Denis Vlasenko | 72cc964 | 2007-11-13 17:13:31 +0000 | [diff] [blame] | 1 | <html> |
| 2 | |
| 3 | <head><title>Command Summary for sed (sed & awk, Second Edition)</title> |
| 4 | </head> |
| 5 | |
| 6 | <body> |
| 7 | |
| 8 | <h2>Command Summary for sed</h2> |
| 9 | |
| 10 | <dl> |
| 11 | |
| 12 | <dt><b>: </b> <b> :</b><em>label</em></dt> |
| 13 | <dd>Label a line in the script for the transfer of control by |
| 14 | <b>b</b> or <b>t</b>. |
| 15 | <em>label</em> may contain up to seven characters. |
| 16 | (The POSIX standard says that an implementation can allow longer |
| 17 | labels if it wishes to. GNU sed allows labels to be of any length.) |
| 18 | </p></dd> |
| 19 | |
| 20 | |
| 21 | <dt><b>=</b> [<em>address</em>]<b>=</b></dt> |
| 22 | <dd>Write to standard output the line number of addressed line.</p></dd> |
| 23 | |
| 24 | |
| 25 | <dt><b>a</b> [<em>address</em>]<b>a\</b></dt> |
| 26 | <dd><em>text</em></p> |
| 27 | |
| 28 | <p>Append <em>text</em> |
| 29 | following each line matched by <em>address</em>. If |
| 30 | <em>text</em> goes over more than one line, newlines |
| 31 | must be "hidden" by preceding them with a backslash. The |
| 32 | <em>text</em> will be terminated by the first |
| 33 | newline that is not hidden in this way. The |
| 34 | <em>text</em> is not available in the pattern space |
| 35 | and subsequent commands cannot be applied to it. The results of this |
| 36 | command are sent to standard output when the list of editing commands |
| 37 | is finished, regardless of what happens to the current line in the |
| 38 | pattern space.</p></dd> |
| 39 | |
| 40 | |
| 41 | <dt><b>b</b> [<em>address1</em>[,<em>address2</em>]]<b>b</b>[<em>label</em>]</dt> |
| 42 | <dd>Transfer control unconditionally (branch) to |
| 43 | <b>:</b><em>label</em> elsewhere in |
| 44 | script. That is, the command following the |
| 45 | <em>label</em> is the next command applied to the |
| 46 | current line. If no <em>label</em> is specified, |
| 47 | control falls through to the end of the script, so no more commands |
| 48 | are applied to the current line.</p></dd> |
| 49 | |
| 50 | |
| 51 | <dt><b>c</b> [<em>address1</em>[,<em>address2</em>]]<b>c\</b></dt> |
| 52 | <dd><em>text</em></p> |
| 53 | |
| 54 | <p>Replace (change) the lines selected by the address with |
| 55 | <em>text</em>. When a range of lines is specified, |
| 56 | all lines as a group are replaced by a single copy of |
| 57 | <em>text</em>. The newline following each line of |
| 58 | <em>text</em> must be escaped by a backslash, except |
| 59 | the last line. The contents of the pattern space are, in effect, |
| 60 | deleted and no subsequent editing commands can be applied to it (or to |
| 61 | <em>text</em>).</p></dd> |
| 62 | |
| 63 | |
| 64 | <dt><b>d</b> [<em>address1</em>[,<em>address2</em>]]<b>d</b></dt> |
| 65 | <dd>Delete line(s) from pattern space. Thus, the line is not passed to standard |
| 66 | output. A new line of input is read and editing resumes with first |
| 67 | command in script.</p></dd> |
| 68 | |
| 69 | |
| 70 | <dt><b>D</b> [<em>address1</em>[,<em>address2</em>]]<b>D</b></dt> |
| 71 | <dd>Delete first part (up to embedded newline) of multiline pattern space created |
| 72 | by <b>N</b> command and resume editing with first command in |
| 73 | script. If this command empties the pattern space, then a new line |
| 74 | of input is read, as if the <b>d</b> command had been executed.</p></dd> |
| 75 | |
| 76 | |
| 77 | <dt><b>g</b> [<em>address1</em>[,<em>address2</em>]]<b>g</b></dt> |
| 78 | <dd>Copy (get) contents of hold space (see <b>h</b> or |
| 79 | <b>H</b> command) into the pattern space, wiping out |
| 80 | previous contents.</p></dd> |
| 81 | |
| 82 | |
| 83 | <dt><b>G</b> [<em>address1</em>[,<em>address2</em>]]<b>G</b></dt> |
| 84 | <dd>Append newline followed by contents of hold space (see |
| 85 | <b>h</b> or <b>H</b> command) to contents of |
| 86 | the pattern space. If hold space is empty, a newline is still |
| 87 | appended to the pattern space.</p></dd> |
| 88 | |
| 89 | |
| 90 | <dt><b>h</b> [<em>address1</em>[,<em>address2</em>]]<b>h</b></dt> |
| 91 | <dd>Copy pattern space into hold space, a special temporary buffer. |
| 92 | Previous contents of hold space are wiped out.</p></dd> |
| 93 | |
| 94 | |
| 95 | <dt><b>H</b> [<em>address1</em>[,<em>address2</em>]]<b>H</b></dt> |
| 96 | <dd>Append newline and contents of pattern space to contents of the hold |
| 97 | space. Even if hold space is empty, this command still appends the |
| 98 | newline first.</p></dd> |
| 99 | |
| 100 | |
| 101 | <dt><b>i</b> [<em>address1</em>]<b>i\</b></dt> |
| 102 | <dd><em>text</em></p> |
| 103 | |
| 104 | <p>Insert <em>text</em> before each line matched by |
| 105 | <em>address</em>. (See <b>a</b> for |
| 106 | details on <em>text</em>.)</p></dd> |
| 107 | |
| 108 | |
| 109 | <dt><b>l</b> [<em>address1</em>[,<em>address2</em>]]<b>l</b></dt> |
| 110 | <dd>List the contents of the pattern space, showing nonprinting characters |
| 111 | as ASCII codes. Long lines are wrapped.</p></dd> |
| 112 | |
| 113 | |
| 114 | <dt><b>n</b> [<em>address1</em>[,<em>address2</em>]]<b>n</b></dt> |
| 115 | <dd>Read next line of input into pattern space. Current line is sent to |
| 116 | standard output. New line becomes current line and increments line |
| 117 | counter. Control passes to command following <b>n</b> |
| 118 | instead of resuming at the top of the script.</p></dd> |
| 119 | |
| 120 | |
| 121 | <dt><b>N</b> [<em>address1</em>[,<em>address2</em>]]<b>N</b></dt> |
| 122 | <dd>Append next input line to contents of pattern space; the new line is |
Denis Vlasenko | f7be20e | 2007-12-24 14:09:19 +0000 | [diff] [blame] | 123 | separated from the previous contents of the pattern space by a newline. |
Denis Vlasenko | 72cc964 | 2007-11-13 17:13:31 +0000 | [diff] [blame] | 124 | (This command is designed to allow pattern matches across two |
| 125 | lines. Using \n to match the embedded newline, you can match |
| 126 | patterns across multiple lines.)</p></dd> |
| 127 | |
| 128 | |
| 129 | <dt><b>p</b> [<em>address1</em>[,<em>address2</em>]]<b>p</b></dt> |
| 130 | <dd>Print the addressed line(s). Note that this can result in duplicate |
| 131 | output unless default output is suppressed by using "#n" or |
| 132 | the <span class="option">-n</span> |
| 133 | |
| 134 | command-line option. Typically used before commands that change flow |
| 135 | control (<b>d</b>, <b>n</b>, |
| 136 | <b>b</b>) and might prevent the current line from being |
| 137 | output.</p></dd> |
| 138 | |
| 139 | |
| 140 | <dt><b>P</b> [<em>address1</em>[,<em>address2</em>]]<b>P</b></dt> |
| 141 | <dd>Print first part (up to embedded newline) of multiline pattern space |
| 142 | created by <b>N</b> command. Same as <b>p</b> |
| 143 | if <b>N</b> has not been applied to a line.</p></dd> |
| 144 | |
| 145 | |
| 146 | <dt><b>q</b> [<em>address</em>]<b>q</b></dt> |
| 147 | <dd>Quit when <em>address</em> is encountered. The |
| 148 | addressed line is first written to output (if default output is not |
| 149 | suppressed), along with any text appended to it by previous |
| 150 | <b>a</b> or <b>r</b> commands.</p></dd> |
| 151 | |
| 152 | |
| 153 | <dt><b>r</b> [<em>address</em>]<b>r</b> <em>file</em></dt> |
| 154 | <dd>Read contents of <em>file</em> and append after the |
| 155 | contents of the pattern space. Exactly one space must be put between |
| 156 | <b>r</b> and the filename.</p></dd> |
| 157 | |
| 158 | |
| 159 | <dt><b>s</b> [<em>address1</em>[,<em>address2</em>]]<b>s</b>/<em>pattern</em>/<em>replacement</em>/[<em>flags</em>]</dt> |
| 160 | <dd>Substitute <em>replacement</em> for |
| 161 | <em>pattern</em> on each addressed line. If pattern |
| 162 | addresses are used, the pattern <b>//</b> represents the |
| 163 | last pattern address specified. The following flags can be specified:</p> |
| 164 | |
| 165 | <dl> |
| 166 | |
| 167 | <dt><b>n</b></dt> |
| 168 | <dd>Replace <em>n</em>th instance of |
| 169 | /<em>pattern</em>/ on each addressed line. |
| 170 | <em>n</em> is any number in the range 1 to 512, and |
| 171 | the default is 1.</p></dd> |
| 172 | |
| 173 | <dt><b>g</b></dt> |
| 174 | <dd>Replace all instances of /<em>pattern</em>/ on each |
| 175 | addressed line, not just the first instance.</p></dd> |
| 176 | |
| 177 | <dt><b>p</b></dt> |
| 178 | <dd>Print the line if a successful substitution is done. If several |
| 179 | successful substitutions are done, multiple copies of the line will be |
| 180 | printed.</p></dd> |
| 181 | |
| 182 | <dt><b>w</b> <em>file</em></dt> |
| 183 | <dd>Write the line to <em>file</em> if a replacement |
| 184 | was done. A maximum of 10 different <em>files</em> can be opened.</p></dd> |
| 185 | |
| 186 | </dl> |
| 187 | |
| 188 | </dd> |
| 189 | |
| 190 | |
| 191 | <dt><b>t</b> [<em>address1</em>[,<em>address2</em>]]<b>t </b>[<em>label</em>]</dt> |
| 192 | <dd>Test if successful substitutions have been made on addressed lines, |
| 193 | and if so, branch to line marked by :<em>label</em>. |
| 194 | (See <b>b</b> and <b>:</b>.) If label is not |
| 195 | specified, control falls through to bottom of script.</p></dd> |
| 196 | |
| 197 | |
| 198 | <dt><b>w</b> [<em>address1</em>[,<em>address2</em>]]<b>w</b> <em>file</em></dt> |
| 199 | <dd>Append contents of pattern space to <em>file</em>. |
| 200 | This action occurs when the command is encountered rather than when |
| 201 | the pattern space is output. Exactly one space must separate the |
| 202 | <b>w</b> and the filename. A maximum of 10 different |
| 203 | files can be opened in a script. This command will create the file if |
| 204 | it does not exist; if the file exists, its contents will be |
| 205 | overwritten each time the script is executed. Multiple write commands |
| 206 | that direct output to the same file append to the end of the file.</p></dd> |
| 207 | |
| 208 | |
| 209 | <dt><b>x</b> [<em>address1</em>[,<em>address2</em>]]<b>x</b></dt> |
| 210 | <dd>Exchange contents of the pattern space with the contents of the hold |
| 211 | space.</p></dd> |
| 212 | |
| 213 | |
| 214 | <dt><b>y</b> [<em>address1</em>[,<em>address2</em>]]<b>y</b>/<em>abc</em>/<em>xyz</em>/</dt> |
| 215 | <dd>Transform each character by position in string |
| 216 | <em>abc</em> to its equivalent in string |
| 217 | <em>xyz</em>.</p></dd> |
| 218 | |
| 219 | |
| 220 | </dl> |