How do I use a filter, that is a taken input from standard input, instead of from the default log file /var/log/messages

Question

How would I start off using get-opts? The output should be similar to this, but I need the filter that is a taken input from standard input

  awk < /var/log/messages '{ print $2, $1, $5}' | uniq -c | awk '{ print $2, $3, $1, $4 }' | cut -d':' -f1

Answer 1

If you write a shell script, then the normal answer is to use "$@", which is somewhat magical:

#!/bin/bash
awk '{ print $2, $1, $5}' "$@" |
uniq -c |
awk '{ print $2, $3, $1, $4 }' |
cut -d':' -f1

The "$@" represents 'all the command line arguments', or nothing if there are no command line arguments. Given that awk is a command that reads the files named on its command line, or standard input if no files are specified, this will work correctly — it is an important part of the design of good, general purpose, Unix tools. If you want to process /var/log/messages in the absence of a command line argument, then you need to use the shell parameter expansion notation:

"${@:-/var/log/messages}"

If there are no arguments in the argument list, then it substitutes /var/log/messages as the file name.

You can find out more about "$@" in the Bash manual under Special Parameters. See also How to iterate over the arguments in a Bash script.

Note that uniq does not sort the data; it looks for lines that are adjacent and the same. You would usually insert a sort before uniq -c:

#!/bin/bash
awk '{ print $2, $1, $5}' "$@" |
sort |
uniq -c |
awk '{ print $2, $3, $1, $4 }' |
cut -d':' -f1