sed

Manuals

man sed
info sed
sed FAQ
sed cheatsheet

Tokens

Token	Explanation
[	Beginning of a character group
]	End of a character group
^	Do not match the following characters
*	Repeat the preceding item zero or more times
$	Match the end of line

Commands

In a script

Command	Remark
sed 's/x/ /g'	Replace x with 7 spaces. Do not forget the “ ' ” 's!
sed 's/x/\t\t\t/g'	Relace x with 3 tabs. Do not forget the “ ' ” 's!

On the commandline

Command	Remark
sed 's/somestring//' file.txt	Delete the string somestring from file.txt and and display the output on the standard output i.e. the terminal
sed '10 d' text123.txt	Delete the 10th line in text123.txt and display the output on the standard output i.e. the terminal
sed '/^\./ d' text123.txt	Delete all lines starting with a . (dot) in file text123.txt and display to the standard output
sed --quiet '/somestring/!p' somefile.txt > otherfile.txt	--quiet: suppress automatic printing of pattern space; p: Print the current pattern space. Remove all lines containing somestring from somefile.txt and save the result in otherfile.txt
sed s/text123/TEXT456/g file.txt > outfile.txt	Replace text123 with TEXT456 in file.txt and put the output in outfile.txt. < file.txt also works
echo 'hello.txt' \| sed 's\|txt$\|org\|'	Replace txt with org
echo 'hello.txt' \| sed 's/txt$/org/'	Replace txt with org
echo one.two.three \| sed 's/../-/'	Replace the first any two characters with a '-'. Result: -e.two.three
echo one.two.three \| sed 's/two…/-/'	Replace the 'two' and the three characters following it with a '-'. Result: one.-ree
sed 's/^[ \t]//' sed 's/^[[:space:]]//g'	Remove all spaces on the beginning of a line
sed -i '/^$/d' file.txt	Delete all blank lines including blank lines containing spaces
sed -i "/^ *$/d" file.txt	Delete all lines only containing spaces
sed -i "/^\s*$/d" file.txt	Delete all blank lines including blank lines containing spaces and tabs
sed -i '1 i\text_to_be_put_on_line_1' somefile	Add the text_to_be_put_on_line_1 on line 1 of somefile
sed 's/[^\| \t\n]//g' fille.in.txt > file.out.txt	Remove all characters from a file except spaces and newline characters

Insert / Add

Insert a newline character

Replace the , between the { and } with a new line character in the whole file

cat somefile | sed 's/},{/}\n{/g' > someotherfilename.txt
sed 's/,/"\n"/g' < file1.txt > file2.txt

Add text beginning line

Add text at the beginning of a line

echo line | sed 's/^/sometext/'

Add text end line

Add text at the end of a line

echo line | sed 's/$/sometext/'

Add dashes to a date

echo 20200326 | sed 's/./&-/4' | sed 's/./&-/7'

Probaly this can be done easier. If you know how please let us know
To use this in Calc or Excel paste a column in to a text editor and save the file to /tmp/filename.txt
In a terminal on Linux (on Windows t.b.d.) run

for i in `cat /tmp/filename.txt`; do echo $i | sed 's/./&-/4' | sed 's/./&-/7' >> otherfile.txt; done

Open otherfile.txt and paste the column back in the spreadsheet

Replace

Search and replace examples

A non breaking space

Replace an UTF-8 non braking space (hex: c2a0) with a “;” (semi colon)

sed 's/\xC2\xA0/;/g'

A newline character

file.txt is a file you can make up wich contains multiple lines. The newline character will be replaced by a space
-z --null-data : separate lines by NUL characters
This also works when the input is from a pipe ( | )

sed --null-data 's/\n/ /g' file.txt # Also works when the input is from a pipe ( | )
cat file.txt | sed --null-data 's/\n/ /g'
sed --null-data 's/\n/\<br\>\n/g' file.txt > filewithhtmlbreakline.txt
sed --null-data 's/\n/\<br\>\n/g' < file.txt > filewithhtmlbreakline.txt
sed -z 's/\x0D\x0A/\x0A/g' file.in > file.out # Replace the carriage return linefeed sequence with a linefeed. The -z (--null-data option is obligatory)
sed -z 's/\x0D\x0A/\x0A/g' file.in | sed -z 's/\x0D\x0D/\x0A/g' | sed -z 's/\x0A\x0D/\x0A/g' | sed -z 's/\x0D/\x0A/g' > file.out # To make sure you got them all

On the other hand:
sed is not ment to replace a newline characters since it works on lines, not files. See the sed FAQ 5.10. Why can't I match or delete a newline using the \n escape sequence? Why can't I match 2 or more lines using \n?
How can I replace a newline (\n) using sed
Solution

The x character is a character that is not in the string (the line sed is processing)
Replace a newline character “ \n ” with four tabs “ \t ” in file filename.txt

cat filename.txt | tr '\n' x | sed 's/x/\t\t\t\t/g'

Replace \n with “\n” in filename.txt

cat filename.txt | tr '\n' @ | sed 's/\@/\"\n\"/g' > filename.quotes.txt

Characters on the end of the line

Work in progress

echo -e "one:123: two\none:456: two\none:789: two" | sed 's/\([0-9]\+\)/(\1,0)/g'
echo -e "one:123: two\none:456: two\none:789: two" | sed 's/[0-9]\{1,\}/(&,0)/'
echo -e "one:123one:456one:789" | sed 's/[0-9]\{1,\}/&\n/g'
sed 's/,[^,]*$/,one/' (replace everything from the first comma to the end of the line with 'one')
sed 's/one.*$/two/g' (replace everything from one and all what is following it to the end of the line with 'two')
echo one.two.three | sed 's/..$/-/' (replace the last two characters with '-')
echo -e "one:123 two:456 three:789" | sed 's/\([0-9]\+\)/\1^/g' | tr -d " " | tr "^" '\n'

Print selected lines

numbers.txt

Print lines 3,4 and 5 then quit

sed -n '3,5p' numbers.txt

Output:

3
4
5

Useful links

Delete all characters between two strings
Removing text between two specific strings

Main subjects on this wiki: Linux, Debian, HTML, Microcontrollers, Privacy

RSS
Disclaimer
Privacy statement
Bugs statement
Cookies
Copyright © : 2014 - 2024 Webevaluation.nl and the authors
Changes reserved.

Webevaluation wiki

Table of Contents

sed