example where the output you said you wanted was not sorted). But, making several more very wild assumptions based on a pathname sample set of size 1, try:

Code:

array=($(ls /tmp/grepd.0*died|awk -F'[.]' 'BEGIN{p="x"}p!=$2{print (p=$2)+0}'))