Bash script to echo the first positional of each line into file with name of the second

Question

I have a text file that is organized as follows:

Group 1
asdsdsdsdf.html  jeff
xcvxcvxcvx.html  bob
vrgeiuvhif.html  sue

Group 2
iwdowijdoi.html  mary
pokpompojm.html  doug
ndkjfsjfbs.html  lisa

I need a bash script that creates a directory named after each group. And then a text file inside named for each person in that group with their corresponding link on the first line.

I have managed to be able create the directories, the named text files, and a file with only the links. But I don't know how to echo each link into each name file or how to sort them by group.

I'm sure there are better ways to accomplish the small amount that I have.

#!/bin/bash

grep Group list.txt > groups
grep -v Group list.txt > links_names
while read file; do mkdir "${file}"; done < groups
while read line; do export name=`echo $line | awk '{print $2}'`
touch "$name"; done < links_names
while read line; do export link=`echo $line | awk '{print $1}'`
echo "$link" >> links ; done < links_names

rm {groups,links_names,links}

Answer 1

With awk:

awk '/^Group/{g=$0; system("mkdir \""g"\""); next} g&&$0{print $1 >g"/"$2}' file

• /^Group/ if the line starts with the string Group...
  • g=$0 set the variable g to the group name, for example Group 1.
  • system(...) call the command mkdir with the system function to create that directory (notice the additional quotes to deal with spaces in group names).
• g&&$0 if the variable g is defined and the line is not empty...
  • print $1 >g"/"$2 write the value in the first field $1 (the link) to a file in the directory g name by the value of the second field $2 (the name).

---

The test:

$ cat Group\ 2/doug
pokpompojm.html

Answer 2

Requirements

I need a bash script that creates a directory named after each group. And then a text file inside named for each person in that group with their corresponding link on the first line.

#!/bin/bash
while read a b
do
    # Skip blank lines
    test -z "$a" && continue

    if [[ "$a" == "Group" ]]
    then
        # Create the group directory
        group_dir="$a $b"
        mkdir "$group_dir"
    else
        # Write the link into the named file
        echo "$a" > "$group_dir/$b"
    fi
done <list.txt

For the purists, this assumes a number of restrictions:

• no line in your text file begins with significant whitespace
• no link contains whitespace
• there are either precisely two words per line separated by whitespace, or that lines are empty
• that the first non-blank line is a Group header

All of these assumptions can be overcome at the expense of additional code (and complexity).

Answer 3

If your textfile is strict, as in your group will always be called "Group" followed by a space and a word and will not have any spaces in the group name, and similarly for the names after .html, this should work.

#!/bin/bash

parseNdostuff(){
    case $1 in
        Group)
            shift
            mkdir $1
            cd $1
            ;;
        *.html)
            link=$1
            shift
            echo $link > $1
            ;;
        *)
            cd ..
            ;;
    esac

}

while read line
do
   parseNdostuff $line
done <  mytext.txt

Just do a while read line and your text file input at the end < mytext.txt, and feed it to a function i wrote. The reason I chose to do it in a function was for simplicity of dealing with each line, plus being able to easily use case and shift.

Some details:

$1 is the first argument of the function (the $ variables are special, look them up)

shift pushes down the arguments ($1 gets dropped, $2 becomes $1, $3 -> $2)

when it creates a directory it cd into that directory so when the statement runs again and sees the .html lines it makes the files and inputs the link all in the proper directory.

*) is all other cases, so if you have a blank line between groups, it will just cd .. and go down a directory, and you'll be back where you started your script. Note, all these cd's are relative to your script placement.