字体:  

How to write a shell script

ELM 发表于: 2007-8-06 23:14 来源: 榆树社区

How to write a shell script
Introduction
A shell is a command line interpretor. It takes commands and executes them. As such, it implements a programming language. The Bourne shell is used to create shell scripts -- ie. programs that are interpreted/executed by the shell. You can write shell scripts with the C-shell; however, this is not covered here.

Creating a Script
Suppose you often type the command

    find . -name file -print

and you'd rather type a simple command, say

    sfind file

Create a shell script

    % cd ~/bin
    % emacs sfind
    % page sfind
    find . -name $1 -print
    % chmod a+x sfind
    % rehash
    % cd /usr/local/bin
    % sfind tcsh
    ./shells/tcsh

Observations
This quick example is far from adequate but some observations:

   1. Shell scripts are simple text files created with an editor.
   2. Shell scripts are marked as executeable

          %chmod a+x sfind

   3. Should be located in your search path and ~/bin should be in your search path.
   4. You likely need to rehash if you're a Csh (tcsh) user (but not again when you login).
   5. Arguments are passed from the command line and referenced. For example, as $1.

#!/bin/sh
All Bourne Shell scripts should begin with the sequence

    #!/bin/sh

From the man page for exec(2):

"On the first line of an interpreter script, following the "#!", is the name of a program which should be used to interpret the contents of the file. For instance, if the first line contains "#! /bin/sh", then the con- tents of the file are executed as a shell script."

You can get away without this, but you shouldn't. All good scripts state the interpretor explicitly. Long ago there was just one (the Bourne Shell) but these days there are many interpretors -- Csh, Ksh, Bash, and others.
Comments
Comments are any text beginning with the pound (#) sign. A comment can start anywhere on a line and continue until the end of the line.
Search Path
All shell scripts should include a search path specifica- tion:

    PATH=/usr/ucb:/usr/bin:/bin; export PATH

A PATH specification is recommended -- often times a script will fail for some people because they have a different or incomplete search path.

The Bourne Shell does not export environment variables to children unless explicitly instructed to do so by using the export command.
Argument Checking
A good shell script should verify that the arguments sup- plied (if any) are correct.

    if [ $# -ne 3 ]; then
         echo 1>&2 Usage: $0 19 Oct 91
         exit 127
    fi

This script requires three arguments and gripes accordingly.

Exit status
All Unix utilities should return an exit status.

    # is the year out of range for me?

    if [ $year -lt 1901  -o  $year -gt 2099 ]; then
         echo 1>&2 Year \"$year\" out of range
         exit 127
    fi

    etc...

    # All done, exit ok

    exit 0

A non-zero exit status indicates an error condition of some sort while a zero exit status indicates things worked as expected.

On BSD systems there's been an attempt to categorize some of the more common exit status codes. See /usr/include/sysexits.h.
Using exit status
Exit codes are important for those who use your code. Many constructs test on the exit status of a command.

The conditional construct is:

    if command; then
         command
    fi

For example,

    if tty -s; then
         echo Enter text end with \^D
    fi

Your code should be written with the expectation that others will use it. Making sure you return a meaningful exit status will help.
Stdin, Stdout, Stderr
Standard input, output, and error are file descriptors 0, 1, and 2. Each has a particular role and should be used accordingly:

    # is the year out of range for me?

    if [ $year -lt 1901  -o  $year -gt 2099 ]; then
         echo 1>&2 Year \"$year\" out of my range
         exit 127
    fi

    etc...

    # ok, you have the number of days since Jan 1, ...

    case `expr $days % 7` in
    0)
         echo Mon;;
    1)
         echo Tue;;

    etc...

Error messages should appear on stderr not on stdout! Output should appear on stdout. As for input/output dialogue:

    # give the fellow a chance to quit

    if tty -s ; then
         echo This will remove all files in $* since ...
         echo $n Ok to procede? $c;      read ans
         case "$ans" in
              n*|N*)
    echo File purge abandoned;
    exit 0   ;;
         esac
         RM="rm -rfi"
    else
         RM="rm -rf"
    fi

Note: this code behaves differently if there's a user to communicate with (ie. if the standard input is a tty rather than a pipe, or file, or etc. See tty(1)).
Language Constructs

      For loop iteration

      Substitute values for variable and perform task:

          for variable in word ...
          do
               command
          done

      For example:

          for i in `cat $LOGS`
          do
                  mv $i $i.$TODAY
                  cp /dev/null $i
                  chmod 664 $i
          done

      Alternatively you may see:

          for variable in word ...; do command; done

    * Case

      Switch to statements depending on pattern match

          case word in
          [ pattern [ | pattern ... ] )
               command ;; ] ...
          esac

      For example:

          case "$year" in

          [0-9][0-9])
                  year=19${year}
                  years=`expr $year - 1901`
                  ;;
          [0-9][0-9][0-9][0-9])
                  years=`expr $year - 1901`
                  ;;
          *)
                  echo 1>&2 Year \"$year\" out of range ...
                  exit 127
                  ;;
          esac

    * Conditional Execution

      Test exit status of command and branch

          if command
          then
               command
          [ else
               command ]
          fi

      For example:

          if [ $# -ne 3 ]; then
                  echo 1>&2 Usage: $0 19 Oct 91
                  exit 127
          fi

      Alternatively you may see:

          if command; then command; [ else command; ] fi

    * While/Until Iteration

      Repeat task while command returns good exit status.

          {while | until} command
          do
               command
          done

      For example:

          # for each argument mentioned, purge that directory

          while [ $# -ge 1 ]; do
                  _purge $1
                  shift
          done

      Alternatively you may see:

          while command; do command; done

    * Variables

      Variables are sequences of letters, digits, or underscores beginning with a letter or underscore. To get the contents of a variable you must prepend the name with a $.

      Numeric variables (eg. like $1, etc.) are positional vari- ables for argument communication.

          o Variable Assignment

            Assign a value to a variable by variable=value. For example:

                PATH=/usr/ucb:/usr/bin:/bin; export PATH

            or

                TODAY=`(set \`date\`; echo $1)`

          o Exporting Variables

            Variables are not exported to children unless explicitly marked.

                # We MUST have a DISPLAY environment variable

                if [ "$DISPLAY" = "" ]; then
                        if tty -s ; then
                 echo "DISPLAY (`hostname`:0.0)? \c";
                 read DISPLAY
                        fi
                        if [ "$DISPLAY" = "" ]; then
                 DISPLAY=`hostname`:0.0
                        fi
                        export DISPLAY
                fi

            Likewise, for variables like the PRINTER which you want hon- ored by lpr(1). From a user's .profile:

                PRINTER=PostScript; export PRINTER

            Note: that the Cshell exports all environment variables.

          o Referencing Variables

            Use $variable (or, if necessary, ${variable}) to reference the value.

                # Most user's have a /bin of their own

                if [ "$USER" != "root" ]; then
                        PATH=$HOME/bin:$PATH
                else
                        PATH=/etc:/usr/etc:$PATH
                fi

            The braces are required for concatenation constructs.

            $p_01

            The value of the variable "p_01".

            ${p}_01

            The value of the variable "p" with "_01" pasted onto the end.

          o Conditional Reference

            ${variable-word}

            If the variable has been set, use it's value, else use word.

            POSTSCRIPT=${POSTSCRIPT-PostScript};
            export POSTSCRIPT

            ${variable:-word}

            If the variable has been set and is not null, use it's value, else use word.

            These are useful constructions for honoring the user envi- ronment. Ie. the user of the script can override variable assignments. Cf. programs like lpr(1) honor the PRINTER environment variable, you can do the same trick with your shell scripts.

            ${variable:?word}

            If variable is set use it's value, else print out word and exit. Useful for bailing out.

          o Arguments

            Command line arguments to shell scripts are positional vari- ables:

            $0, $1, ...

            The command and arguments. With $0 the command and the rest the arguments.

            $#

            The number of arguments.

            $*, $@

            All the arguments as a blank separated string. Watch out for "$*" vs. "$@".
            And, some commands:

            shift

            Shift the postional variables down one and decrement number of arguments.

            set arg arg ...

            Set the positional variables to the argument list.

            Command line parsing uses shift:

                # parse argument list

                while [ $# -ge 1 ]; do
                        case $1 in
                     process arguments...
                        esac
                        shift
                done

            A use of the set command:

                # figure out what day it is

                TODAY=`(set \`date\`; echo $1)`

                cd $SPOOL

                for i in `cat $LOGS`
                do
                        mv $i $i.$TODAY
                        cp /dev/null $i
                        chmod 664 $i
                done

          o Special Variables

            $$

            Current process id. This is very useful for constructing temporary files.

                     tmp=/tmp/cal0$$
                     trap "rm -f $tmp /tmp/cal1$$ /tmp/cal2$$"
                     trap exit 1 2 13 15
                     /usr/lib/calprog >$tmp

            $?

            The exit status of the last command.

                     $command
                     # Run target file if no errors and ...

                     if [ $? -eq 0 ]
                     then
              etc...
                     fi


    * Quotes/Special Characters

      Special characters to terminate words:

            ; & ( ) | ^ < > new-line space tab

      These are for command sequences, background jobs, etc. To quote any of these use a backslash (\) or bracket with quote marks ("" or '').

      Single Quotes

      Within single quotes all characters are quoted -- including the backslash. The result is one word.

               grep :${gid}: /etc/group | awk -F: '{print $1}'

      Double Quotes

      Within double quotes you have variable subsitution (ie. the dollar sign is interpreted) but no file name generation (ie. * and ? are quoted). The result is one word.

               if [ ! "${parent}" ]; then
                 parent=${people}/${group}/${user}
               fi

      Back Quotes

      Back quotes mean run the command and substitute the output.

               if [ "`echo -n`" = "-n" ]; then
                n=""
                c="\c"
               else
                n="-n"
                c=""
               fi

      and

               TODAY=`(set \`date\`; echo $1)`

    * Functions

      Functions are a powerful feature that aren't used often enough. Syntax is

          name ()
          {
               commands
          }

      For example:

          # Purge a directory

          _purge()
          {
                  # there had better be a directory

                  if [ ! -d $1 ]; then
           echo $1: No such directory 1>&2
           return
                  fi

               etc...
          }

      Within a function the positional parmeters $0, $1, etc. are the arguments to the function (not the arguments to the script).

      Within a function use return instead of exit.

      Functions are good for encapsulations. You can pipe, redi- rect input, etc. to functions. For example:

          # deal with a file, add people one at a time

          do_file()
          {
                  while parse_one

                  etc...
          }

          etc...

          # take standard input (or a specified file) and do it.

          if [ "$1" != "" ]; then
                  cat $1 | do_file
          else
                  do_file
          fi

    * Sourcing commands

      You can execute shell scripts from within shell scripts. A couple of choices:

      sh command

      This runs the shell script as a separate shell. For example, on Sun machines in /etc/rc:

               sh /etc/rc.local

      . command

      This runs the shell script from within the current shell script. For example:

               # Read in configuration information
               .  /etc/hostconfig

      What are the virtues of each? What's the difference? The second form is useful for configuration files where environment variable are set for the script. For example:

          for HOST in $HOSTS; do

            # is there a config file for this host?

            if [ -r ${BACKUPHOME}/${HOST} ]; then
      .  ${BACKUPHOME}/${HOST}
            fi
          etc...

      Using configuration files in this manner makes it possible to write scripts that are automatically tailored for differ- ent situations.

Some Tricks

    * Test

      The most powerful command is test(1).

          if test expression; then

               etc...

      and (note the matching bracket argument)

          if [ expression ]; then

               etc...

      On System V machines this is a builtin (check out the com- mand /bin/test).

      On BSD systems (like the Suns) compare the command /usr/bin/test with /usr/bin/[.

      Useful expressions are:

      test { -w, -r, -x, -s, ... } filename

      is file writeable, readable, executeable, empty, etc?

      test n1 { -eq, -ne, -gt, ... } n2

      are numbers equal, not equal, greater than, etc.?

      test s1 { =, != } s2

      Are strings the same or different?

      test cond1 { -o, -a } cond2

      Binary or; binary and; use ! for unary negation.

      For example

          if [ $year -lt 1901  -o  $year -gt 2099 ]; then
               echo 1>&2 Year \"$year\" out of range
               exit 127
          fi

      Learn this command inside out! It does a lot for you.

    * String matching

      The test command provides limited string matching tests. A more powerful trick is to match strings with the case switch.

          # parse argument list

          while [ $# -ge 1 ]; do
                  case $1 in
                  -c*)    rate=`echo $1 | cut -c3-`;;
                  -c)     shift;  rate=$1 ;;
                  -p*)    prefix=`echo $1 | cut -c3-`;;
                  -p)     shift;  prefix=$1 ;;
                  -*)     echo $Usage; exit 1 ;;
                  *)      disks=$*;       break   ;;
                  esac

                  shift

          done

      Of course getopt would work much better.

    * SysV vs BSD echo

      On BSD systems to get a prompt you'd say:

          echo -n Ok to procede?;  read ans

      On SysV systems you'd say:

          echo Ok to procede? \c; read ans

      In an effort to produce portable code we've been using:

          # figure out what kind of echo to use

          if [ "`echo -n`" = "-n" ]; then
                  n="";  c="\c"
          else
                  n="-n";     c=""
          fi

          etc...

          echo $n Ok to procede? $c; read ans

    * Is there a person?

      The Unix tradition is that programs should execute as qui- etly as possible. Especially for pipelines, cron jobs, etc.

      User prompts aren't required if there's no user.

          # If there's a person out there, prod him a bit.

          if tty -s; then
               echo Enter text end with \^D
          fi

      The tradition also extends to output.

          # If the output is to a terminal, be verbose

          if tty -s <&1; then
               verbose=true
          else
               verbose=false
          fi

      Beware: just because stdin is a tty that doesn't mean that stdout is too. User prompts should be directed to the user terminal.

          # If there's a person out there, prod him a bit.

          if tty -s; then
               echo Enter text end with \^D >&0
          fi

      Have you ever had a program stop waiting for keyboard input when the output is directed elsewhere?

    * Creating Input

      We're familiar with redirecting input. For example:

          # take standard input (or a specified file) and do it.

          if [ "$1" != "" ]; then
                  cat $1 | do_file
          else
                  do_file
          fi

      alternatively, redirection from a file:

          # take standard input (or a specified file) and do it.

          if [ "$1" != "" ]; then
                  do_file < $1
          else
                  do_file
          fi

      You can also construct files on the fly.

          rmail bsmtp <<$1@newshost.uwo.ca>
          rcpt to:
          data
          from: <$1@newshost.uwo.ca>
          to:
          Subject: Signon $2

          subscribe $2 Usenet Feeder at UWO
          .
          quit
          EOF

      Note: that variables are expanded in the input.

    * String Manipulations

      One of the more common things you'll need to do is parse strings. Some tricks

          TIME=`date | cut -c12-19`

          TIME=`date | sed 's/.* .* .* \(.*\) .* .*/\1/'`

          TIME=`date | awk '{print $4}'`

          TIME=`set \`date\`; echo $4`

          TIME=`date | (read u v w x y z; echo $x)`

      With some care, redefining the input field separators can help.

          #!/bin/sh
          # convert IP number to in-addr.arpa name

          name()
          {    set `IFS=".";echo $1`
               echo $4.$3.$2.$1.in-addr.arpa
          }

          if [ $# -ne 1 ]; then
               echo 1>&2 Usage: bynum IP-address
               exit 127
          fi

          add=`name $1`

          nslookup < < EOF | grep "$add" | sed 's/.*= //'
          set type=any
          $add
          EOF

    * Debugging

      The shell has a number of flags that make debugging easier:

      sh -n command

      Read the shell script but don't execute the commands. IE. check syntax.

      sh -x command

      Display commands and arguments as they're executed. In a lot of my shell scripts you'll see

          # Uncomment the next line for testing
          # set -x

Based on An Introduction to Shell Programing by:
Reg Quinton
Computing and Communications Services
The University of Western Ontario
London, Ontario N6A 5B7
Canada

Press here to return to the General Unix Software Menu.

From: http://www.hsrl.rutgers.edu/ug/shell_help.html