textproc.html

一本完整的描述Unix Shell 编程的工具书的所有范例

>13  70  447 README</TT
>
 [13 lines  70 words  447 characters]</PRE
></TD
></TR
></TABLE
></P
><P
><TT
CLASS="USERINPUT"
><B
>wc -w</B
></TT
> gives only the word count.</P
><P
><TT
CLASS="USERINPUT"
><B
>wc -l</B
></TT
> gives only the line count.</P
><P
><TT
CLASS="USERINPUT"
><B
>wc -c</B
></TT
> gives only the byte count.</P
><P
><TT
CLASS="USERINPUT"
><B
>wc -m</B
></TT
> gives only the character count.</P
><P
><TT
CLASS="USERINPUT"
><B
>wc -L</B
></TT
> gives only the length of the longest line.</P
><P
>Using <B
CLASS="COMMAND"
>wc</B
> to count how many
	    <I
CLASS="EMPHASIS"
>.txt</I
> files are in current working directory:
	      <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;$ ls *.txt | wc -l
   2&nbsp;# Will work as long as none of the "*.txt" files have a linefeed in their name.
   3&nbsp;
   4&nbsp;# Alternative ways of doing this are:
   5&nbsp;#      find . -maxdepth 1 -name \*.txt -print0 | grep -cz .
   6&nbsp;#      (shopt -s nullglob; set -- *.txt; echo $#)
   7&nbsp;
   8&nbsp;# Thanks, S.C.</PRE
></TD
></TR
></TABLE
>
	    </P
><P
>Using <B
CLASS="COMMAND"
>wc</B
> to total up the size of all the
	      files whose names begin with letters in the range d - h
	      <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="SCREEN"
> <TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>wc [d-h]* | grep total | awk '{print $3}'</B
></TT
>
 <TT
CLASS="COMPUTEROUTPUT"
>71832</TT
>
 	      </PRE
></TD
></TR
></TABLE
>
	    </P
><P
>Using <B
CLASS="COMMAND"
>wc</B
> to count the instances of the
	      word <SPAN
CLASS="QUOTE"
>"Linux"</SPAN
> in the main source file for
	      this book.
	      <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="SCREEN"
> <TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep Linux abs-book.sgml | wc -l</B
></TT
>
 <TT
CLASS="COMPUTEROUTPUT"
>50</TT
>
 	      </PRE
></TD
></TR
></TABLE
>
	    </P
><P
>See also <A
HREF="filearchiv.html#EX52"
>Example 12-35</A
> and <A
HREF="redircb.html#REDIR4"
>Example 16-8</A
>.</P
><P
>Certain commands include some of the
	      functionality of <B
CLASS="COMMAND"
>wc</B
> as options.
	      
	    <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;... | grep foo | wc -l
   2&nbsp;# This frequently used construct can be more concisely rendered.
   3&nbsp;
   4&nbsp;... | grep -c foo
   5&nbsp;# Just use the "-c" (or "--count") option of grep.
   6&nbsp;
   7&nbsp;# Thanks, S.C.</PRE
></TD
></TR
></TABLE
></P
></DD
><DT
><A
NAME="TRREF"
></A
><B
CLASS="COMMAND"
>tr</B
></DT
><DD
><P
>character translation filter.</P
><DIV
CLASS="CAUTION"
><TABLE
CLASS="CAUTION"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="common/caution.png"
HSPACE="5"
ALT="Caution"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
><A
HREF="special-chars.html#UCREF"
>Must use quoting and/or
	      brackets</A
>, as appropriate. Quotes prevent the
	      shell from reinterpreting the special characters in
	      <B
CLASS="COMMAND"
>tr</B
> command sequences. Brackets should be
	      quoted to prevent expansion by the shell.  </P
></TD
></TR
></TABLE
></DIV
><P
>Either <TT
CLASS="USERINPUT"
><B
>tr "A-Z" "*" &#60;filename</B
></TT
>
	      or <TT
CLASS="USERINPUT"
><B
>tr A-Z \* &#60;filename</B
></TT
> changes
	      all the uppercase letters in <TT
CLASS="FILENAME"
>filename</TT
>
	      to asterisks (writes to <TT
CLASS="FILENAME"
>stdout</TT
>).
	      On some systems this may not work, but <TT
CLASS="USERINPUT"
><B
>tr A-Z
	      '[**]'</B
></TT
> will.</P
><P
>The <TT
CLASS="OPTION"
>-d</TT
> option deletes a range of
	      characters.
	    <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;echo "abcdef"                 # abcdef
   2&nbsp;echo "abcdef" | tr -d b-d     # aef
   3&nbsp;
   4&nbsp;
   5&nbsp;tr -d 0-9 &#60;filename
   6&nbsp;# Deletes all digits from the file "filename".</PRE
></TD
></TR
></TABLE
></P
><P
>The <TT
CLASS="OPTION"
>--squeeze-repeats</TT
> (or
              <TT
CLASS="OPTION"
>-s</TT
>) option deletes all but the
              first instance of a string of consecutive characters.
              This option is useful for removing excess <A
HREF="special-chars.html#WHITESPACEREF"
>whitespace</A
>.



	      <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="SCREEN"
> <TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo "XXXXX" | tr --squeeze-repeats 'X'</B
></TT
>
 <TT
CLASS="COMPUTEROUTPUT"
>X</TT
></PRE
></TD
></TR
></TABLE
></P
><P
>The <TT
CLASS="OPTION"
>-c</TT
> <SPAN
CLASS="QUOTE"
>"complement"</SPAN
>
	      option <I
CLASS="EMPHASIS"
>inverts</I
> the character set to
	      match. With this option, <B
CLASS="COMMAND"
>tr</B
> acts only
	      upon those characters <I
CLASS="EMPHASIS"
>not</I
> matching
	      the specified set.</P
><P
>	      <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="SCREEN"
> <TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo "acfdeb123" | tr -c b-d +</B
></TT
>
 <TT
CLASS="COMPUTEROUTPUT"
>+c+d+b++++</TT
></PRE
></TD
></TR
></TABLE
>
            </P
><P
>Note that <B
CLASS="COMMAND"
>tr</B
> recognizes <A
HREF="regexp.html#POSIXREF"
>POSIX character classes</A
>.
	         <A
NAME="AEN8254"
HREF="#FTN.AEN8254"
>[1]</A
>
	      </P
><P
>	      <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="SCREEN"
> <TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>echo "abcd2ef1" | tr '[:alpha:]' -</B
></TT
>
 <TT
CLASS="COMPUTEROUTPUT"
>----2--1</TT
>
 	      </PRE
></TD
></TR
></TABLE
>
	    </P
><DIV
CLASS="EXAMPLE"
><HR><A
NAME="EX49"
></A
><P
><B
>Example 12-18. <B
CLASS="COMMAND"
>toupper</B
>: Transforms a file to all uppercase.</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;#!/bin/bash
   2&nbsp;# Changes a file to all uppercase.
   3&nbsp;
   4&nbsp;E_BADARGS=65
   5&nbsp;
   6&nbsp;if [ -z "$1" ]  # Standard check for command line arg.
   7&nbsp;then
   8&nbsp;  echo "Usage: `basename $0` filename"
   9&nbsp;  exit $E_BADARGS
  10&nbsp;fi  
  11&nbsp;
  12&nbsp;tr a-z A-Z &#60;"$1"
  13&nbsp;
  14&nbsp;# Same effect as above, but using POSIX character set notation:
  15&nbsp;#        tr '[:lower:]' '[:upper:]' &#60;"$1"
  16&nbsp;# Thanks, S.C.
  17&nbsp;
  18&nbsp;exit 0
  19&nbsp;
  20&nbsp;#  Exercise:
  21&nbsp;#  Rewrite this script to give the option of changing a file
  22&nbsp;#+ to *either* upper or lowercase.</PRE
></TD
></TR
></TABLE
><HR></DIV
><DIV
CLASS="EXAMPLE"
><HR><A
NAME="LOWERCASE"
></A
><P
><B
>Example 12-19. <B
CLASS="COMMAND"
>lowercase</B
>: Changes all filenames in working directory to lowercase.</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;#!/bin/bash
   2&nbsp;#
   3&nbsp;#  Changes every filename in working directory to all lowercase.
   4&nbsp;#
   5&nbsp;#  Inspired by a script of John Dubois,
   6&nbsp;#+ which was translated into Bash by Chet Ramey,
   7&nbsp;#+ and considerably simplified by the author of the ABS Guide.
   8&nbsp;
   9&nbsp;
  10&nbsp;for filename in *                # Traverse all files in directory.
  11&nbsp;do
  12&nbsp;   fname=`basename $filename`
  13&nbsp;   n=`echo $fname | tr A-Z a-z`  # Change name to lowercase.
  14&nbsp;   if [ "$fname" != "$n" ]       # Rename only files not already lowercase.
  15&nbsp;   then
  16&nbsp;     mv $fname $n
  17&nbsp;   fi  
  18&nbsp;done   
  19&nbsp;
  20&nbsp;exit $?
  21&nbsp;
  22&nbsp;
  23&nbsp;# Code below this line will not execute because of "exit".
  24&nbsp;#--------------------------------------------------------#
  25&nbsp;# To run it, delete script above line.
  26&nbsp;
  27&nbsp;# The above script will not work on filenames containing blanks or newlines.
  28&nbsp;# Stephane Chazelas therefore suggests the following alternative:
  29&nbsp;
  30&nbsp;
  31&nbsp;for filename in *    # Not necessary to use basename,
  32&nbsp;                     # since "*" won't return any file containing "/".
  33&nbsp;do n=`echo "$filename/" | tr '[:upper:]' '[:lower:]'`
  34&nbsp;#                             POSIX char set notation.
  35&nbsp;#                    Slash added so that trailing newlines are not
  36&nbsp;#                    removed by command substitution.
  37&nbsp;   # Variable substitution:
  38&nbsp;   n=${n%/}          # Removes trailing slash, added above, from filename.
  39&nbsp;   [[ $filename == $n ]] || mv "$filename" "$n"
  40&nbsp;                     # Checks if filename already lowercase.
  41&nbsp;done
  42&nbsp;
  43&nbsp;exit $?</PRE
></TD
></TR
></TABLE
><HR></DIV
><DIV
CLASS="EXAMPLE"
><HR><A
NAME="DU"
></A
><P
><B
>Example 12-20. <B
CLASS="COMMAND"
>Du</B
>: DOS to UNIX text file conversion.</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;#!/bin/bash
   2&nbsp;# Du.sh: DOS to UNIX text file converter.
   3&nbsp;
   4&nbsp;E_WRONGARGS=65
   5&nbsp;
   6&nbsp;if [ -z "$1" ]
   7&nbsp;then
   8&nbsp;  echo "Usage: `basename $0` filename-to-convert"
   9&nbsp;  exit $E_WRONGARGS
  10&nbsp;fi
  11&nbsp;
  12&nbsp;NEWFILENAME=$1.unx
  13&nbsp;
  14&nbsp;CR='\015'  # Carriage return.
  15&nbsp;           # 015 is octal ASCII code for CR.
  16&nbsp;           # Lines in a DOS text file end in CR-LF.
  17&nbsp;           # Lines in a UNIX text file end in LF only.
  18&nbsp;
  19&nbsp;tr -d $CR &#60; $1 &#62; $NEWFILENAME
  20&nbsp;# Delete CR's and write to new file.
  21&nbsp;
  22&nbsp;echo "Original DOS text file is \"$1\"."
  23&nbsp;echo "Converted UNIX text file is \"$NEWFILENAME\"."
  24&nbsp;
  25&nbsp;exit 0
  26&nbsp;
  27&nbsp;# Exercise:
  28&nbsp;# --------
  29&nbsp;# Change the above script to convert from UNIX to DOS.</PRE
></TD
></TR
></TABLE
><HR></DIV
><DIV
CLASS="EXAMPLE"
><HR><A
NAME="ROT13"
></A
><P
><B
>Example 12-21. <B
CLASS="COMMAND"
>rot13</B
>: rot13, ultra-weak encryption.</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;#!/bin/bash
   2&nbsp;# rot13.sh: Classic rot13 algorithm,
   3&nbsp;#           encryption that might fool a 3-year old.
   4&nbsp;
   5&nbsp;# Usage: ./rot13.sh filename
   6&nbsp;# or     ./rot13.sh &#60;filename
   7&nbsp;# or     ./rot13.sh and supply keyboard input (stdin)
   8&nbsp;
   9&nbsp;cat "$@" | tr 'a-zA-Z' 'n-za-mN-ZA-M'   # "a" goes to "n", "b" to "o", etc.
  10&nbsp;#  The 'cat "$@"' construction
  11&nbsp;#+ permits getting input either from stdin or from files.
  12&nbsp;
  13&nbsp;exit 0</PRE
></TD
></TR
></TABLE
><HR></DIV
><DIV
CLASS="EXAMPLE"
><HR><A
NAME="CRYPTOQUOTE"
></A
><P
><B
>Example 12-22. Generating <SPAN
CLASS="QUOTE"
>"Crypto-Quote"</SPAN
> Puzzles</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;#!/bin/bash
   2&nbsp;# crypto-quote.sh: Encrypt quotes
   3&nbsp;
   4&nbsp;#  Will encrypt famous quotes in a simple monoalphabetic substitution.
   5&nbsp;#  The result is similar to the "Crypto Quote" puzzles
   6&nbsp;#+ seen in the Op Ed pages of the Sunday paper.
   7&nbsp;
   8&nbsp;
   9&nbsp;key=ETAOINSHRDLUBCFGJMQPVWZYXK
  10&nbsp;# The "key" is nothing more than a scrambled alphabet.
  11&nbsp;# Changing the "key" changes the encryption.
  12&nbsp;
  13&nbsp;# The 'cat "$@"' construction gets input either from stdin or from files.
  14&nbsp;# If using stdin, terminate input with a Control-D.
  15&nbsp;# Otherwise, specify filename as command-line parameter.
  16&nbsp;
  17&nbsp;cat "$@" | tr "a-z" "A-Z" | tr "A-Z" "$key"
  18&nbsp;#        |  to uppercase  |     encrypt       
  19&nbsp;# Will work on lowercase, uppercase, or mixed-case quotes.
  20&nbsp;# Passes non-alphabetic characters through unchanged.
  21&nbsp;
  22&nbsp;
  23&nbsp;# Try this script with something like:
  24&nbsp;# "Nothing so needs reforming as other people's habits."
  25&nbsp;# --Mark Twain
  26&nbsp;#
  27&nbsp;# Output is:
  28&nbsp;# "CFPHRCS QF CIIOQ MINFMBRCS EQ FPHIM GIFGUI'Q HETRPQ."
  29&nbsp;# --BEML PZERC
  30&nbsp;
  31&nbsp;# To reverse the encryption:
  32&nbsp;# cat "$@" | tr "$key" "A-Z"
  33&nbsp;
  34&nbsp;
  35&nbsp;#  This simple-minded cipher can be broken by an average 12-year old
  36&nbsp;#+ using only pencil and paper.
  37&nbsp;
  38&nbsp;exit 0
  39&nbsp;
  40&nbsp;#  Exercise:
  41&nbsp;#  --------
  42&nbsp;#  Modify the script so that it will either encrypt or decrypt,
  43&nbsp;#+ depending on command-line argument(s).</PRE
></TD
></TR
></TABLE
><HR></DIV
><TABLE
CLASS="SIDEBAR"
BORDER="1"
CELLPADDING="5"
><TR
><TD
><DIV
CLASS="SIDEBAR"
><A
NAME="AEN8282"
></A
><P
><B
><B
CLASS="COMMAND"
>tr</B
> variants</B
></P
><P
>	    The <B
CLASS="COMMAND"
>tr</B
> utility has two historic
	    variants. The BSD version does not use brackets
	    (<TT
CLASS="USERINPUT"
><B
>tr a-z A-Z</B
></TT
>), but the SysV one does
	    (<TT
CLASS="USERINPUT"
><B
>tr '[a-z]' '[A-Z]'</B
></TT