How do I fix this syntax error from awk in my shell script?

Question

I have a script with two codes to pull the sum of data of a given file and print to another file and a second pulls the total number of missing data points to another file. Finally the two output files are placed together in to a third file. All three codes work when run alone. Due to the large number of files needing to have this done I was attempting to automate in a single script file. Running bash shell. The Colavg.sh file has been execute permissions. When I attempt to run this as a single script I get the following error:

awk: cmd. line:1: ./Colavg.sh
awk: cmd. line:1: ^ syntax error
awk: cmd. line:1: ./Colavg.sh
awk: cmd. line:1:   ^ unterminated regexp

#!/bin/awk

awk -F ',' 'NF==4{count++} END {print "missing: "count}' County029-CO-0023-T0000JD.out > avg.out;
awk -F ',' '$4==-999{c999++}; END{print "missing: "c999}' County029-CO-0023-T0000JD.out > avg1.out;
cat avg1.out avg.out >> County029-CO-0023-T0000JD-avg.out

Answer 1

The issue is your script's #!-line. It should read #!/bin/sh as it's a shell script, not an awk script.

What's really happening here is that the shell, with your erroneous #!-line, would execute

/bin/awk ./Colavg.sh

This means using the text string ./Colavg.sh as the awk program. As an awk program, this does not make sense, and your awk interpreter identifies two separate errors before deciding to bail out.

---

The script could be simplified:

#!/bin/sh

name=County029-CO-0023-T0000JD

awk -F , '
    $4 == -999 { c999++  }
    NF == 4    { count++ } 
    END { printf "missing: %d\n", c999  }
    END { printf "missing: %d\n", count }' "$name.out" >>"$name-avg.out"

Change >> at the end to > if you want the output file to be truncated (emptied) before the results are written to it.