textproc.html

一本完整的描述Unix Shell 编程的工具书的所有范例

><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="common/tip.png"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>To force <B
CLASS="COMMAND"
>grep</B
> to show the filename
	      when searching only one target file, simply give
	      <TT
CLASS="FILENAME"
>/dev/null</TT
> as the second file.</P
><P
>	      <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="SCREEN"
> <TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep Linux osinfo.txt /dev/null</B
></TT
>
 <TT
CLASS="COMPUTEROUTPUT"
>osinfo.txt:This is a file containing information about Linux.
 osinfo.txt:The GPL governs the distribution of the Linux operating system.</TT
>
 	      </PRE
></TD
></TR
></TABLE
>
	    </P
></TD
></TR
></TABLE
></DIV
><P
>If there is a successful match, <B
CLASS="COMMAND"
>grep</B
>
	      returns an <A
HREF="exit-status.html#EXITSTATUSREF"
>exit status</A
>
	      of 0, which makes it useful in a condition test in a
	      script, especially in combination with the <TT
CLASS="OPTION"
>-q</TT
>
	      option to suppress output.
	        <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;SUCCESS=0                      # if grep lookup succeeds
   2&nbsp;word=Linux
   3&nbsp;filename=data.file
   4&nbsp;
   5&nbsp;grep -q "$word" "$filename"    # The "-q" option causes nothing to echo to stdout.
   6&nbsp;
   7&nbsp;if [ $? -eq $SUCCESS ]
   8&nbsp;# if grep -q "$word" "$filename"   can replace lines 5 - 7.
   9&nbsp;then
  10&nbsp;  echo "$word found in $filename"
  11&nbsp;else
  12&nbsp;  echo "$word not found in $filename"
  13&nbsp;fi</PRE
></TD
></TR
></TABLE
>
            </P
><P
><A
HREF="debugging.html#ONLINE"
>Example 29-6</A
> demonstrates how to use
	      <B
CLASS="COMMAND"
>grep</B
> to search for a word pattern in
	      a system logfile.</P
><DIV
CLASS="EXAMPLE"
><HR><A
NAME="GRP"
></A
><P
><B
>Example 12-15. Emulating <SPAN
CLASS="QUOTE"
>"grep"</SPAN
> in a script</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;#!/bin/bash
   2&nbsp;# grp.sh: Very crude reimplementation of 'grep'.
   3&nbsp;
   4&nbsp;E_BADARGS=65
   5&nbsp;
   6&nbsp;if [ -z "$1" ]    # Check for argument to script.
   7&nbsp;then
   8&nbsp;  echo "Usage: `basename $0` pattern"
   9&nbsp;  exit $E_BADARGS
  10&nbsp;fi  
  11&nbsp;
  12&nbsp;echo
  13&nbsp;
  14&nbsp;for file in *     # Traverse all files in $PWD.
  15&nbsp;do
  16&nbsp;  output=$(sed -n /"$1"/p $file)  # Command substitution.
  17&nbsp;
  18&nbsp;  if [ ! -z "$output" ]           # What happens if "$output" is not quoted?
  19&nbsp;  then
  20&nbsp;    echo -n "$file: "
  21&nbsp;    echo $output
  22&nbsp;  fi              #  sed -ne "/$1/s|^|${file}: |p"  is equivalent to above.
  23&nbsp;
  24&nbsp;  echo
  25&nbsp;done  
  26&nbsp;
  27&nbsp;echo
  28&nbsp;
  29&nbsp;exit 0
  30&nbsp;
  31&nbsp;# Exercises:
  32&nbsp;# ---------
  33&nbsp;# 1) Add newlines to output, if more than one match in any given file.
  34&nbsp;# 2) Add features.</PRE
></TD
></TR
></TABLE
><HR></DIV
><P
>How can <B
CLASS="COMMAND"
>grep</B
> search for two (or
	      more) separate patterns? What if you want
	      <B
CLASS="COMMAND"
>grep</B
> to display all lines in a file
	      or files that contain both <SPAN
CLASS="QUOTE"
>"pattern1"</SPAN
>
	      <I
CLASS="EMPHASIS"
>and</I
> <SPAN
CLASS="QUOTE"
>"pattern2"</SPAN
>?</P
><P
>One method is to <A
HREF="special-chars.html#PIPEREF"
>pipe</A
> the result of <B
CLASS="COMMAND"
>grep
	      pattern1</B
> to <B
CLASS="COMMAND"
>grep pattern2</B
>.</P
><P
>For example, given the following file:</P
><P
>	    <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;# Filename: tstfile
   2&nbsp;
   3&nbsp;This is a sample file.
   4&nbsp;This is an ordinary text file.
   5&nbsp;This file does not contain any unusual text.
   6&nbsp;This file is not unusual.
   7&nbsp;Here is some text.</PRE
></TD
></TR
></TABLE
>
            </P
><P
>Now, let's search this file for lines containing
	      <I
CLASS="EMPHASIS"
>both</I
> <SPAN
CLASS="QUOTE"
>"file"</SPAN
> and
	      <SPAN
CLASS="QUOTE"
>"text"</SPAN
> . . . </P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="SCREEN"
> <TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep file tstfile</B
></TT
>
 <TT
CLASS="COMPUTEROUTPUT"
># Filename: tstfile
 This is a sample file.
 This is an ordinary text file.
 This file does not contain any unusual text.
 This file is not unusual.</TT
>
 
 <TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>grep file tstfile | grep text</B
></TT
>
 <TT
CLASS="COMPUTEROUTPUT"
>This is an ordinary text file.
 This file does not contain any unusual text.</TT
></PRE
></TD
></TR
></TABLE
><P
>--</P
><P
><A
NAME="EGREPREF"
></A
><B
CLASS="COMMAND"
>egrep</B
>
	      - <I
CLASS="EMPHASIS"
>extended grep</I
> - is the same
	      as <B
CLASS="COMMAND"
>grep -E</B
>. This uses a somewhat
	      different, extended set of <A
HREF="regexp.html#REGEXREF"
>Regular
	      Expressions</A
>, which can make the search a bit more
	      flexible.</P
><P
><B
CLASS="COMMAND"
>fgrep</B
> - <I
CLASS="EMPHASIS"
>fast grep</I
>
	      - is the same as <B
CLASS="COMMAND"
>grep -F</B
>. It does
	      a literal string search (no Regular Expressions), which
	      usually speeds things up a bit.</P
><DIV
CLASS="NOTE"
><TABLE
CLASS="NOTE"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="common/note.png"
HSPACE="5"
ALT="Note"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>On some Linux distros, <B
CLASS="COMMAND"
>egrep</B
> and
	      <B
CLASS="COMMAND"
>fgrep</B
> are symbolic links to, or aliases for
	      <B
CLASS="COMMAND"
>grep</B
>, but invoked with the
	      <TT
CLASS="OPTION"
>-E</TT
> and <TT
CLASS="OPTION"
>-F</TT
> options,
	      respectively.</P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="EXAMPLE"
><HR><A
NAME="DICTLOOKUP"
></A
><P
><B
>Example 12-16. Looking up definitions in <I
CLASS="CITETITLE"
>Webster's 1913 Dictionary</I
></B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;#!/bin/bash
   2&nbsp;# dict-lookup.sh
   3&nbsp;
   4&nbsp;#  This script looks up definitions in the 1913 Webster's Dictionary.
   5&nbsp;#  This Public Domain dictionary is available for download
   6&nbsp;#+ from various sites, including
   7&nbsp;#+ Project Gutenberg (http://www.gutenberg.org/etext/247).
   8&nbsp;#
   9&nbsp;#  Convert it from DOS to UNIX format (only LF at end of line)
  10&nbsp;#+ before using it with this script.
  11&nbsp;#  Store the file in plain, uncompressed ASCII.
  12&nbsp;#  Set DEFAULT_DICTFILE variable below to path/filename.
  13&nbsp;
  14&nbsp;
  15&nbsp;E_BADARGS=65
  16&nbsp;MAXCONTEXTLINES=50                        # Maximum number of lines to show.
  17&nbsp;DEFAULT_DICTFILE="/usr/share/dict/webster1913-dict.txt"
  18&nbsp;                                          # Default dictionary file pathname.
  19&nbsp;                                          # Change this as necessary.
  20&nbsp;#  Note:
  21&nbsp;#  ----
  22&nbsp;#  This particular edition of the 1913 Webster's
  23&nbsp;#+ begins each entry with an uppercase letter
  24&nbsp;#+ (lowercase for the remaining characters).
  25&nbsp;#  Only the *very first line* of an entry begins this way,
  26&nbsp;#+ and that's why the search algorithm below works.
  27&nbsp;
  28&nbsp;
  29&nbsp;
  30&nbsp;if [[ -z $(echo "$1" | sed -n '/^[A-Z]/p') ]]
  31&nbsp;#  Must at least specify word to look up, and
  32&nbsp;#+ it must start with an uppercase letter.
  33&nbsp;then
  34&nbsp;  echo "Usage: `basename $0` Word-to-define [dictionary-file]"
  35&nbsp;  echo
  36&nbsp;  echo "Note: Word to look up must start with capital letter,"
  37&nbsp;  echo "with the rest of the word in lowercase."
  38&nbsp;  echo "--------------------------------------------"
  39&nbsp;  echo "Examples: Abandon, Dictionary, Marking, etc."
  40&nbsp;  exit $E_BADARGS
  41&nbsp;fi
  42&nbsp;
  43&nbsp;
  44&nbsp;if [ -z "$2" ]                            #  May specify different dictionary
  45&nbsp;                                          #+ as an argument to this script.
  46&nbsp;then
  47&nbsp;  dictfile=$DEFAULT_DICTFILE
  48&nbsp;else
  49&nbsp;  dictfile="$2"
  50&nbsp;fi
  51&nbsp;
  52&nbsp;# ---------------------------------------------------------
  53&nbsp;Definition=$(fgrep -A $MAXCONTEXTLINES "$1 \\" "$dictfile")
  54&nbsp;#                  Definitions in form "Word \..."
  55&nbsp;#
  56&nbsp;#  And, yes, "fgrep" is fast enough
  57&nbsp;#+ to search even a very large text file.
  58&nbsp;
  59&nbsp;
  60&nbsp;# Now, snip out just the definition block.
  61&nbsp;
  62&nbsp;echo "$Definition" |
  63&nbsp;sed -n '1,/^[A-Z]/p' |
  64&nbsp;#  Print from first line of output
  65&nbsp;#+ to the first line of the next entry.
  66&nbsp;sed '$d' | sed '$d'
  67&nbsp;#  Delete last two lines of output
  68&nbsp;#+ (blank line and first line of next entry).
  69&nbsp;# ---------------------------------------------------------
  70&nbsp;
  71&nbsp;exit 0
  72&nbsp;
  73&nbsp;# Exercises:
  74&nbsp;# ---------
  75&nbsp;# 1)  Modify the script to accept any type of alphabetic input
  76&nbsp;#   + (uppercase, lowercase, mixed case), and convert it
  77&nbsp;#   + to an acceptable format for processing.
  78&nbsp;#
  79&nbsp;# 2)  Convert the script to a GUI application,
  80&nbsp;#   + using something like "gdialog" . . .
  81&nbsp;#     The script will then no longer take its argument(s)
  82&nbsp;#   + from the command line.
  83&nbsp;#
  84&nbsp;# 3)  Modify the script to parse one of the other available
  85&nbsp;#   + Public Domain Dictionaries, such as the U.S. Census Bureau Gazetteer.</PRE
></TD
></TR
></TABLE
><HR></DIV
><P
><B
CLASS="COMMAND"
>agrep</B
> (<I
CLASS="EMPHASIS"
>approximate
	      grep</I
>) extends the capabilities of
	      <B
CLASS="COMMAND"
>grep</B
> to approximate matching. The search
	      string may differ by a specified number of characters
	      from the resulting matches. This utility is not part of
	      the core Linux distribution.</P
><DIV
CLASS="TIP"
><TABLE
CLASS="TIP"
WIDTH="90%"
BORDER="0"
><TR
><TD
WIDTH="25"
ALIGN="CENTER"
VALIGN="TOP"
><IMG
SRC="common/tip.png"
HSPACE="5"
ALT="Tip"></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
><P
>To search compressed files, use
	      <B
CLASS="COMMAND"
>zgrep</B
>, <B
CLASS="COMMAND"
>zegrep</B
>, or
	      <B
CLASS="COMMAND"
>zfgrep</B
>. These also work on non-compressed
	      files, though slower than plain <B
CLASS="COMMAND"
>grep</B
>,
	      <B
CLASS="COMMAND"
>egrep</B
>, <B
CLASS="COMMAND"
>fgrep</B
>.
	      They are handy for searching through a mixed set of files,
	      some compressed, some not.</P
><P
>To search <A
HREF="filearchiv.html#BZIPREF"
>bzipped</A
>
	      files, use <B
CLASS="COMMAND"
>bzgrep</B
>.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><B
CLASS="COMMAND"
>look</B
></DT
><DD
><P
>The command <B
CLASS="COMMAND"
>look</B
> works like
	      <B
CLASS="COMMAND"
>grep</B
>, but does a lookup on
	      a <SPAN
CLASS="QUOTE"
>"dictionary"</SPAN
>, a sorted word list.
	      By default, <B
CLASS="COMMAND"
>look</B
> searches for a match
	      in <TT
CLASS="FILENAME"
>/usr/dict/words</TT
>, but a different
	      dictionary file may be specified.</P
><DIV
CLASS="EXAMPLE"
><HR><A
NAME="LOOKUP"
></A
><P
><B
>Example 12-17. Checking words in a list for validity</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;#!/bin/bash
   2&nbsp;# lookup: Does a dictionary lookup on each word in a data file.
   3&nbsp;
   4&nbsp;file=words.data  # Data file from which to read words to test.
   5&nbsp;
   6&nbsp;echo
   7&nbsp;
   8&nbsp;while [ "$word" != end ]  # Last word in data file.
   9&nbsp;do
  10&nbsp;  read word      # From data file, because of redirection at end of loop.
  11&nbsp;  look $word &#62; /dev/null  # Don't want to display lines in dictionary file.
  12&nbsp;  lookup=$?      # Exit status of 'look' command.
  13&nbsp;
  14&nbsp;  if [ "$lookup" -eq 0 ]
  15&nbsp;  then
  16&nbsp;    echo "\"$word\" is valid."
  17&nbsp;  else
  18&nbsp;    echo "\"$word\" is invalid."
  19&nbsp;  fi  
  20&nbsp;
  21&nbsp;done &#60;"$file"    # Redirects stdin to $file, so "reads" come from there.
  22&nbsp;
  23&nbsp;echo
  24&nbsp;
  25&nbsp;exit 0
  26&nbsp;
  27&nbsp;# ----------------------------------------------------------------
  28&nbsp;# Code below line will not execute because of "exit" command above.
  29&nbsp;
  30&nbsp;
  31&nbsp;# Stephane Chazelas proposes the following, more concise alternative:
  32&nbsp;
  33&nbsp;while read word &#38;&#38; [[ $word != end ]]
  34&nbsp;do if look "$word" &#62; /dev/null
  35&nbsp;   then echo "\"$word\" is valid."
  36&nbsp;   else echo "\"$word\" is invalid."
  37&nbsp;   fi
  38&nbsp;done &#60;"$file"
  39&nbsp;
  40&nbsp;exit 0</PRE
></TD
></TR
></TABLE
><HR></DIV
></DD
><DT
><B
CLASS="COMMAND"
>sed</B
>, <B
CLASS="COMMAND"
>awk</B
></DT
><DD
><P
>Scripting languages especially suited for parsing text
	      files and command output. May be embedded singly or in
	      combination in pipes and shell scripts.</P
></DD
><DT
><B
CLASS="COMMAND"
><A
HREF="sedawk.html#SEDREF"
>sed</A
></B
></DT
><DD
><P
>Non-interactive <SPAN
CLASS="QUOTE"
>"stream editor"</SPAN
>, permits using
	      many <B
CLASS="COMMAND"
>ex</B
> commands in batch mode. It
	      finds many uses in shell scripts.</P
></DD
><DT
><B
CLASS="COMMAND"
><A
HREF="awk.html#AWKREF"
>awk</A
></B
></DT
><DD
><P
>Programmable file extractor and formatter, good for
	      manipulating and/or extracting fields (columns) in
	      structured text files. Its syntax is similar to C.</P
></DD
><DT
><B
CLASS="COMMAND"
>wc</B
></DT
><DD
><P
><I
CLASS="EMPHASIS"
>wc</I
> gives a <SPAN
CLASS="QUOTE"
>"word count"</SPAN
> on a file or I/O stream:
	      <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="SCREEN"
> <TT
CLASS="PROMPT"
>bash $ </TT
><TT
CLASS="USERINPUT"
><B
>wc /usr/share/doc/sed-4.1.2/README</B
></TT
>
 <TT
CLASS="COMPUTEROUTPUT"