Shell Programming and Scripting

I have a list of filenames that I want created - they must be created via a certain naming convention due to the software I'm using. Once they are created I have another file that will be used to populate them with data. So far this is what I have:

Code:

#For each line in text file "foo_txt" create a file with that name:

for i in $(cat create/foo_txt); do
        touch create/foodir/$i;

#For each line in text file "bar_txt" store the line of text in each file created by foo_txt:

for i in $(cat create/bar_txt); do echo $i >> create/foodir/<each_file_from_foo_txt>; done

For example, foo_txt would contain the following:

Code:

file_a
file_b
file_c
file_d

And four files would be created. bar_txt would contain the following:

Code:

 new_id1, file_a, 1
new_id2, file_b, 2
new_id3, file_c, 3
new_id4, file_a, 4
new_id5, file_a, 5
new_id6, file_b, 6

Some of the lines from bar_txt would be appended to a file created from foo_txt which is why I'm using the ">>", and that's why the files from foo_txt would contain more than 1 line from bar_txt - whatever filename is listed in foo_txt at the time each line of bar_txt is read.

Your description was somewhat confusing to me. If I understand what you want, this may get you started:

Code:

#!/usr/bin/env ksh
while read file junk
do
    touch $file
done <foo

while read junk1 file junk2
do
    if [[ -f ${file%%,*} ]]                # only if previously created
    then
        echo "$junk1 $file $junk2" >>${file%%,*}
    fi
done <bar

I avoid using constructs like for x in $(cat some-file) do because the number of lines in the file might exceed the system parm limits, and/or there might be multiple words on the file where only one of which is the desired filename. The using while read does the same thing without these dangers.

I made the assumption that lines from 'bar' were only to be placed into a newly created file if the filename existed, and the line was to be written only to the named file. Thus filea would only get the lines from bar that had filea listed as the second token.

You should be able to augment this example with proper directories etc. and please be advised that I did NOT test this code.

It is confusing and I appreciate your help and decrypting my request. Let me just add a little more explanation. When I originally referred to the "bar_txt" file - that's a file containing a list of /etc/passwd entries. The "foo_txt" file contains the name of the server where the /etc/passwd entry will be created. So, I usually have a "foo_txt" file that has repeating entries depending on how many lines of the bar_txt file will be input to the respective file created from foo_txt. I tried to denote this as the server name is listed in each line of the bar_txt file - it would be where you see "file_a" or "file_b" etc. I probably could have written the question a little better.

In any event I appreciate your reply and I'll give it a chance when I can - just thought I'd throw up a quick post and see if I could get some guidance.