text into csv conversion

nails February 17, 2011 at 08:40:43

I am having trouble determining your data format, Do you want everything on one line where column 1 is listed first seperated by commas follwed by column 2 seperated by commas? Everything downshifted?

yes....all columns are in one line seperated by commas.. and next columns of results are downshifted ..............for example pass_min_days, pass_min_length,Pass_warn_age,pass_age_tries,pass_al_warn 0 , 5 , 7 , 5, yes please help thanks in advance

nails February 18, 2011 at 13:15:48

I choose to preprocess the file first: change the entire file to lowercase, eliminate the two lines not needed and determine the length of the file. The awk script prints out the first column, saves the second column in array twocol, and prints out the twocol array when the process ENDs.: #!/bin/bash # downshift the data. assumes data file contains no blank lines tr '[:upper:]' '[:lower:]' < data > newdata # eliminate the lines you don't need sed -e '/pass_max_days/d' -e '/getpass_asterisks/d' newdata > mynewdata # get the number of lines in the new file numlines=$(wc -l < mynewdata) awk ' { # save the 2nd col twocol[NR]=$2 str=$1 # override str variable if needed if($1 == "pass_always_warn") str="pass_al_warn" if(NR < len) printf("%s,", str) else # no comma printf("%s ", str) } END { for(i=1; i <= len; i++) if(i < len) printf("%s,", twocol[i]) else printf("%s ", twocol[i]) } ' len="$numlines" mynewdata # end script Some of your output columns are different from the input file. I only changed pass_always_warn. I leave changing the others to you.