To pass a parameter with sed

Question

For my script I have to read the data from a file.

For example I have 2 files: one is a text file, the second is a csv file with 10 columns.

What I am trying to do is to read the csv and change the parameters in the first file.

1. Using sed -n to read the data one row at a time and save this in a fileC
2. Using sed "s/&replacefile1param/&readFileC/g"

What I have issue is to read the cat of readFileC as a parameter in the second step. How can I do this?

The first file is a sample script file with 2 parameters inside. The second file is a CSV file with 200 lines with 6 columns.

I need to change the template file by reading the columns from the CSV, and substituting the parameters. The result need to be 200 file1 and each result needs to be saved in the location specified in the the 4th column of the CSV as well.

Sample template file

The name =scriptname

The staments go here

job name = jobname

exit

Sample CSV data

scriptname,jobname,date,location,updated,env
kkkkkk,filetest,2001-01-01,/san00/project,test
kkkkk2,filetest2,2001-01-01,/san00/project,test
kkkkk3,filetest3,2001-01-01,/san00/project2,real

I need to replace the parameters in the template file with values taken from the CSV, and save the file in the /san00 location from the 4th column of the CSV.

Desired output

The script name kkkkkk

The staments go here

job name = filetest

exit

This should be saved in san00/project.

Answer 1

file1 is the template and file2 the csv. What happens if two rows have the same location, is undefined, the solution below overwrites the location file with the latest data. It is also unclear to me if the /san00 directory is the only one used in the location part and if it is supposed to be an absolute path.

Solution creating locations relative to current directory (notice the . in .$dir and .$location).

while IFS=, read scriptname jobname date location updated env; do
    dir=$(dirname $location)
    [ -d .$dir ] || mkdir -p ".$dir"
    sed -e "s/scriptname/$scriptname/" \
        -e "s/jobname/$jobname/" file1 > ".$location"
done < <(tail -n +2 file2)

Solution with absolute paths in location, target directories must pre-exist

while IFS=, read scriptname jobname date location updated env; do
    sed -e "s/scriptname/$scriptname/" \
        -e "s/jobname/$jobname/" file1 > "$location"
done < <(tail -n +2 file2)

There are many ways to extract the information of file2, the above is just one of them, and depending on the nature of the data it might be more complicated than this.

Answer 2

This might work for you:

sed '1d;s|\([^,]*\),\([^,]*\),[^,]*,\([^,]*\).*|sed "1s/scriptname/\1/;5s/jobname/\2/" template >\3|' file.csv |
sh