Chapter 32. Debugging

 

Debugging is twice as hard as writing the code in the first place. Therefore, if you write the code as cleverly as possible, you are, by definition, not smart enough to debug it.

--Brian Kernighan

The Bash shell contains no built-in debugger, and only bare-bones debugging-specific commands and constructs. Syntax errors or outright typos in the script generate cryptic error messages that are often of no help in debugging a non-functional script.


Example 32-1. A buggy script

   1 #!/bin/bash
   2 # ex74.sh
   3 
   4 # This is a buggy script.
   5 # Where, oh where is the error?
   6 
   7 a=37
   8 
   9 if [$a -gt 27 ]
  10 then
  11   echo $a
  12 fi  
  13 
  14 exit $?   # 0! Why?

Output from script:
 ./ex74.sh: [37: command not found
What's wrong with the above script? Hint: after the if.


Example 32-2. Missing keyword

   1 #!/bin/bash
   2 # missing-keyword.sh
   3 # What error message will this script generate? And why?
   4 
   5 for a in 1 2 3
   6 do
   7   echo "$a"
   8 # done     # Required keyword 'done' commented out in line 8.
   9 
  10 exit 0     # Will not exit here!
  11 
  12 # === #
  13 
  14 # From command line, after script terminates:
  15   echo $?    # 2

Output from script:
 missing-keyword.sh: line 10: syntax error: unexpected end of file
 	
Note that the error message does not necessarily reference the line in which the error occurs, but the line where the Bash interpreter finally becomes aware of the error.

Error messages may disregard comment lines in a script when reporting the line number of a syntax error.

What if the script executes, but does not work as expected? This is the all too familiar logic error.


Example 32-3. test24: another buggy script

   1 #!/bin/bash
   2 
   3 #  This script is supposed to delete all filenames in current directory
   4 #+ containing embedded spaces.
   5 #  It doesn't work.
   6 #  Why not?
   7 
   8 
   9 badname=`ls | grep ' '`
  10 
  11 # Try this:
  12 # echo "$badname"
  13 
  14 rm "$badname"
  15 
  16 exit 0

Try to find out what's wrong with Example 32-3 by uncommenting the echo "$badname" line. Echo statements are useful for seeing whether what you expect is actually what you get.

In this particular case, rm "$badname" will not give the desired results because $badname should not be quoted. Placing it in quotes ensures that rm has only one argument (it will match only one filename). A partial fix is to remove to quotes from $badname and to reset $IFS to contain only a newline, IFS=$'\n'. However, there are simpler ways of going about it.
   1 # Correct methods of deleting filenames containing spaces.
   2 rm *\ *
   3 rm *" "*
   4 rm *' '*
   5 # Thank you. S.C.

Summarizing the symptoms of a buggy script,

  1. It bombs with a "syntax error" message, or

  2. It runs, but does not work as expected (logic error).

  3. It runs, works as expected, but has nasty side effects (logic bomb).

Tools for debugging non-working scripts include

  1. Inserting echo statements at critical points in the script to trace the variables, and otherwise give a snapshot of what is going on.

    Tip

    Even better is an echo that echoes only when debug is on.

       1 ### debecho (debug-echo), by Stefano Falsetto ###
       2 ### Will echo passed parameters only if DEBUG is set to a value. ###
       3 debecho () {
       4   if [ ! -z "$DEBUG" ]; then
       5      echo "$1" >&2
       6      #         ^^^ to stderr
       7   fi
       8 }
       9 
      10 DEBUG=on
      11 Whatever=whatnot
      12 debecho $Whatever   # whatnot
      13 
      14 DEBUG=
      15 Whatever=notwhat
      16 debecho $Whatever   # (Will not echo.)

  2. Using the tee filter to check processes or data flows at critical points.

  3. Setting option flags -n -v -x

    sh -n scriptname checks for syntax errors without actually running the script. This is the equivalent of inserting set -n or set -o noexec into the script. Note that certain types of syntax errors can slip past this check.

    sh -v scriptname echoes each command before executing it. This is the equivalent of inserting set -v or set -o verbose in the script.

    The -n and -v flags work well together. sh -nv scriptname gives a verbose syntax check.

    sh -x scriptname echoes the result each command, but in an abbreviated manner. This is the equivalent of inserting set -x or set -o xtrace in the script.

    Inserting set -u or set -o nounset in the script runs it, but gives an unbound variable error message and aborts the script.
       1 set -u   # Or   set -o nounset
       2 
       3 # Setting a variable to null will not trigger the error/abort.
       4 # unset_var=
       5 
       6 echo $unset_var   # Unset (and undeclared) variable.
       7 
       8 echo "Should not echo!"
       9 
      10 # sh t2.sh
      11 # t2.sh: line 6: unset_var: unbound variable

  4. Using an "assert" function to test a variable or condition at critical points in a script. (This is an idea borrowed from C.)


    Example 32-4. Testing a condition with an assert

       1 #!/bin/bash
       2 # assert.sh
       3 
       4 #######################################################################
       5 assert ()                 #  If condition false,
       6 {                         #+ exit from script
       7                           #+ with appropriate error message.
       8   E_PARAM_ERR=98
       9   E_ASSERT_FAILED=99
      10 
      11 
      12   if [ -z "$2" ]          #  Not enough parameters passed
      13   then                    #+ to assert() function.
      14     return $E_PARAM_ERR   #  No damage done.
      15   fi
      16 
      17   lineno=$2
      18 
      19   if [ ! $1 ] 
      20   then
      21     echo "Assertion failed:  \"$1\""
      22     echo "File \"$0\", line $lineno"    # Give name of file and line number.
      23     exit $E_ASSERT_FAILED
      24   # else
      25   #   return
      26   #   and continue executing the script.
      27   fi  
      28 } # Insert a similar assert() function into a script you need to debug.    
      29 #######################################################################
      30 
      31 
      32 a=5
      33 b=4
      34 condition="$a -lt $b"     #  Error message and exit from script.
      35                           #  Try setting "condition" to something else
      36                           #+ and see what happens.
      37 
      38 assert "$condition" $LINENO
      39 # The remainder of the script executes only if the "assert" does not fail.
      40 
      41 
      42 # Some commands.
      43 # Some more commands . . .
      44 echo "This statement echoes only if the \"assert\" does not fail."
      45 # . . .
      46 # More commands . . .
      47 
      48 exit $?

  5. Using the $LINENO variable and the caller builtin.

  6. Trapping at exit.

    The exit command in a script triggers a signal 0, terminating the process, that is, the script itself. [1] It is often useful to trap the exit, forcing a "printout" of variables, for example. The trap must be the first command in the script.

Trapping signals

trap

Specifies an action on receipt of a signal; also useful for debugging.

A simple instance:
   1 trap '' 2
   2 # Ignore interrupt 2 (Control-C), with no action specified. 
   3 
   4 trap 'echo "Control-C disabled."' 2
   5 # Message when Control-C pressed.


Example 32-5. Trapping at exit

   1 #!/bin/bash
   2 # Hunting variables with a trap.
   3 
   4 trap 'echo Variable Listing --- a = $a  b = $b' EXIT
   5 #  EXIT is the name of the signal generated upon exit from a script.
   6 #
   7 #  The command specified by the "trap" doesn't execute until
   8 #+ the appropriate signal is sent.
   9 
  10 echo "This prints before the \"trap\" --"
  11 echo "even though the script sees the \"trap\" first."
  12 echo
  13 
  14 a=39
  15 
  16 b=36
  17 
  18 exit 0
  19 #  Note that commenting out the 'exit' command makes no difference,
  20 #+ since the script exits in any case after running out of commands.


Example 32-6. Cleaning up after Control-C

   1 #!/bin/bash
   2 # logon.sh: A quick 'n dirty script to check whether you are on-line yet.
   3 
   4 umask 177  # Make sure temp files are not world readable.
   5 
   6 
   7 TRUE=1
   8 LOGFILE=/var/log/messages
   9 #  Note that $LOGFILE must be readable
  10 #+ (as root, chmod 644 /var/log/messages).
  11 TEMPFILE=temp.$$
  12 #  Create a "unique" temp file name, using process id of the script.
  13 #     Using 'mktemp' is an alternative.
  14 #     For example:
  15 #     TEMPFILE=`mktemp temp.XXXXXX`
  16 KEYWORD=address
  17 #  At logon, the line "remote IP address xxx.xxx.xxx.xxx"
  18 #                      appended to /var/log/messages.
  19 ONLINE=22
  20 USER_INTERRUPT=13
  21 CHECK_LINES=100
  22 #  How many lines in log file to check.
  23 
  24 trap 'rm -f $TEMPFILE; exit $USER_INTERRUPT' TERM INT
  25 #  Cleans up the temp file if script interrupted by control-c.
  26 
  27 echo
  28 
  29 while [ $TRUE ]  #Endless loop.
  30 do
  31   tail -n $CHECK_LINES $LOGFILE> $TEMPFILE
  32   #  Saves last 100 lines of system log file as temp file.
  33   #  Necessary, since newer kernels generate many log messages at log on.
  34   search=`grep $KEYWORD $TEMPFILE`
  35   #  Checks for presence of the "IP address" phrase,
  36   #+ indicating a successful logon.
  37 
  38   if [ ! -z "$search" ] #  Quotes necessary because of possible spaces.
  39   then
  40      echo "On-line"
  41      rm -f $TEMPFILE    #  Clean up temp file.
  42      exit $ONLINE
  43   else
  44      echo -n "."        #  The -n option to echo suppresses newline,
  45                         #+ so you get continuous rows of dots.
  46   fi
  47 
  48   sleep 1  
  49 done  
  50 
  51 
  52 #  Note: if you change the KEYWORD variable to "Exit",
  53 #+ this script can be used while on-line
  54 #+ to check for an unexpected logoff.
  55 
  56 # Exercise: Change the script, per the above note,
  57 #           and prettify it.
  58 
  59 exit 0
  60 
  61 
  62 # Nick Drage suggests an alternate method:
  63 
  64 while true
  65   do ifconfig ppp0 | grep UP 1> /dev/null && echo "connected" && exit 0
  66   echo -n "."   # Prints dots (.....) until connected.
  67   sleep 2
  68 done
  69 
  70 # Problem: Hitting Control-C to terminate this process may be insufficient.
  71 #+         (Dots may keep on echoing.)
  72 # Exercise: Fix this.
  73 
  74 
  75 
  76 # Stephane Chazelas has yet another alternative:
  77 
  78 CHECK_INTERVAL=1
  79 
  80 while ! tail -n 1 "$LOGFILE" | grep -q "$KEYWORD"
  81 do echo -n .
  82    sleep $CHECK_INTERVAL
  83 done
  84 echo "On-line"
  85 
  86 # Exercise: Discuss the relative strengths and weaknesses
  87 #           of each of these various approaches.


Example 32-7. A Simple Implementation of a Progress Bar

   1 #! /bin/bash
   2 # progress-bar2.sh
   3 # Author: Graham Ewart (with reformatting by ABS Guide author).
   4 # Used in ABS Guide with permission (thanks!).
   5 
   6 # Invoke this script with bash. It doesn't work with sh.
   7 
   8 interval=1
   9 long_interval=10
  10 
  11 {
  12      trap "exit" SIGUSR1
  13      sleep $interval; sleep $interval
  14      while true
  15      do
  16        echo -n '.'     # Use dots.
  17        sleep $interval
  18      done; } &         # Start a progress bar as a background process.
  19 
  20 pid=$!
  21 trap "echo !; kill -USR1 $pid; wait $pid"  EXIT        # To handle ^C.
  22 
  23 echo -n 'Long-running process '
  24 sleep $long_interval
  25 echo ' Finished!'
  26 
  27 kill -USR1 $pid
  28 wait $pid              # Stop the progress bar.
  29 trap EXIT
  30 
  31 exit $?

Note

The DEBUG argument to trap causes a specified action to execute after every command in a script. This permits tracing variables, for example.


Example 32-8. Tracing a variable

   1 #!/bin/bash
   2 
   3 trap 'echo "VARIABLE-TRACE> \$variable = \"$variable\""' DEBUG
   4 # Echoes the value of $variable after every command.
   5 
   6 variable=29; line=$LINENO
   7 
   8 echo "  Just initialized \$variable to $variable in line number $line."
   9 
  10 let "variable *= 3"; line=$LINENO
  11 echo "  Just multiplied \$variable by 3 in line number $line."
  12 
  13 exit 0
  14 
  15 #  The "trap 'command1 . . . command2 . . .' DEBUG" construct is
  16 #+ more appropriate in the context of a complex script,
  17 #+ where inserting multiple "echo $variable" statements might be
  18 #+ awkward and time-consuming.
  19 
  20 # Thanks, Stephane Chazelas for the pointer.
  21 
  22 
  23 Output of script:
  24 
  25 VARIABLE-TRACE> $variable = ""
  26 VARIABLE-TRACE> $variable = "29"
  27   Just initialized $variable to 29.
  28 VARIABLE-TRACE> $variable = "29"
  29 VARIABLE-TRACE> $variable = "87"
  30   Just multiplied $variable by 3.
  31 VARIABLE-TRACE> $variable = "87"

Of course, the trap command has other uses aside from debugging, such as disabling certain keystrokes within a script (see Example A-43).


Example 32-9. Running multiple processes (on an SMP box)

   1 #!/bin/bash
   2 # parent.sh
   3 # Running multiple processes on an SMP box.
   4 # Author: Tedman Eng
   5 
   6 #  This is the first of two scripts,
   7 #+ both of which must be present in the current working directory.
   8 
   9 
  10 
  11 
  12 LIMIT=$1         # Total number of process to start
  13 NUMPROC=4        # Number of concurrent threads (forks?)
  14 PROCID=1         # Starting Process ID
  15 echo "My PID is $$"
  16 
  17 function start_thread() {
  18         if [ $PROCID -le $LIMIT ] ; then
  19                 ./child.sh $PROCID&
  20                 let "PROCID++"
  21         else
  22            echo "Limit reached."
  23            wait
  24            exit
  25         fi
  26 }
  27 
  28 while [ "$NUMPROC" -gt 0 ]; do
  29         start_thread;
  30         let "NUMPROC--"
  31 done
  32 
  33 
  34 while true
  35 do
  36 
  37 trap "start_thread" SIGRTMIN
  38 
  39 done
  40 
  41 exit 0
  42 
  43 
  44 
  45 # ======== Second script follows ========
  46 
  47 
  48 #!/bin/bash
  49 # child.sh
  50 # Running multiple processes on an SMP box.
  51 # This script is called by parent.sh.
  52 # Author: Tedman Eng
  53 
  54 temp=$RANDOM
  55 index=$1
  56 shift
  57 let "temp %= 5"
  58 let "temp += 4"
  59 echo "Starting $index  Time:$temp" "$@"
  60 sleep ${temp}
  61 echo "Ending $index"
  62 kill -s SIGRTMIN $PPID
  63 
  64 exit 0
  65 
  66 
  67 # ======================= SCRIPT AUTHOR'S NOTES ======================= #
  68 #  It's not completely bug free.
  69 #  I ran it with limit = 500 and after the first few hundred iterations,
  70 #+ one of the concurrent threads disappeared!
  71 #  Not sure if this is collisions from trap signals or something else.
  72 #  Once the trap is received, there's a brief moment while executing the
  73 #+ trap handler but before the next trap is set.  During this time, it may
  74 #+ be possible to miss a trap signal, thus miss spawning a child process.
  75 
  76 #  No doubt someone may spot the bug and will be writing 
  77 #+ . . . in the future.
  78 
  79 
  80 
  81 # ===================================================================== #
  82 
  83 
  84 
  85 # ----------------------------------------------------------------------#
  86 
  87 
  88 
  89 #################################################################
  90 # The following is the original script written by Vernia Damiano.
  91 # Unfortunately, it doesn't work properly.
  92 #################################################################
  93 
  94 #!/bin/bash
  95 
  96 #  Must call script with at least one integer parameter
  97 #+ (number of concurrent processes).
  98 #  All other parameters are passed through to the processes started.
  99 
 100 
 101 INDICE=8        # Total number of process to start
 102 TEMPO=5         # Maximum sleep time per process
 103 E_BADARGS=65    # No arg(s) passed to script.
 104 
 105 if [ $# -eq 0 ] # Check for at least one argument passed to script.
 106 then
 107   echo "Usage: `basename $0` number_of_processes [passed params]"
 108   exit $E_BADARGS
 109 fi
 110 
 111 NUMPROC=$1              # Number of concurrent process
 112 shift
 113 PARAMETRI=( "$@" )      # Parameters of each process
 114 
 115 function avvia() {
 116          local temp
 117          local index
 118          temp=$RANDOM
 119          index=$1
 120          shift
 121          let "temp %= $TEMPO"
 122          let "temp += 1"
 123          echo "Starting $index Time:$temp" "$@"
 124          sleep ${temp}
 125          echo "Ending $index"
 126          kill -s SIGRTMIN $$
 127 }
 128 
 129 function parti() {
 130          if [ $INDICE -gt 0 ] ; then
 131               avvia $INDICE "${PARAMETRI[@]}" &
 132                 let "INDICE--"
 133          else
 134                 trap : SIGRTMIN
 135          fi
 136 }
 137 
 138 trap parti SIGRTMIN
 139 
 140 while [ "$NUMPROC" -gt 0 ]; do
 141          parti;
 142          let "NUMPROC--"
 143 done
 144 
 145 wait
 146 trap - SIGRTMIN
 147 
 148 exit $?
 149 
 150 : <<SCRIPT_AUTHOR_COMMENTS
 151 I had the need to run a program, with specified options, on a number of
 152 different files, using a SMP machine. So I thought [I'd] keep running
 153 a specified number of processes and start a new one each time . . . one
 154 of these terminates.
 155 
 156 The "wait" instruction does not help, since it waits for a given process
 157 or *all* process started in background. So I wrote [this] bash script
 158 that can do the job, using the "trap" instruction.
 159   --Vernia Damiano
 160 SCRIPT_AUTHOR_COMMENTS

Note

trap '' SIGNAL (two adjacent apostrophes) disables SIGNAL for the remainder of the script. trap SIGNAL restores the functioning of SIGNAL once more. This is useful to protect a critical portion of a script from an undesirable interrupt.

   1 	trap '' 2  # Signal 2 is Control-C, now disabled.
   2 	command
   3 	command
   4 	command
   5 	trap 2     # Reenables Control-C
   6 	

Notes

[1]

By convention, signal 0 is assigned to exit.