Getopts in sourced Bash function works interactively, but not in test script? - bash

I have a Bash function library and one function is proving problematic for testing. prunner is a function that is meant to provide some of the functionality of GNU Parallel, and avoid the scoping issues of trying to use other Bash functions in Perl. It supports setting a command to run against the list of arguments with -c, and setting the number of background jobs to run concurrently with -t.
In testing it, I have ended up with the following scenario:
prunner -c "gzip -fk" *.out - works as expected in test.bash and interactively.
find . -maxdepth 1 -name "*.out" | prunner -c echo -t 6 - does not work, seemingly ignoring -c echo.
Testing was performed on Ubuntu 16.04 with Bash 4.3 and on Mac OS X with Bash 4.4.
What appears to be happening with the latter in test.bash is that getopts is refusing to process -c, and thus prunner will try to directly execute the argument without the prefix command it was given. The strange part is that I am able to observe it accepting the -t option, so getopts is at least partially working. Bash debugging with set -x has not been able to shed any light on why this is happening for me.
Here is the function in question, lightly modified to use echo instead of log and quit so that it can be used separately from the rest of my library:
prunner () {
local PQUEUE=()
while getopts ":c:t:" OPT ; do
case ${OPT} in
c) local PCMD="${OPTARG}" ;;
t) local THREADS="${OPTARG}" ;;
:) echo "ERROR: Option '-${OPTARG}' requires an argument." ;;
*) echo "ERROR: Option '-${OPTARG}' is not defined." ;;
esac
done
shift $(($OPTIND-1))
for ARG in "$#" ; do
PQUEUE+=("$ARG")
done
if [ ! -t 0 ] ; then
while read -r LINE ; do
PQUEUE+=("$LINE")
done
fi
local QCOUNT="${#PQUEUE[#]}"
local INDEX=0
echo "Starting parallel execution of $QCOUNT jobs with ${THREADS:-8} threads using command prefix '$PCMD'."
until [ ${#PQUEUE[#]} == 0 ] ; do
if [ "$(jobs -rp | wc -l)" -lt "${THREADS:-8}" ] ; then
echo "Starting command in parallel ($(($INDEX+1))/$QCOUNT): ${PCMD} ${PQUEUE[$INDEX]}"
eval "${PCMD} ${PQUEUE[$INDEX]}" || true &
unset PQUEUE[$INDEX]
((INDEX++)) || true
fi
done
wait
echo "Parallel execution finished for $QCOUNT jobs."
}
Can anyone please help me to determine why -c options are not working correctly for prunner when lines are piped to stdin?

My guess is that you are executing the two commands in the same shell. In that case, in the second invocation, OPTIND will have the value 3 (which is where it got to on the first invocation) and that is where getopts will start scanning.
If you use getopts to parse arguments to a function (as opposed to a script), declare local OPTIND=1 to avoid invocations from interfering with each other.

Perhaps you are already doing this, but make sure to pass the top-level shell parameters to your function. The function will receive the parameters via the call, for example:
xyz () {
echo "First arg: ${1}"
echo "Second arg: ${2}"
}
xyz "This is" "very simple"
In your example, you should always be calling the function with the standard args so that they can be processed in the method via getopts.
prunner "$#"
Note that pruner will not modify the standard args outside of the function.

Related

Shell script not running on Mac OS

I have what amounts to a very simple bash script that executes a deployment. Here is my code:
#!/usr/bin/env bash
function print_help
{
echo '
Deploy the application.
Usage:
-r reinstall
-h show help
'
}
reinstall=false
while getopts "rh" opt; do
case ${opt} in
r)
echo "clean"
reinstall=true
;;
h)
echo "help"
print_help
exit 0
;;
esac
done
I am calling the script as follows:
. deploy.sh -h
No matter what I do, neither option (i.e. -r, -h) results in the respective echo and in the case of -h the print_help function isn't called.
What am doing wrong?
getopts uses a global variable OPTIND to keep track about which argument it processes currently. Each option it parses, it increments/changes OPTIND to keep track which argument will be next.
If you call getopt without changing OPTIND it will start from where it last ended. If it already parsed first argument, it would want to continue parsing from the second argument, etc. Because there is no second argument the second (or later) time you source your script, there is only -h, getopt will just fail, because it thinks it already parsed -h.
If you want to re-parse arguments in current shell, you need to just reset OPTIND=1. Or start a fresh new shell, which will reset OPTIND to 1 by itself.
If there's a space between the dot and your script name then you're not lauching the script but just sourcing it to your current shell session !
if you want to run your scipt you should do :
# chmod +x ./deploy.sh
# ./deploy.sh -h
if you source it the functtions and variables inside your scipt will be available in your current shell session this could allow you to do things like that :
# . ./deploy.sh
# print_help

Reliable way to require only bash shell in script

I wonder if there are any reliable methods (cross-shell compatible) to require bash as shell interpreter for my script.
For example, I have shell script that can be run only with bash interpreter. Despite of #!/usr/bin/bash at the beginning of my script some smart user/hacker can run it directly with another shell: $ csh script.sh
This can lead to unwanted consequences.
I already thought about testing echo $0 output and exiting with error code but syntax for if statements (as long as for another conditional statements) is different among various shell interpreters. Testing directly for $BASH_VERSION variable is unreliable due to the same limitations.
Are there any cross-shell compatible and reliable way to determine current interpreter?
Thank you!
EDIT: as for now I have the following basic check for compatibility:
### error codes
E_NOTABASH=1
E_OLD_BASH=2
# perform some checks
if [ -z "$BASH_VERSION" ]
then
echo -e "ERROR: this script support only BASH interpreter! Exiting" >&2
exit $E_NOTABASH
fi
if [[ "${BASH_VERSINFO[0]}" -lt 4 ]]
then
echo -e "ERROR: this script needs BASH 4.0 or greater! Your current version is $BASH_VERSION. Exiting" >&2
exit $E_OLD_BASH
fi
Not entirely sure I understand the scope of the question.
A #! /usr/bin/env bash shebang will fail if there's no bash, but to keep it from being explicitly parsed by another shell, um...
How about -
case "$BASH_VERSION" in
4.*) : bash version 4+ so ok ;;
*) echo "please run only with bash v4+. Aborting."
exit 1 ;;
esac
If the syntax works, it is either right or hacked.
If it crashes, you're good. :)
you could check for the parent process id, command respectively
pstree -p $$ | grep -m 1 -oE '^[^\(]+'
or
ps $(ps -o ppid=$$)

getopt erroneously caches arguments

I've created a script in my bash_aliases to make SSH'ing onto servers easier. However, I'm getting some odd behavior that I don't understand. The below script works as you'd expect, except for when it's re-used.
If I use it like this for this first time in a shell, it works exactly as expected:
$>sdev -s myservername
ssh -i ~/.ssh/id_rsa currentuser#myservername.devdomain.com
However, if I run that a second time, without specifying -s|--server, it will use the server name from the last time I ran this, having seemingly cached it:
$>sdev
ssh -i ~/.ssh/id_rsa currentuser#myservername.devdomain.com
It should have exited with an error and output this message: /bin/bash: A server name (-s|--server) is required.
This happens with any of the arguments; that is, if I specify an argument, and then the next time I don't, this method will use the argument from the last time it was supplied.
Obviously, this is not the behavior I want. What's responsible in my script for doing that, and how do I fix it?
#!/bin/bash
sdev() {
getopt --test > /dev/null
if [[ $? -ne 4 ]]; then
echo "`getopt --test` failed in this environment"
exit 1
fi
OPTIONS=u:,k:,p,s:
LONGOPTIONS=user:,key:,prod,server:
# -temporarily store output to be able to check for errors
# -e.g. use “--options” parameter by name to activate quoting/enhanced mode
# -pass arguments only via -- "$#" to separate them correctly
PARSED=$(getopt --options=$OPTIONS --longoptions=$LONGOPTIONS --name "$0" -- "$#")
if [[ $? -ne 0 ]]; then
# e.g. $? == 1
# then getopt has complained about wrong arguments to stdout
exit 2
fi
# read getopt’s output this way to handle the quoting right:
eval set -- "$PARSED"
domain=devdomain
user="$(whoami)"
key=id_rsa
# now enjoy the options in order and nicely split until we see --
while true; do
case "$1" in
-u|--user)
user="$2"
shift 2
;;
-k|--key)
key="$2".pem
shift 2
;;
-p|--prod)
domain=proddomain
shift
;;
-s|--server)
server="$2"
shift 2
;;
--)
shift
break
;;
*)
echo "Programming error"
exit 3
;;
esac
done
if [ -z "$server" ]; then
echo "$0: A server name (-s|--server) is required."
kill -INT $$
fi
echo "ssh -i ~/.ssh/$key.pem $user#$server.$domain.com"
ssh -i ~/.ssh/$key $user#$server.$domain.com
}
server is a global shell variable, so it's shared between runs of the function (as long as they're run in the same shell). That is, when you run sdev -s myservername, it sets the variable server to "myservername". Later, when you run just sdev, it checks to see if $server is empty, finds it's not, and goes ahead and uses it.
Solution: use local variables! Actually, it'd be best to declare all of the variables you use in the function as local; that way, you don't run the risk of interfering with something else that's trying to use the same variable name. I'd also recommend avoiding all-caps variable names (like OPTIONS, LONGOPTIONS, and PARSED) -- there are a bunch of all-caps variables that have special meanings to the shell and/or other programs, and if you use one of those by mistake it can cause weird problems.
Anyway, here's the simple solution: add this near the beginning of the script:
local server=""

Control operator as optional parameter in bash script

I am trying to run a program optionally in the background. Is there a way to pass a control operator optionally. Something like:
if some_condition
bg=&
fi
myprog $bg
However, as I can see, bash is (rightly) treating $bg as an argument to myprog. I am trying to get myprog running in the background.
Is there a way to do this?
How about:
if some_condition
then
myprog &
else
myprog
fi
I'm in favour of James Brown's answer, however, if you're only looking to specify myprog once, then maybe you can use a function to determine if it should run in the background or not. However, this does mean you'll be writing more lines initially, but you could benefit from this if you're potentially running a lot in the background...
run_in_background=false
run_cmd () {
echo "Running $#"
if $run_in_background; then
"$#" &
else
"$#"
fi
}
if [[ "$1" == '--bg' ]]; then
run_in_background=true
shift
fi
run_cmd "$#"
Then you can run it like so:
./script.bash --bg sleep 1
Or if you could run it within the script itself:
... (Continuing from inside the script above)
run_cmd sleep 1
run_cmd echo hello
wait # Waits for background processes to finish
And then you can determine if the commands being passed to run_cmd will be in the foreground or background by either omitting or using the --bg flag.

Parsing mixed arguments in a script bash

I need to implement a script called with mixed (optional and non-optional) arguments for example -
./scriptfile -m "(argument of -m)" file1 -p file2 -u "(argument of -u)"
in a random order. I've read a lot about the getopts builtin command, but I think it doesn't solve my problem. I can't change the order of arguments, so I don't understand how I can read the arguments one by one.
Someone have any ideas?
You should really give a try to getopts, it is designed for that purpose :
Ex :
#!/bin/bash
while getopts ":a:x:" opt; do
case $opt in
a)
echo "-a was triggered with $OPTARG" >&2
;;
x)
echo "-x was triggered with $OPTARG" >&2
;;
\?)
echo "Invalid option: -$OPTARG" >&2
;;
esac
done
Running the script with different switches ordering :
$ bash /tmp/l.sh -a foo -x bar
-a was triggered with foo
-x was triggered with bar
$ bash /tmp/l.sh -x bar -a foo
-x was triggered with bar
-a was triggered with foo
As you can see, there's no problem to change the order of the switches
See http://wiki.bash-hackers.org/howto/getopts_tutorial
Consider using Python and its excellent built-in library argparse. It will support almost any reasonable and conventional command line options, and with less hassle than bash (which is, strangely, a fairly poor language when it comes to command line argument processing).

Resources