bash script loop breaks [duplicate] - linux

I have the following shell script. The purpose is to loop thru each line of the target file (whose path is the input parameter to the script) and do work against each line. Now, it seems only work with the very first line in the target file and stops after that line got processed. Is there anything wrong with my script?
#!/bin/bash
# SCRIPT: do.sh
# PURPOSE: loop thru the targets
FILENAME=$1
count=0
echo "proceed with $FILENAME"
while read LINE; do
let count++
echo "$count $LINE"
sh ./do_work.sh $LINE
done < $FILENAME
echo "\ntotal $count targets"
In do_work.sh, I run a couple of ssh commands.

The problem is that do_work.sh runs ssh commands and by default ssh reads from stdin which is your input file. As a result, you only see the first line processed, because the command consumes the rest of the file and your while loop terminates.
This happens not just for ssh, but for any command that reads stdin, including mplayer, ffmpeg, HandBrakeCLI, httpie, brew install, and more.
To prevent this, pass the -n option to your ssh command to make it read from /dev/null instead of stdin. Other commands have similar flags, or you can universally use < /dev/null.

A very simple and robust workaround is to change the file descriptor from which the read command receives input.
This is accomplished by two modifications: the -u argument to read, and the redirection operator for < $FILENAME.
In BASH, the default file descriptor values (i.e. values for -u in read) are:
0 = stdin
1 = stdout
2 = stderr
So just choose some other unused file descriptor, like 9 just for fun.
Thus, the following would be the workaround:
while read -u 9 LINE; do
let count++
echo "$count $LINE"
sh ./do_work.sh $LINE
done 9< $FILENAME
Notice the two modifications:
read becomes read -u 9
< $FILENAME becomes 9< $FILENAME
As a best practice, I do this for all while loops I write in BASH.
If you have nested loops using read, use a different file descriptor for each one (9,8,7,...).

More generally, a workaround which isn't specific to ssh is to redirect standard input for any command which might otherwise consume the while loop's input.
while read -r line; do
((count++))
echo "$count $line"
sh ./do_work.sh "$line" </dev/null
done < "$filename"
The addition of </dev/null is the crucial point here, though the corrected quoting is also somewhat important for robustness; see also When to wrap quotes around a shell variable?. You will want to use read -r unless you specifically require the slightly odd legacy behavior you get for backslashes in the input without -r. Finally, avoid upper case for your private variables.
Another workaround of sorts which is somewhat specific to ssh is to make sure any ssh command has its standard input tied up, e.g. by changing
ssh otherhost some commands here
to instead read the commands from a here document, which conveniently (for this particular scenario) ties up the standard input of ssh for the commands:
ssh otherhost <<'____HERE'
some commands here
____HERE

ssh -n option prevents checking the exit status of ssh when using HEREdoc while piping output to another program.
So use of /dev/null as stdin is preferred.
#!/bin/bash
while read ONELINE ; do
ssh ubuntu#host_xyz </dev/null <<EOF 2>&1 | filter_pgm
echo "Hi, $ONELINE. You come here often?"
process_response_pgm
EOF
if [ ${PIPESTATUS[0]} -ne 0 ] ; then
echo "aborting loop"
exit ${PIPESTATUS[0]}
fi
done << input_list.txt

This was happening to me because I had set -e and a grep in a loop was returning with no output (which gives a non-zero error code).

Related

Loop ends prematurely when executing a command via SSH in a Bash function [duplicate]

I have the following shell script. The purpose is to loop thru each line of the target file (whose path is the input parameter to the script) and do work against each line. Now, it seems only work with the very first line in the target file and stops after that line got processed. Is there anything wrong with my script?
#!/bin/bash
# SCRIPT: do.sh
# PURPOSE: loop thru the targets
FILENAME=$1
count=0
echo "proceed with $FILENAME"
while read LINE; do
let count++
echo "$count $LINE"
sh ./do_work.sh $LINE
done < $FILENAME
echo "\ntotal $count targets"
In do_work.sh, I run a couple of ssh commands.
The problem is that do_work.sh runs ssh commands and by default ssh reads from stdin which is your input file. As a result, you only see the first line processed, because the command consumes the rest of the file and your while loop terminates.
This happens not just for ssh, but for any command that reads stdin, including mplayer, ffmpeg, HandBrakeCLI, httpie, brew install, and more.
To prevent this, pass the -n option to your ssh command to make it read from /dev/null instead of stdin. Other commands have similar flags, or you can universally use < /dev/null.
A very simple and robust workaround is to change the file descriptor from which the read command receives input.
This is accomplished by two modifications: the -u argument to read, and the redirection operator for < $FILENAME.
In BASH, the default file descriptor values (i.e. values for -u in read) are:
0 = stdin
1 = stdout
2 = stderr
So just choose some other unused file descriptor, like 9 just for fun.
Thus, the following would be the workaround:
while read -u 9 LINE; do
let count++
echo "$count $LINE"
sh ./do_work.sh $LINE
done 9< $FILENAME
Notice the two modifications:
read becomes read -u 9
< $FILENAME becomes 9< $FILENAME
As a best practice, I do this for all while loops I write in BASH.
If you have nested loops using read, use a different file descriptor for each one (9,8,7,...).
More generally, a workaround which isn't specific to ssh is to redirect standard input for any command which might otherwise consume the while loop's input.
while read -r line; do
((count++))
echo "$count $line"
sh ./do_work.sh "$line" </dev/null
done < "$filename"
The addition of </dev/null is the crucial point here, though the corrected quoting is also somewhat important for robustness; see also When to wrap quotes around a shell variable?. You will want to use read -r unless you specifically require the slightly odd legacy behavior you get for backslashes in the input without -r. Finally, avoid upper case for your private variables.
Another workaround of sorts which is somewhat specific to ssh is to make sure any ssh command has its standard input tied up, e.g. by changing
ssh otherhost some commands here
to instead read the commands from a here document, which conveniently (for this particular scenario) ties up the standard input of ssh for the commands:
ssh otherhost <<'____HERE'
some commands here
____HERE
ssh -n option prevents checking the exit status of ssh when using HEREdoc while piping output to another program.
So use of /dev/null as stdin is preferred.
#!/bin/bash
while read ONELINE ; do
ssh ubuntu#host_xyz </dev/null <<EOF 2>&1 | filter_pgm
echo "Hi, $ONELINE. You come here often?"
process_response_pgm
EOF
if [ ${PIPESTATUS[0]} -ne 0 ] ; then
echo "aborting loop"
exit ${PIPESTATUS[0]}
fi
done << input_list.txt
This was happening to me because I had set -e and a grep in a loop was returning with no output (which gives a non-zero error code).

'read -r' doesn't read beyond first line in a loop that does ssh [duplicate]

I have the following shell script. The purpose is to loop thru each line of the target file (whose path is the input parameter to the script) and do work against each line. Now, it seems only work with the very first line in the target file and stops after that line got processed. Is there anything wrong with my script?
#!/bin/bash
# SCRIPT: do.sh
# PURPOSE: loop thru the targets
FILENAME=$1
count=0
echo "proceed with $FILENAME"
while read LINE; do
let count++
echo "$count $LINE"
sh ./do_work.sh $LINE
done < $FILENAME
echo "\ntotal $count targets"
In do_work.sh, I run a couple of ssh commands.
The problem is that do_work.sh runs ssh commands and by default ssh reads from stdin which is your input file. As a result, you only see the first line processed, because the command consumes the rest of the file and your while loop terminates.
This happens not just for ssh, but for any command that reads stdin, including mplayer, ffmpeg, HandBrakeCLI, httpie, brew install, and more.
To prevent this, pass the -n option to your ssh command to make it read from /dev/null instead of stdin. Other commands have similar flags, or you can universally use < /dev/null.
A very simple and robust workaround is to change the file descriptor from which the read command receives input.
This is accomplished by two modifications: the -u argument to read, and the redirection operator for < $FILENAME.
In BASH, the default file descriptor values (i.e. values for -u in read) are:
0 = stdin
1 = stdout
2 = stderr
So just choose some other unused file descriptor, like 9 just for fun.
Thus, the following would be the workaround:
while read -u 9 LINE; do
let count++
echo "$count $LINE"
sh ./do_work.sh $LINE
done 9< $FILENAME
Notice the two modifications:
read becomes read -u 9
< $FILENAME becomes 9< $FILENAME
As a best practice, I do this for all while loops I write in BASH.
If you have nested loops using read, use a different file descriptor for each one (9,8,7,...).
More generally, a workaround which isn't specific to ssh is to redirect standard input for any command which might otherwise consume the while loop's input.
while read -r line; do
((count++))
echo "$count $line"
sh ./do_work.sh "$line" </dev/null
done < "$filename"
The addition of </dev/null is the crucial point here, though the corrected quoting is also somewhat important for robustness; see also When to wrap quotes around a shell variable?. You will want to use read -r unless you specifically require the slightly odd legacy behavior you get for backslashes in the input without -r. Finally, avoid upper case for your private variables.
Another workaround of sorts which is somewhat specific to ssh is to make sure any ssh command has its standard input tied up, e.g. by changing
ssh otherhost some commands here
to instead read the commands from a here document, which conveniently (for this particular scenario) ties up the standard input of ssh for the commands:
ssh otherhost <<'____HERE'
some commands here
____HERE
ssh -n option prevents checking the exit status of ssh when using HEREdoc while piping output to another program.
So use of /dev/null as stdin is preferred.
#!/bin/bash
while read ONELINE ; do
ssh ubuntu#host_xyz </dev/null <<EOF 2>&1 | filter_pgm
echo "Hi, $ONELINE. You come here often?"
process_response_pgm
EOF
if [ ${PIPESTATUS[0]} -ne 0 ] ; then
echo "aborting loop"
exit ${PIPESTATUS[0]}
fi
done << input_list.txt
This was happening to me because I had set -e and a grep in a loop was returning with no output (which gives a non-zero error code).

How to portable read a text file line by line in bash

For processing a text file in bash line by line, I usually implement a while loop like this:
function doSomething() {
local inputFile="$1"
local fd=""
local line=""
exec {fd}<"$inputFile" # open file
echo "Opened ${inputFile} for read using descriptor ${fd}"
while IFS='' read -r -u $fd line || [[ -n "$line" ]]; do
echo "read = \"$line\""
done
exec {fd}<&- # close file
return 0
}
This works on my Linux but unfortunately not in OSX. For OSX I currently have to change the code to something like this:
exec 3<"$inputFile" # open file
while IFS='' read -r -u 3 line || [[ -n "$line" ]]; do
echo "read = \"$line\""
done
exec 3<&- # close file
But this has the disadvantage, that I have to manage the file descriptor numbers by myself (in the first script, I let bash choose an available file descriptor number).
Did someone have a solution for this which works for both Linux and OSX?
Note that for some reason, I don't want to use piping or I/O redirection to the complete loop like this (because I don't want to execute the loop in a different process):
while IFS='' read -r line || [[ -n "$line" ]]; do
echo "read = \"$line\""
done < "$inputFile"
The last loop will not fork a new process. You can verify that by printing "$BASHPID" in and outside of the loop.
New processes are only created for pipelines. Simple redirections are handled by temporary dups within the bash process.
Feel free to use standard stdin/stdout redirection. It's no more expensive than redirection done with the exec builtin.

pseudo-terminal error will not be allocated because stdin is not a terminal - sudo

There are other threads with this same topic but my issue is unique. I am running a bash script that has a function that sshes to a remote server and runs a sudo command on the remote server. I'm using the ssh -t option to avoid the requiretty issue. The offending line of code works fine as long as it's NOT being called from within the while loop. The while loop basically reads from a csv file on the local server and calls the checkAuthType function:
while read inputline
do
ARRAY=(`echo $inputline | tr ',' ' '`)
HOSTNAME=${ARRAY[0]}
OS_TYPE=${ARRAY[1]}
checkAuthType $HOSTNAME $OS_TYPE
<more irrelevant code>
done < configfile.csv
This is the function that sits at the top of the script (outside of any while loops):
function checkAuthType()
{
if [ $2 == linux ]; then
LINE=`ssh -t $1 'sudo grep "PasswordAuthentication" /etc/ssh/sshd_config | grep -v "yes\|Yes\|#"'`
fi
if [ $2 == unix ]; then
LINE=`ssh -n $1 'grep "PasswordAuthentication" /usr/local/etc/sshd_config | grep -v "yes\|Yes\|#"'`
fi
<more irrelevant code>
}
So, the offending line is the line that has the sudo command within the function. I can change the command to something simple like "sudo ls -l" and I will still get the "stdin is not a terminal" error. I've also tried "ssh -t -t" but to no avail. But if I call the checkAuthType function from outside of the while loop, it works fine. What is it about the while loop that changes the terminal and how do I fix it? Thank you one thousand times in advance.
Another option to try to get around the problem would be to redirect the file to a different file descriptor and force read to read from it instead.
while read inputline <&3
do
ARRAY=(`echo $inputline | tr ',' ' '`)
HOSTNAME=${ARRAY[0]}
OS_TYPE=${ARRAY[1]}
checkAuthType $HOSTNAME $OS_TYPE
<more irrelevant code>
done 3< configfile.csv
I am guessing you are testing with linux. You should try add the -n flag to your (linux) ssh command to avoid having ssh read from stdin - as it normally reads from stdin the while loop is feeding it your csv.
UPDATE
You should (usually) use the -n flag when scripting with SSH, and the flag is typically needed for 'expected behavior' when using a while read-loop. It does not seem to be the main issue here, though.
There are probably other solutions to this, but you could try adding another -t flag to force pseudo-tty allocation when stdin is not a terminal:
ssh -n -t -t
BroSlow's approach with a different file descriptor seems to work! Since the read command reads from fd 3 and not stdin,
ssh and hence sudo still have or get a tty/pty as stdin.
# simple test case
while read line <&3; do
sudo -k
echo "$line"
ssh -t localhost 'sudo ls -ld /'
done 3<&- 3< <(echo 1; sleep 3; echo 2; sleep 3)

Bash script does not continue to read the next line of file

I have a shell script that saves the output of a command that is executed to a CSV file. It reads the command it has to execute from a shell script which is in this format:
ffmpeg -i /home/test/videos/avi/418kb.avi /home/test/videos/done/418kb.flv
ffmpeg -i /home/test/videos/avi/1253kb.avi /home/test/videos/done/1253kb.flv
ffmpeg -i /home/test/videos/avi/2093kb.avi /home/test/videos/done/2093kb.flv
You can see each line is an ffmpeg command. However, the script just executes the first line. Just a minute ago it was doing nearly all of the commands. It was missing half for some reason. I edited the text file that contained the commands and now it will only do the first line. Here is my bash script:
#!/bin/bash
# Shell script utility to read a file line line.
# Once line is read it will run processLine() function
#Function processLine
processLine(){
line="$#"
START=$(date +%s.%N)
eval $line > /dev/null 2>&1
END=$(date +%s.%N)
DIFF=$(echo "$END - $START" | bc)
echo "$line, $START, $END, $DIFF" >> file.csv 2>&1
echo "It took $DIFF seconds"
echo $line
}
# Store file name
FILE=""
# get file name as command line argument
# Else read it from standard input device
if [ "$1" == "" ]; then
FILE="/dev/stdin"
else
FILE="$1"
# make sure file exist and readable
if [ ! -f $FILE ]; then
echo "$FILE : does not exists"
exit 1
elif [ ! -r $FILE ]; then
echo "$FILE: can not read"
exit 2
fi
fi
# read $FILE using the file descriptors
# Set loop separator to end of line
BAKIFS=$IFS
IFS=$(echo -en "\n\b")
exec 3<&0
exec 0<$FILE
while read line
do
# use $line variable to process line in processLine() function
processLine $line
done
exec 0<&3
# restore $IFS which was used to determine what the field separators are
BAKIFS=$ORIGIFS
exit 0
Thank you for any help.
UPDATE 2
Its the ffmpeg commands rather than the shell script that isn't working. But I should of been using just "\b" as Paul pointed out. I am also making use of Johannes's shorter script.
I think that should do the same and seems to be correct:
#!/bin/bash
CSVFILE=/tmp/file.csv
cat "$#" | while read line; do
echo "Executing '$line'"
START=$(date +%s)
eval $line &> /dev/null
END=$(date +%s)
let DIFF=$END-$START
echo "$line, $START, $END, $DIFF" >> "$CSVFILE"
echo "It took ${DIFF}s"
done
no?
ffmpeg reads STDIN and exhausts it. The solution is to call ffmpeg with:
ffmpeg </dev/null ...
See the detailed explanation here: http://mywiki.wooledge.org/BashFAQ/089
Update:
Since ffmpeg version 1.0, there is also the -nostdin option, so this can be used instead:
ffmpeg -nostdin ...
I just had the same problem.
I believe ffmpeg is responsible for this behaviour.
My solution for this problem:
1) Call ffmpeg with an "&" at the end of your ffmpeg command line
2) Since now the skript will not wait till completion of the ffmpeg process,
we have to prevent our script from starting several ffmpeg processes.
We achieve this goal by delaying the loop pass while there is at least
one running ffmpeg process.
#!/bin/bash
cat FileList.txt |
while read VideoFile; do
<place your ffmpeg command line here> &
FFMPEGStillRunning="true"
while [ "$FFMPEGStillRunning" = "true" ]; do
Process=$(ps -C ffmpeg | grep -o -e "ffmpeg" )
if [ -n "$Process" ]; then
FFMPEGStillRunning="true"
else
FFMPEGStillRunning="false"
fi
sleep 2s
done
done
I would add echos before and after the eval to see what it's about to eval (in case it's treating the whole file as one big long line) and after (in case one of the ffmpeg commands is taking forever).
Unless you are planning to read something from standard input after the loop, you don't need to preserve and restore the original standard input (though it is good to see you know how).
Similarly, I don't see a reason for dinking with IFS at all. There is certainly no need to restore the value of IFS before exit - this is a real shell you are using, not a DOS BAT file.
When you do:
read var1 var2 var3
the shell assigns the first field to $var1, the second to $var2, and the rest of the line to $var3. In the case where there's just one variable - your script, for example - the whole line goes into the variable, just as you want it to.
Inside the process line function, you probably don't want to throw away error output from the executed command. You probably do want to think about checking the exit status of the command. The echo with error redirection is ... unusual, and overkill. If you're sufficiently sure that the commands can't fail, then go ahead with ignoring the error. Is the command 'chatty'; if so, throw away the chat by all means. If not, maybe you don't need to throw away standard output, either.
The script as a whole should probably diagnose when it is given multiple files to process since it ignores the extraneous ones.
You could simplify your file handling by using just:
cat "$#" |
while read line
do
processline "$line"
done
The cat command automatically reports errors (and continues after them) and processes all the input files, or reads standard input if there are no arguments left. The use of double quotes around the variable means that it is passed as a single unit (and therefore unparsed into separate words).
The use of date and bc is interesting - I'd not seen that before.
All in all, I'd be looking at something like:
#!/bin/bash
# Time execution of commands read from a file, line by line.
# Log commands and times to CSV logfile "file.csv"
processLine(){
START=$(date +%s.%N)
eval "$#" > /dev/null
STATUS=$?
END=$(date +%s.%N)
DIFF=$(echo "$END - $START" | bc)
echo "$line, $START, $END, $DIFF, $STATUS" >> file.csv
echo "${DIFF}s: $STATUS: $line"
}
cat "$#" |
while read line
do
processLine "$line"
done

Resources