Sed 整理

1. basic options

sed [options] {command} {input-file}

sed -n

sed -n 'p' a.out

打印a.out文件內容

-n: 取消默認打印到屏幕功能

sed -f

sed [options] -f {sed-commands-in-file} {input-file}

test-script.sed

/^root/ p

/^nobody/ p

sed -n -f test-script.sed /etc/passwd

執行test-script.sed中的sed語句

sed -e

sed [options] -e {sed-command-1} -e {sed-command-2} {input-file}

sed -n -e '/^root/ p' -e '/^nobody/ p' /etc/passwd

sed -n \ -e '/^root/ p' \ -e '/^nobody/ p' \ /etc/passwd

逐條執行sed語句（而沒必要把語句卸載文件中，adhoc分析用）

sed '{}'

sed [options] '{

sed-command-1

sed-command-2

}' {input-file}

sed -n '{

逐條執行sed語句（而沒必要把語句卸載文件中，adhoc分析用）

/^root/ p

/^nobody/ p

}' /etc/passwd

sed scripting flow

read

excute

print

repeat

sed 'p'

sed 'p' employee.txt

屏幕上打印每條語句兩次

sed -n 'p' employee.txt

屏幕上打印每條語句一次， 至關於cat

sed range(selective lines to execute)

range by comma

sed -n '2 p' employee.txt

只打印第二行

sed -n '1,4 p' employee.txt

打印第一至四行

sed -n '2,$ p' employee.txt

打印第二行到最後一行

range by plus

sed -n '1,+2 p' employee.txt

從第一行開始，再打印緊接着的兩行

n,+m

range by tilde

n~m

從第n行開始，每一個m行打印

n, n+m, n+2m,n+3m,…

sed -n '1~2p' employee.txt

打印奇數行的內容

Pattern matching

sed -n  '/Jane/ p' employee.txt

輸出包含Jane的行

sed -n '/Jane/,4 p' employee.txt

輸出從第一個包含Jane的行開始，到第四行結束

若是1到4行沒有包含Jane 的行，則輸出後面包含Jane的行

sed -n '/Raj/,$ p' a.txt

輸出從第一個包含Raj的行開始，到最後一行

sed -n '/Raj/,/Jane/ p' a.txt

輸出從第一個包含Raj的行開始，到緊接的包含Jane的行結束

sed -n '/Raj/,+2p' a.txt

輸出從第一個包含Raj的行，及緊接着的兩行

sed 'd'

patten match 的逆運算

sed '2d' a.txt

刪除第二行；結果輸出出第二行之外的全部行

sed '2,$ d' a.txt

刪除第二行至最後一行； 輸出剩下的行

sed '/Manager/ d' a.txt

刪除包含Manager的行；輸出剩下的行

sed '/Jason/,4 d' a.txt

刪除第一個包含Jason的行至第四行

若是前四行中沒有包含Jason的行，則將刪除剩下的行

sed '/Raj/,$ d' a.txt

刪除從第一個包含Raj的行至最後一行

sed '/Raj/, /Jane/, d' a.txt

刪除包含Raj的行至後面第一個包含Jane的行

sed '/Jason/,+2d' a.txt

刪除從包含Jason的行及緊接着的兩行

sed '/^$/ d' a.txt

刪除全部空行

sed '/^#/d' a.txt

刪除全部註釋行

sed 'w'

sed 'w out.txt' a.txt

把a.txt內容寫入out.txt

sed -n 'w out.txt' a.txt

直接寫入，取消將內容打印到屏幕

sed -n '1,4 w out.txt' a.txt

將一至四行，寫入到out.txt 文件

sed -n '2，$ w out.txt' a.txt

2至最後一行 寫入

sed -n '1~2 w out.txt' a.txt

1,3,5,7,..行寫入

sed -n '/Jane/ w out.txt' a.txt

==

sed -n '/Jane/ p' a.txt > out.txt

sed -n 'Jason,4 w out.txt' a.txt

….

sed substitution command

sed '[address-range|pattern-range] s/orginal-string/replacement-string/[substitute-flags]' input-file

sed 's/Manager/Director/' a.txt

用Director替換全部的Manager

sed '/Sales/s/Manager/Director/' a.txt

只對包含Sales的行進行替換

g flag

默認狀況下，替換隻替換每一行中第一次出現的pattern，而不是一行中全部的pattern

g能夠替換全部的pattern

sed 's/a/A/g' a.txt

將全部的小寫字母a替換爲大寫字母A

number flag

sed 's/a/A/2' a.txt

每行的第二個小寫字母a替換爲大寫字母A

print flag

sed -n 's/John/Johnny/p' a.txt

打印全部被替換後的行

w write flag

sed -n 's/John/Johnny/w output.txt' a.txt

i ignore case flag

sed 's/john/Johnny/i' a.txt

e excute flag

execute as shell command

file.txt

/etc/passwd

/etc/group

sed 's/^/ls -l /e' file.txt

==

執行

ls -l /etc/passwd

ls -l /etc/group

sed substition delimeter

sed 's/…/…/' a.txt

sed 's^…^…^' a.txt

sed 's@…@…@' a.txt

sed 's!…!…!' a.txt

multiple sustitute command affecting the same line

list of command

command-1 on the pattern space

command-2 on the newly changed pattern by command-1

….

sed '{

s/Developer/IT Manager/

s/Manager/Director/

}'

a.txt

& get matched pattern

sed 's/^[0-9][0-9][0-9]/[&]/g a.txt

101,…

[101],..

102,..

[102],…

sed 's/^.*/<&>/' a.txt

全部的行，前加<後加>

\1, \2,..single group

sed 's/\([^,]*\).*/\1/g' a.txt

\([^,]*\)

匹配到第一個非，的字符

\1

用第一個匹配進行替換

sed 's/\([^:]\)/\1/g' /etc/passwd

匹配第一個field

echo "The Geek Stuff" | sed 's/\(\b[A-Z]\)/\(\1\)/g'

(T)he (G)eek (S)tuff

Chapter 3

Regular Expression basic

^

beginning of line

$

end of line

.

single character

*

zero or more occurences

\+

one or more occurences

\?

zero or one occurences

[0-9]

character class

|

or operation

sed -n '/101\|102/p' a.txt

{m}

exactly m occurrences

\{m\}

{m,n}

m to n occurrences

\b

word boundary

\n

back reference

sed -e 's/#.*//; /^$/ d' employee.txt

conver Dos file format to Unix file format

sed 's/.$//' filename

Chapter 4

Sed execution

#sed comments begin with #

to find the sed interpreter, sed file begins with this line

#!/bin/sed -f

chmod u+x script.sed

./script.sed input.txt

first line

#!/bin/sed -nf

suppress the output

'-i modifye the input file

sed -i 's/John/Johnny/' a.txt

replace and modify a.txt

sed -ibak 's/John/Johnny/' a.txt

replace and modify a.txt

also save a copy of a.txt as a.txtbak

Chapter 5 Addition Sed Commands

a append line

sed '[address] a the-line-to-append' input-file

sed '2 a 203,Jack Johnson,Engineer' employee.txt

第二行後添加

sed '$ a 106,Jack Johnson,Engineer' employee.txt

最後一行後添加

sed '/Jason/a\

203,Jack Johnson,Engineer\

204,Mark Smith,Sales Engineer' employee.txt

有Jason的行後添加兩行

i-command insert before

sed '[address] i the-line-to-append' input-file

c-command change line, replace it with the specified

sed '[address] c the-line-to-append' input-file

sed '/Jason/ {

a\

204,Jack Johnson,Engineer

i\

202,Mark Smith,Sales Engineer

c\

203,Joe Mason,Sysadmin

}' employee.txt

l: print hidden characters

sed -n l a.txt

=: show  print line number

sed = a.txt

sed '/Jane/ =' a.txt

sed -n '/Raj/ =' a.txt

just show the number of the line containig Raj

sed -n '$ =' a.txt

show the total number of lines

y: transform operation

sed 'y/abcde/ABCDE/' a.txt

字母對應替換： a-A, b-B, c-C, d-D, e-E

multiple inputs in command line

sed -n '/root/p' /etc/passwd /etc/group

q: quit command

r: read from file to print

sed '$ r log.txt' a.txt

print a.txt & log.txt

sed '/Raj/ r log.txt' a.txt

print the first line in a.txt containing Raj

and then print log.txt

simulating unix command

grep -v Jane employee.txt

sed -n '/Jane/ !p' employee.txt

options summary

-n

suprress the default printing

-f

sed command in file

-e

execute sed command

-i

change the original file

-l

sepecify the line length

-l 20

or 'l 20'

n: print current pattern space and fetchs the next line fro mthe input-file

Chapter 6 Sed Hold and Pattern Space Commands

Pattern Space

the internal sed buffer where sed places, and modifies, the line it reas from the input file

Hold Space

additional buffer where sed hold temporary data

support move data from between Pattern Space and Hold Space

but cannot execute sed command on hold space

Contents is retained from one cycle to the next, not deleted between cycles

x:

swap pattern space with hold space

better for the two consecutive lines with one case infor

John Doe

CEO

Jason Smith

IT Manager

Raj Reddy

Sysadmin

sed -n -e '{x;n}' -e '/Manager/ {x;p}' a.txt

h

copy pattern space to hold space

sed -n -e '/Manager/ !h' -e '/Manager/{x;p}' a.txt

/Manager/!h

if line doesn't contain Manager, then copy

H

append pattern space to hold space

sed -n -e '/Manager/!h' -e '/Manager/{H;x;p}' a.txt

after appending, it will be line2\nline1 with in two lines

sed -n -e '/Manager/!h' -e '/Manager/{H;x;s/\n/:/;p}' a.txt

g

copy hold space to pattern space

sed -n -e '/Manager/!h' -e '/Manager/{g;p}'

G

append hold space to pattern space

sed -n -e '/Manager/!h' -e '/Manager/{x;G;s/\n/:/;p}' empnametitle.txt

Chapter 7 Sed Multi-Line Commands and Loops

N

append next line to pattern space

sed -e '{N; s/\n/:/}' a.txt

sed -e '=' employee.txt | sed -e '{N;s/\n/ /}'

print line numbers

D

???????????

loop and branch

:label

defines the label

b label

branches the execution flow to the label

#!/bin/sed -nf

h;n;H;x

s/\n/:/

/Manager/!b end

s/^/*/

:end

p

t

?????????????????///