extmisc.html

Shall高级编程

	      options this powerful utility takes.</P
><DIV
CLASS="EXAMPLE"
><HR><A
NAME="SELFCOPY"
></A
><P
><B
>Example 15-55. A script that copies itself</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;#!/bin/bash
   2&nbsp;# self-copy.sh
   3&nbsp;
   4&nbsp;# This script copies itself.
   5&nbsp;
   6&nbsp;file_subscript=copy
   7&nbsp;
   8&nbsp;dd if=$0 of=$0.$file_subscript 2&#62;/dev/null
   9&nbsp;# Suppress messages from dd:   ^^^^^^^^^^^
  10&nbsp;
  11&nbsp;exit $?
  12&nbsp;
  13&nbsp;#  A program whose only output is its own source code
  14&nbsp;#+ is called a "quine" per Willard Quine.
  15&nbsp;#  Does this script qualify as a quine?</PRE
></TD
></TR
></TABLE
><HR></DIV
><DIV
CLASS="EXAMPLE"
><HR><A
NAME="EXERCISINGDD"
></A
><P
><B
>Example 15-56. Exercising <I
CLASS="FIRSTTERM"
>dd</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;#!/bin/bash
   2&nbsp;# exercising-dd.sh
   3&nbsp;
   4&nbsp;# Script by Stephane Chazelas.
   5&nbsp;# Somewhat modified by ABS Guide author.
   6&nbsp;
   7&nbsp;infile=$0       # This script.
   8&nbsp;outfile=log.txt # This output file left behind.
   9&nbsp;n=3
  10&nbsp;p=5
  11&nbsp;
  12&nbsp;dd if=$infile of=$outfile bs=1 skip=$((n-1)) count=$((p-n+1)) 2&#62; /dev/null
  13&nbsp;# Extracts characters n to p (3 to 5) from this script.
  14&nbsp;
  15&nbsp;# --------------------------------------------------------
  16&nbsp;
  17&nbsp;echo -n "hello world" | dd cbs=1 conv=unblock 2&#62; /dev/null
  18&nbsp;# Echoes "hello world" vertically.
  19&nbsp;# Why? Newline after each character dd emits.
  20&nbsp;
  21&nbsp;exit 0</PRE
></TD
></TR
></TABLE
><HR></DIV
><P
><A
NAME="DDKEYSTROKES"
></A
></P
><P
>To demonstrate just how versatile <B
CLASS="COMMAND"
>dd</B
> is,
	     let's use it to capture keystrokes.</P
><DIV
CLASS="EXAMPLE"
><HR><A
NAME="DDKEYPRESS"
></A
><P
><B
>Example 15-57. Capturing Keystrokes</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;#!/bin/bash
   2&nbsp;# dd-keypress.sh: Capture keystrokes without needing to press ENTER.
   3&nbsp;
   4&nbsp;
   5&nbsp;keypresses=4                      # Number of keypresses to capture.
   6&nbsp;
   7&nbsp;
   8&nbsp;old_tty_setting=$(stty -g)        # Save old terminal settings.
   9&nbsp;
  10&nbsp;echo "Press $keypresses keys."
  11&nbsp;stty -icanon -echo                # Disable canonical mode.
  12&nbsp;                                  # Disable local echo.
  13&nbsp;keys=$(dd bs=1 count=$keypresses 2&#62; /dev/null)
  14&nbsp;# 'dd' uses stdin, if "if" (input file) not specified.
  15&nbsp;
  16&nbsp;stty "$old_tty_setting"           # Restore old terminal settings.
  17&nbsp;
  18&nbsp;echo "You pressed the \"$keys\" keys."
  19&nbsp;
  20&nbsp;# Thanks, Stephane Chazelas, for showing the way.
  21&nbsp;exit 0</PRE
></TD
></TR
></TABLE
><HR></DIV
><P
><A
NAME="DDRANDOM"
></A
></P
><P
>The <B
CLASS="COMMAND"
>dd</B
> command can do random access on a
	      data stream.

	        <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;echo -n . | dd bs=1 seek=4 of=file conv=notrunc
   2&nbsp;#  The "conv=notrunc" option means that the output file
   3&nbsp;#+ will not be truncated.
   4&nbsp;
   5&nbsp;# Thanks, S.C.</PRE
></TD
></TR
></TABLE
>
            </P
><P
><A
NAME="DDCOPY"
></A
></P
><P
>The <B
CLASS="COMMAND"
>dd</B
> command can copy raw data
	      and disk images to and from devices, such as floppies and
	      tape drives (<A
HREF="contributed-scripts.html#COPYCD"
>Example A-5</A
>). A common use is
	      creating boot floppies.</P
><P
>		<TT
CLASS="USERINPUT"
><B
>dd if=kernel-image of=/dev/fd0H1440</B
></TT
>
             </P
><P
>Similarly, <B
CLASS="COMMAND"
>dd</B
> can copy the entire
	       contents of a floppy, even one formatted with a
	       <SPAN
CLASS="QUOTE"
>"foreign"</SPAN
> OS, to the hard drive as an
	       image file.</P
><P
>		<TT
CLASS="USERINPUT"
><B
>dd if=/dev/fd0 of=/home/bozo/projects/floppy.img</B
></TT
>
             </P
><P
><A
NAME="DDSWAP"
></A
></P
><P
>	      Other applications of <B
CLASS="COMMAND"
>dd</B
> include
	      initializing temporary swap files (<A
HREF="zeros.html#EX73"
>Example 28-2</A
>)
	      and ramdisks (<A
HREF="zeros.html#RAMDISK"
>Example 28-3</A
>). It can even do a
	      low-level copy of an entire hard drive partition, although
	      this is not necessarily recommended.</P
><P
>People (with presumably nothing better to do with
	      their time) are constantly thinking of interesting
	      applications of <B
CLASS="COMMAND"
>dd</B
>.</P
><P
><A
NAME="DDFDEL"
></A
></P
><DIV
CLASS="EXAMPLE"
><HR><A
NAME="BLOTOUT"
></A
><P
><B
>Example 15-58. Securely deleting a file</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;#!/bin/bash
   2&nbsp;# blot-out.sh: Erase "all" traces of a file.
   3&nbsp;
   4&nbsp;#  This script overwrites a target file alternately
   5&nbsp;#+ with random bytes, then zeros before finally deleting it.
   6&nbsp;#  After that, even examining the raw disk sectors by conventional methods
   7&nbsp;#+ will not reveal the original file data.
   8&nbsp;
   9&nbsp;PASSES=7         #  Number of file-shredding passes.
  10&nbsp;                 #  Increasing this slows script execution,
  11&nbsp;                 #+ especially on large target files.
  12&nbsp;BLOCKSIZE=1      #  I/O with /dev/urandom requires unit block size,
  13&nbsp;                 #+ otherwise you get weird results.
  14&nbsp;E_BADARGS=70     #  Various error exit codes.
  15&nbsp;E_NOT_FOUND=71
  16&nbsp;E_CHANGED_MIND=72
  17&nbsp;
  18&nbsp;if [ -z "$1" ]   # No filename specified.
  19&nbsp;then
  20&nbsp;  echo "Usage: `basename $0` filename"
  21&nbsp;  exit $E_BADARGS
  22&nbsp;fi
  23&nbsp;
  24&nbsp;file=$1
  25&nbsp;
  26&nbsp;if [ ! -e "$file" ]
  27&nbsp;then
  28&nbsp;  echo "File \"$file\" not found."
  29&nbsp;  exit $E_NOT_FOUND
  30&nbsp;fi  
  31&nbsp;
  32&nbsp;echo; echo -n "Are you absolutely sure you want to blot out \"$file\" (y/n)? "
  33&nbsp;read answer
  34&nbsp;case "$answer" in
  35&nbsp;[nN]) echo "Changed your mind, huh?"
  36&nbsp;      exit $E_CHANGED_MIND
  37&nbsp;      ;;
  38&nbsp;*)    echo "Blotting out file \"$file\".";;
  39&nbsp;esac
  40&nbsp;
  41&nbsp;
  42&nbsp;flength=$(ls -l "$file" | awk '{print $5}')  # Field 5 is file length.
  43&nbsp;pass_count=1
  44&nbsp;
  45&nbsp;chmod u+w "$file"   # Allow overwriting/deleting the file.
  46&nbsp;
  47&nbsp;echo
  48&nbsp;
  49&nbsp;while [ "$pass_count" -le "$PASSES" ]
  50&nbsp;do
  51&nbsp;  echo "Pass #$pass_count"
  52&nbsp;  sync         # Flush buffers.
  53&nbsp;  dd if=/dev/urandom of=$file bs=$BLOCKSIZE count=$flength
  54&nbsp;               # Fill with random bytes.
  55&nbsp;  sync         # Flush buffers again.
  56&nbsp;  dd if=/dev/zero of=$file bs=$BLOCKSIZE count=$flength
  57&nbsp;               # Fill with zeros.
  58&nbsp;  sync         # Flush buffers yet again.
  59&nbsp;  let "pass_count += 1"
  60&nbsp;  echo
  61&nbsp;done  
  62&nbsp;
  63&nbsp;
  64&nbsp;rm -f $file    # Finally, delete scrambled and shredded file.
  65&nbsp;sync           # Flush buffers a final time.
  66&nbsp;
  67&nbsp;echo "File \"$file\" blotted out and deleted."; echo
  68&nbsp;
  69&nbsp;
  70&nbsp;exit 0
  71&nbsp;
  72&nbsp;#  This is a fairly secure, if inefficient and slow method
  73&nbsp;#+ of thoroughly "shredding" a file.
  74&nbsp;#  The "shred" command, part of the GNU "fileutils" package,
  75&nbsp;#+ does the same thing, although more efficiently.
  76&nbsp;
  77&nbsp;#  The file cannot not be "undeleted" or retrieved by normal methods.
  78&nbsp;#  However . . .
  79&nbsp;#+ this simple method would *not* likely withstand
  80&nbsp;#+ sophisticated forensic analysis.
  81&nbsp;
  82&nbsp;#  This script may not play well with a journaled file system.
  83&nbsp;#  Exercise (difficult): Fix it so it does.
  84&nbsp;
  85&nbsp;
  86&nbsp;
  87&nbsp;#  Tom Vier's "wipe" file-deletion package does a much more thorough job
  88&nbsp;#+ of file shredding than this simple script.
  89&nbsp;#     http://www.ibiblio.org/pub/Linux/utils/file/wipe-2.0.0.tar.bz2
  90&nbsp;
  91&nbsp;#  For an in-depth analysis on the topic of file deletion and security,
  92&nbsp;#+ see Peter Gutmann's paper,
  93&nbsp;#+     "Secure Deletion of Data From Magnetic and Solid-State Memory".
  94&nbsp;#       http://www.cs.auckland.ac.nz/~pgut001/pubs/secure_del.html</PRE
></TD
></TR
></TABLE
><HR></DIV
><P
>See also the <A
HREF="biblio.html#DDLINK"
>dd
	      thread</A
> entry in the <A
HREF="biblio.html#BIBLIOREF"
>bibliography</A
>.</P
></DD
><DT
><A
NAME="ODREF"
></A
><B
CLASS="COMMAND"
>od</B
></DT
><DD
><P
>The <B
CLASS="COMMAND"
>od</B
>, or <I
CLASS="FIRSTTERM"
>octal
	      dump</I
> filter converts input (or files) to octal
	      (base-8) or other bases. This is useful for viewing or
	      processing binary data files or otherwise unreadable system
	      device files, such as <TT
CLASS="FILENAME"
>/dev/urandom</TT
>,
	      and as a filter for binary data.</P
><P
>	      <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;head -c4 /dev/urandom | od -N4 -tu4 | sed -ne '1s/.* //p'
   2&nbsp;# Sample output: 1324725719, 3918166450, 2989231420, etc.
   3&nbsp;
   4&nbsp;# From rnd.sh example script, by St閜hane Chazelas</PRE
></TD
></TR
></TABLE
>
	  </P
><P
>See also <A
HREF="randomvar.html#SEEDINGRANDOM"
>Example 9-31</A
> and <A
HREF="contributed-scripts.html#INSERTIONSORT"
>Example A-38</A
>.</P
></DD
><DT
><A
NAME="HEXDUMPREF"
></A
><B
CLASS="COMMAND"
>hexdump</B
></DT
><DD
><P
>Performs a hexadecimal, octal, decimal, or ASCII
	      dump of a binary file. This command is the rough equivalent
	      of <B
CLASS="COMMAND"
>od</B
>, above, but not nearly as
	      useful. May be used to view the contents of a binary file,
	      in combination with <A
HREF="extmisc.html#DDREF"
>dd</A
> and <A
HREF="filearchiv.html#LESSREF"
>less</A
>.</P
><P
>	  <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;dd if=/bin/ls | hexdump -C | less
   2&nbsp;# The -C option nicely formats the output in tabular form.</PRE
></TD
></TR
></TABLE
>
	    </P
></DD
><DT
><A
NAME="OBJDUMPREF"
></A
><B
CLASS="COMMAND"
>objdump</B
></DT
><DD
><P
>Displays information about an object file or binary
	      executable in either hexadecimal form or as a disassembled
	      listing (with the <TT
CLASS="OPTION"
>-d</TT
> option).</P
><P
>	      <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="SCREEN"
> <TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>objdump -d /bin/ls</B
></TT
>
 <TT
CLASS="COMPUTEROUTPUT"
>/bin/ls:     file format elf32-i386

 Disassembly of section .init:

 080490bc &#60;.init&#62;:
  80490bc:       55                      push   %ebp
  80490bd:       89 e5                   mov    %esp,%ebp
  . . .</TT
>
 	      </PRE
></TD
></TR
></TABLE
>
	    </P
></DD
><DT
><A
NAME="MCOOKIEREF"
></A
><B
CLASS="COMMAND"
>mcookie</B
></DT
><DD
><P
>This command generates a <SPAN
CLASS="QUOTE"
>"magic cookie,"</SPAN
> a
	      128-bit (32-character) pseudorandom hexadecimal number,
	      normally used as an authorization <SPAN
CLASS="QUOTE"
>"signature"</SPAN
>
	      by the X server. This also available for use in a script
	      as a <SPAN
CLASS="QUOTE"
>"quick 'n dirty"</SPAN
> random number.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;random000=$(mcookie)</PRE
></TD
></TR
></TABLE
></P
><P
>Of course, a script could use <A
HREF="filearchiv.html#MD5SUMREF"
>md5</A
> for the same purpose.</P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;# Generate md5 checksum on the script itself.
   2&nbsp;random001=`md5sum $0 | awk '{print $1}'`
   3&nbsp;# Uses 'awk' to strip off the filename.</PRE
></TD
></TR
></TABLE
></P
><P
>The <B
CLASS="COMMAND"
>mcookie</B
> command gives yet another way
	      to generate a <SPAN
CLASS="QUOTE"
>"unique"</SPAN
> filename.</P
><DIV
CLASS="EXAMPLE"
><HR><A
NAME="TEMPFILENAME"
></A
><P
><B
>Example 15-59. Filename generator</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;#!/bin/bash
   2&nbsp;# tempfile-name.sh:  temp filename generator
   3&nbsp;
   4&nbsp;BASE_STR=`mcookie`   # 32-character magic cookie.
   5&nbsp;POS=11               # Arbitrary position in magic cookie string.
   6&nbsp;LEN=5                # Get $LEN consecutive characters.
   7&nbsp;
   8&nbsp;prefix=temp          #  This is, after all, a "temp" file.
   9&nbsp;                     #  For more "uniqueness," generate the
  10&nbsp;                     #+ filename prefix using the same method
  11&nbsp;                     #+ as the suffix, below.
  12&nbsp;
  13&nbsp;suffix=${BASE_STR:POS:LEN}
  14&nbsp;                     #  Extract a 5-character string,
  15&nbsp;                     #+ starting at position 11.
  16&nbsp;
  17&nbsp;temp_filename=$prefix.$suffix
  18&nbsp;                     # Construct the filename.
  19&nbsp;
  20&nbsp;echo "Temp filename = "$temp_filename""
  21&nbsp;
  22&nbsp;# sh tempfile-name.sh
  23&nbsp;# Temp filename = temp.e19ea
  24&nbsp;
  25&nbsp;#  Compare this method of generating "unique" filenames
  26&nbsp;#+ with the 'date' method in ex51.sh.
  27&nbsp;
  28&nbsp;exit 0</PRE
></TD
></TR
></TABLE
><HR></DIV
></DD
><DT
><A
NAME="UNITSREF"
></A
><B
CLASS="COMMAND"
>units</B
></DT
><DD
><P
>This utility converts between different <I
CLASS="FIRSTTERM"
>units
	      of measure</I
>. While normally invoked in interactive
	      mode, <B
CLASS="COMMAND"
>units</B
> may find use in a
	      script.</P