Awk/Cut variable denying further use of mutt?

Question

So I am building a script to check for files with certain errors in a bunch of files, based on output from an SQL DB. The file with the error shall be sent to me via mail.

The problem is that when I try to send the mail, I get the message "script.sh: 9: mutt: not found" Which does not occur, if I send the mail before the PATH variable is created.

The script looks as following:

JOB=$(sudo cat /tmp/sqltest.txt | awk '{ print $5 }')
DATE=$(sudo cat /tmp/sqltest.txt | awk '{ print $1 }')
CODE=$(sudo cat /tmp/sqltest.txt | awk '{ print $3 }')
PATH=$(grep ${CODE} /tmp/unzip/* | awk '{ print $1 }' | cut -d':' -f1 | head -n 1)
echo "File containing error message for job "${JOB}" at "${DATE}"" | mutt -a "/tmp/sqltest.txt" -s "Mail title" -- <mail@address>

In short, grep finds the file where the error code is, awk picks out the column with the path to the file, the column also comes with a timestamp which cut removes and head ensures that I only get one result, if the error is reported several places.

I can send the mail with mutt if I use it after variable CODE, instead of PATH, though I unfortunately need PATH instead of /tmp/sqltest.txt

Do you have any ideas on what might cause this?

Answer 1

What we got here is a classic case of trying to use an environment variable (and a pretty important one !) : just use another variable name to get rid of the error. As some suggested, it is good practice to try to avoid full-uppercase variables.

Environment Variables There is a couple of environment variables inside Bash, PATH being one of it. You can get the list of both environment and shell variables using the set command.

Source : Environment Variable on Wikipedia

Answer 2

You can simply add the output obtained by command substitution to the PATH.

Change the line setting the PATH to

PATH=$(grep ${CODE} /tmp/unzip/* | awk '{ print $1 }' | cut -d':' -f1 | head -n 1):${PATH}

The change in PATH would be valid for the duration of the script.