special-chars.html

Shall高级编程

></TR
></TABLE
></P
></TD
></TR
></TABLE
></DIV
><DIV
CLASS="FORMALPARA"
><P
><B
>array initialization. </B
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;Array=(element1 element2 element3)</PRE
></TD
></TR
></TABLE
></P
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>{xxx,yyy,zzz,...}</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>Brace expansion. </B
><A
NAME="BRACEEXPREF"
></A
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;cat {file1,file2,file3} &#62; combined_file
   2&nbsp;# Concatenates the files file1, file2, and file3 into combined_file.
   3&nbsp;
   4&nbsp;
   5&nbsp;cp file22.{txt,backup}
   6&nbsp;# Copies "file22.txt" to "file22.backup"</PRE
></TD
></TR
></TABLE
></P
></DIV
><P
>A command may act upon a comma-separated list of file specs within
	  <TT
CLASS="REPLACEABLE"
><I
>braces</I
></TT
>.

	     <A
NAME="AEN1000"
HREF="#FTN.AEN1000"
>[3]</A
>
	  
	  Filename expansion (<A
HREF="globbingref.html"
>globbing</A
>)
	  applies to the file specs between the braces.</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
>No spaces allowed within the braces
	    <SPAN
CLASS="emphasis"
><I
CLASS="EMPHASIS"
>unless</I
></SPAN
> the spaces are quoted or escaped.</P
><P
><TT
CLASS="USERINPUT"
><B
>echo {file1,file2}\ :{\ A," B",' C'}</B
></TT
></P
><P
><TT
CLASS="COMPUTEROUTPUT"
>file1 : A file1 : B file1 : C file2 : A file2 : B file2 : C</TT
></P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="BRACEEXPREF33"
></A
><SPAN
CLASS="TOKEN"
>{a..z}</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>Extended Brace expansion. </B
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;echo {a..z} # a b c d e f g h i j k l m n o p q r s t u v w x y z
   2&nbsp;# Echoes characters between a and z.
   3&nbsp;
   4&nbsp;echo {0..3} # 0 1 2 3
   5&nbsp;# Echoes characters between 0 and 3.</PRE
></TD
></TR
></TABLE
>
	      </P
></DIV
><P
>The <I
CLASS="FIRSTTERM"
>{a..z}</I
>
	    <A
HREF="bashver3.html#BRACEEXPREF3"
>extended brace
	    expansion</A
> construction is a feature introduced
	    in <A
HREF="bashver3.html#BASH3REF"
>version 3</A
> of
	    <I
CLASS="FIRSTTERM"
>Bash</I
>.</P
></DD
><DT
><A
NAME="CODEBLOCKREF"
></A
><SPAN
CLASS="TOKEN"
>{}</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>Block of code [curly brackets]. </B
>Also referred to as an <I
CLASS="FIRSTTERM"
>inline group</I
>,
		this construct, in effect, creates an <I
CLASS="FIRSTTERM"
>anonymous
		function</I
> (a function without a
		name). However, unlike in a <SPAN
CLASS="QUOTE"
>"standard"</SPAN
> <A
HREF="functions.html#FUNCTIONREF"
>function</A
>, the variables
		inside a code block remain visible to the remainder of the
		script.</P
></DIV
><P
> <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="SCREEN"
> <TT
CLASS="PROMPT"
>bash$ </TT
><TT
CLASS="USERINPUT"
><B
>{ local a;
	      a=123; }</B
></TT
>
 <TT
CLASS="COMPUTEROUTPUT"
>bash: local: can only be used in a
function</TT
>
 	      </PRE
></TD
></TR
></TABLE
> </P
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;a=123
   2&nbsp;{ a=321; }
   3&nbsp;echo "a = $a"   # a = 321   (value inside code block)
   4&nbsp;
   5&nbsp;# Thanks, S.C.</PRE
></TD
></TR
></TABLE
></P
><P
><A
NAME="BLOCKIO"
></A
></P
><P
>The code block enclosed in braces may have <A
HREF="io-redirection.html#IOREDIRREF"
>I/O redirected</A
> to and from
	      it.</P
><DIV
CLASS="EXAMPLE"
><HR><A
NAME="EX8"
></A
><P
><B
>Example 3-1. Code blocks and I/O redirection</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;#!/bin/bash
   2&nbsp;# Reading lines in /etc/fstab.
   3&nbsp;
   4&nbsp;File=/etc/fstab
   5&nbsp;
   6&nbsp;{
   7&nbsp;read line1
   8&nbsp;read line2
   9&nbsp;} &#60; $File
  10&nbsp;
  11&nbsp;echo "First line in $File is:"
  12&nbsp;echo "$line1"
  13&nbsp;echo
  14&nbsp;echo "Second line in $File is:"
  15&nbsp;echo "$line2"
  16&nbsp;
  17&nbsp;exit 0
  18&nbsp;
  19&nbsp;# Now, how do you parse the separate fields of each line?
  20&nbsp;# Hint: use awk, or . . .
  21&nbsp;# . . . Hans-Joerg Diers suggests using the "set" Bash builtin.</PRE
></TD
></TR
></TABLE
><HR></DIV
><P
><A
NAME="BLOCKIO2"
></A
></P
><DIV
CLASS="EXAMPLE"
><HR><A
NAME="RPMCHECK"
></A
><P
><B
>Example 3-2. Saving the output of a code block to a file</B
></P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;#!/bin/bash
   2&nbsp;# rpm-check.sh
   3&nbsp;
   4&nbsp;#  Queries an rpm file for description, listing,
   5&nbsp;#+ and whether it can be installed.
   6&nbsp;#  Saves output to a file.
   7&nbsp;# 
   8&nbsp;#  This script illustrates using a code block.
   9&nbsp;
  10&nbsp;SUCCESS=0
  11&nbsp;E_NOARGS=65
  12&nbsp;
  13&nbsp;if [ -z "$1" ]
  14&nbsp;then
  15&nbsp;  echo "Usage: `basename $0` rpm-file"
  16&nbsp;  exit $E_NOARGS
  17&nbsp;fi  
  18&nbsp;
  19&nbsp;{ # Begin code block.
  20&nbsp;  echo
  21&nbsp;  echo "Archive Description:"
  22&nbsp;  rpm -qpi $1       # Query description.
  23&nbsp;  echo
  24&nbsp;  echo "Archive Listing:"
  25&nbsp;  rpm -qpl $1       # Query listing.
  26&nbsp;  echo
  27&nbsp;  rpm -i --test $1  # Query whether rpm file can be installed.
  28&nbsp;  if [ "$?" -eq $SUCCESS ]
  29&nbsp;  then
  30&nbsp;    echo "$1 can be installed."
  31&nbsp;  else
  32&nbsp;    echo "$1 cannot be installed."
  33&nbsp;  fi  
  34&nbsp;  echo              # End code block.
  35&nbsp;} &#62; "$1.test"       # Redirects output of everything in block to file.
  36&nbsp;
  37&nbsp;echo "Results of rpm test in file $1.test"
  38&nbsp;
  39&nbsp;# See rpm man page for explanation of options.
  40&nbsp;
  41&nbsp;exit 0</PRE
></TD
></TR
></TABLE
><HR></DIV
><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
>Unlike a command group within (parentheses),
	      as above, a code block enclosed by {braces} will
	      <SPAN
CLASS="emphasis"
><I
CLASS="EMPHASIS"
>not</I
></SPAN
> normally launch a <A
HREF="subshells.html#SUBSHELLSREF"
>subshell</A
>.

		<A
NAME="AEN1075"
HREF="#FTN.AEN1075"
>[4]</A
>

	      </P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><SPAN
CLASS="TOKEN"
>{}</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>placeholder for text. </B
>Used after <A
HREF="moreadv.html#XARGSCURLYREF"
>xargs
		<TT
CLASS="OPTION"
>-i</TT
></A
> (<I
CLASS="FIRSTTERM"
>replace
		strings</I
> option). The <SPAN
CLASS="TOKEN"
>{}</SPAN
> double
		curly brackets are a placeholder for output text.</P
></DIV
><P
><TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;ls . | xargs -i -t cp ./{} $1
   2&nbsp;#            ^^         ^^
   3&nbsp;
   4&nbsp;# From "ex42.sh" (copydir.sh) example.</PRE
></TD
></TR
></TABLE
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>{} \;</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>pathname. </B
>Mostly used in <A
HREF="moreadv.html#FINDREF"
>find</A
>
		constructs.  This is <SPAN
CLASS="emphasis"
><I
CLASS="EMPHASIS"
>not</I
></SPAN
> a shell
		<A
HREF="internal.html#BUILTINREF"
>builtin</A
>.</P
></DIV
><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
>The <SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="TOKEN"
>;</SPAN
>"</SPAN
> ends
		the <TT
CLASS="OPTION"
>-exec</TT
> option of a
		<B
CLASS="COMMAND"
>find</B
> command sequence.  It needs
		to be escaped to protect it from interpretation by the
		shell.</P
></TD
></TR
></TABLE
></DIV
></DD
><DT
><A
NAME="LEFTBRACKET"
></A
><SPAN
CLASS="TOKEN"
>[ ]</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>test. </B
><A
NAME="BRACKTEST"
></A
></P
></DIV
><P
><A
HREF="tests.html#IFTHEN"
>Test</A
> expression between
	       <B
CLASS="COMMAND"
>[ ]</B
>.	Note that <B
CLASS="COMMAND"
>[</B
>
	       is part of the shell <I
CLASS="FIRSTTERM"
>builtin</I
> <A
HREF="tests.html#TTESTREF"
>test</A
> (and a synonym for it),
	       <SPAN
CLASS="emphasis"
><I
CLASS="EMPHASIS"
>not</I
></SPAN
> a link to the external command
	       <TT
CLASS="FILENAME"
>/usr/bin/test</TT
>.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>[[ ]]</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>test. </B
></P
></DIV
><P
>Test expression between <SPAN
CLASS="TOKEN"
>[[ ]]</SPAN
>. This is a shell
	       <A
HREF="internal.html#KEYWORDREF"
>keyword</A
>.</P
><P
>See the discussion on the <A
HREF="tests.html#DBLBRACKETS"
>[[ ... ]] construct</A
>.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>[ ]</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>array element. </B
></P
></DIV
><P
>In the context of an <A
HREF="arrays.html#ARRAYREF"
>array</A
>,
	       brackets set off the numbering of each element of that array.
	         <TABLE
BORDER="0"
BGCOLOR="#E0E0E0"
WIDTH="90%"
><TR
><TD
><PRE
CLASS="PROGRAMLISTING"
>   1&nbsp;Array[1]=slot_1
   2&nbsp;echo ${Array[1]}</PRE
></TD
></TR
></TABLE
></P
></DD
><DT
><SPAN
CLASS="TOKEN"
>[ ]</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>range of characters. </B
></P
></DIV
><P
>As part of a <A
HREF="regexp.html#REGEXREF"
>regular
	       expression</A
>, brackets delineate a <A
HREF="regexp.html#BRACKETSREF"
>range of characters</A
> to
	       match.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>(( ))</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
>integer expansion. </B
></P
></DIV
><P
>Expand and evaluate integer expression between
	       <SPAN
CLASS="TOKEN"
>(( ))</SPAN
>.</P
><P
>See the discussion on the <A
HREF="dblparens.html"
>(( ... )) construct</A
>.</P
></DD
><DT
><SPAN
CLASS="TOKEN"
>&#62;</SPAN
> <SPAN
CLASS="TOKEN"
>&#38;&#62;</SPAN
> <SPAN
CLASS="TOKEN"
>&#62;&#38;</SPAN
> <SPAN
CLASS="TOKEN"
>&#62;&#62;</SPAN
> <SPAN
CLASS="TOKEN"
>&#60;</SPAN
> <SPAN
CLASS="TOKEN"
>&#60;&#62;</SPAN
></DT
><DD
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="io-redirection.html#IOREDIRREF"
>redirection</A
>. </B
></P
></DIV
><P
><TT
CLASS="USERINPUT"
><B
>scriptname &#62;filename</B
></TT
> redirects the output of
	      <TT
CLASS="FILENAME"
>scriptname</TT
> to file
	      <TT
CLASS="FILENAME"
>filename</TT
>. Overwrite
	      <TT
CLASS="FILENAME"
>filename</TT
> if it already exists.</P
><P
><A
NAME="REDIROUTERROR"
></A
></P
><P
><TT
CLASS="USERINPUT"
><B
>command &#38;&#62;filename</B
></TT
> redirects
	      both the <A
HREF="ioredirintro.html#STDINOUTDEF"
><TT
CLASS="FILENAME"
>stdout</TT
></A
>
	      and the
	      <TT
CLASS="FILENAME"
>stderr</TT
> of <TT
CLASS="FILENAME"
>command</TT
>
	      to <TT
CLASS="FILENAME"
>filename</TT
>.</P
><P
><A
NAME="REDIROUTERROR2"
></A
></P
><P
><TT
CLASS="USERINPUT"
><B
>command &#62;&#38;2</B
></TT
> redirects
	      <TT
CLASS="FILENAME"
>stdout</TT
> of <TT
CLASS="FILENAME"
>command</TT
>
	      to <TT
CLASS="FILENAME"
>stderr</TT
>.</P
><P
><TT
CLASS="USERINPUT"
><B
>scriptname &#62;&#62;filename</B
></TT
> appends
	      the output of <TT
CLASS="FILENAME"
>scriptname</TT
>
	      to file <TT
CLASS="FILENAME"
>filename</TT
>. If
	      <TT
CLASS="FILENAME"
>filename</TT
> does not already exist,
	      it is created.</P
><P
><A
NAME="REDIRRW"
></A
></P
><P
><TT
CLASS="USERINPUT"
><B
>[i]&#60;&#62;filename</B
></TT
>
               opens file <TT
CLASS="FILENAME"
>filename</TT
> for reading
               and writing, and assigns <A
HREF="io-redirection.html#FDREF"
>file
               descriptor</A
> <SPAN
CLASS="TOKEN"
>i</SPAN
> to it. If
               <TT
CLASS="FILENAME"
>filename</TT
> does not exist, it is
               created.</P
><DIV
CLASS="FORMALPARA"
><P
><B
><A
HREF="process-sub.html#PROCESSSUBREF"
>process substitution</A
>. </B
></P
></DIV
><P
><TT
CLASS="USERINPUT"
><B
>(command)&#62;</B
></TT
></P
><P
><TT
CLASS="USERINPUT"
><B
>&#60;(command)</B
></TT
></P
><P
><A
HREF="comparison-ops.html#LTREF"
>In a different context</A
>,
	      the <SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="TOKEN"
>&#60;</SPAN
>"</SPAN
> and
	      <SPAN
CLASS="QUOTE"
>"<SPAN
CLASS="TOKEN"
>&#62;</SPAN
>"</SPAN
> characters act
	      as <A
HREF="comparison-ops.html#SCOMPARISON1"
>string comparison
	      operators</A
>.</P
><P
><A
HREF="comparison-ops.html#INTLT"
>In yet another context</A
>,