Sed cheat sheet

Published:: 2025-01-06
Updated:: 2025-01-06
Author:: Jonas Jared Jacek
License:: CC BY-SA 4.0

Cheat sheet for GNU sed, the powerful Unix utility designed for non-interactive text manipulation and automation like substitution, deletion, and transformation. Contains many sed command examples.

Introduction #

sed is an indispensable tool for text processing and automation in Unix-like environments. Its concise syntax and powerful features enable users to perform complex editing tasks efficiently. Whether you are replacing text, deleting lines, or performing intricate transformations, sed provides a versatile solution tailored to a wide range of scripting and command-line needs.

What is `sed`? #

sed, short for stream editor, is a powerful Unix utility used for parsing and transforming text in a pipeline. Unlike interactive text editors, sed processes text non-interactively, making it ideal for automating repetitive editing tasks, especially in shell scripts.

How `sed` Works #

sed reads input line by line (streaming), applies specified editing commands to each line, and outputs the result. It can perform a variety of text manipulations, including substitution, deletion, insertion, and more, using a concise and expressive syntax.

Basic Workflow #

Input: sed takes input from standard input (stdin) or from a file.
Processing: sed applies the editing commands to each line of the input.
Output: The transformed text is sent to standard output (stdout).

Basic Syntax #

sed [OPTIONS] 'SCRIPT' [INPUTFILE...]

OPTIONS: Flags that modify sed behavior (e.g., -e, -f, -i).
SCRIPT: One or more editing commands enclosed in single quotes.
INPUTFILE: Files to be processed. If omitted, sed reads from stdin.

Substitution (`s`) #

Replace occurrences of a pattern with a replacement.

sed 's/pattern/replacement/' file.txt

Possible flags:

g: Global replacement (all occurrences in a line).
i: Case-insensitive matching.
number: Replace the nth occurrence.

Replace first occurrence of “foo” with “bar” in each line #

sed 's/foo/bar/' file.txt

Replace all occurrences of “foo” with “bar” #

sed 's/foo/bar/g' file.txt

Change file name extensions #

sed 's/\.txt$/.md/' filenames.txt

Replace the second occurrence of “foo” with “bar” #

sed 's/foo/bar/2' file.txt

Replace “foo” with “bar” in lines 10 to 20 #

sed '10,20s/foo/bar/' file.txt  # Substitute on lines 10 to 20

Replace “foo” with “bar” only in lines containing “baz” #

sed '/baz/s/foo/bar/' file.txt

Trim leading and trailing whitespace #

sed 's/^[ \t]*//;s/[ \t]*$//' file.txt

Replace tabs with spaces #

sed 's/\t/    /g' file.txt  # 4 spaces in this case

Deletion (`d`) #

Delete lines matching a pattern or at specific addresses.

sed '/pattern/d' file.txt

Remove blank lines #

sed '/^$/d' file.txt

Delete second line #

sed '2d' file.txt

Delete lines 3 to 5 #

sed '3,5d' file.txt

Delete lines containing “error” #

sed '/error/d' file.txt

Insertion (`i`) and Appending (`a`) #

Insert or append text before or after a line matching a pattern or at a specific address.

sed 'ADDRESS i\ 
Text to insert' file.txt

and

sed 'ADDRESS a\ 
Text to append' file.txt

Insert “Start of File” before the first line #

sed '1i\
Start of File' file.txt

Append “End of File” after the last line #

sed '$a\
End of File' file.txt

Insert line numbers #

sed = file.txt | sed 'N;s/\n/\t/'

Change (`c`) #

Replace entire lines matching a pattern or address.

sed 'ADDRESS c\ 
New text' file.txt

Change lines containing “foo” to “bar”: #

sed '/foo/c\
bar' file.txt

Print (`p`) #

Print specific lines. Often used with the -n option to suppress automatic printing.

sed -n 'ADDRESS p' file.txt

Print only lines containing “success” #

sed -n '/success/p' file.txt

Addressing in sed #

Addresses in sed determine which lines the commands apply to. These can be specific line numbers, patterns, or a combination of both. Below are examples showcasing different types of addressing.

Single line addressing #

Apply a command to a specific line by specifying its line number.

Delete the 3rd line of a file #

sed '3d' file.txt

Replace the content of the 5th line #

sed '5c\

Range Addressing #

Apply commands to a range of lines using the START,END syntax.

Delete lines 2 to 4 #

sed '2,4d' file.txt

Replace the text “error” with “fixed” in lines 10 through 20 #

sed '10,20s/error/fixed/g' file.txt

Pattern Addressing #

Apply commands to lines matching a specific pattern.

Delete lines containing “DEBUG” #

sed '/DEBUG/d' file.txt

Add a prefix “LOG: “ to lines containing “ERROR” #

sed '/ERROR/s/^/LOG: /' file.txt

Range with Patterns #

Use patterns to define a range of lines.

Delete all lines between those matching “START” and “END” #

sed '/START/,/END/d' file.txt

Replace “foo” with “bar” in lines between “begin” and “stop” #

sed '/begin/,/stop/s/foo/bar/g' file.txt

Special Line Addressing #

Use special symbols like $ and ~ for specific addressing.

Delete the last line of a file #

sed '$d' file.txt

Replace “END” with “FINISH” in the last line #

sed '$s/END/FINISH/' file.txt

Delete every 4th line (GNU sed) #

sed '4~4d' file.txt

Combined Addressing #

Combine line numbers, ranges, and patterns for precise targeting.

Delete lines 1 to 5 and any line containing “TODO” #

sed '1,5d; /TODO/d' file.txt

Replace “abc” with “xyz” on the 2nd line and any line containing “keyword” #

sed '2s/abc/xyz/; /keyword/s/abc/xyz/' file.txt

Negating Patterns #

Use ! to negate an address, applying commands to lines not matching a condition.

Print all lines except those containing “SKIP” #

sed '/SKIP/!p' file.txt

Add a header to all lines except the first line #

sed '1!s/^/Header: /' file.txt

Table of addressing techniques #

Address Type	Syntax	Example Command
Single Line	`N`	`sed '3d' file.txt`
Line Range	`START,END`	`sed '5,10d' file.txt`
Pattern Match	`/PATTERN/`	`sed '/error/d' file.txt`
Range with Patterns	`/START/,/END/`	`sed '/begin/,/stop/d' file.txt`
Last Line	`$`	`sed '$s/END/FINISH/' file.txt`
Every nth Line	`N~M` (GNU `sed`)	`sed '2~3d' file.txt`
Negate Pattern	`!`	`sed '/pattern/!d' file.txt`

Multiple Commands #

There are three different ways to apply multiple sed commands in a single execution:

Using `-e` option #

Use the -e option to replace “foo” with “bar” and delete lines containing “baz”:

sed -e 's/foo/bar/' -e '/baz/d' file.txt

Using a `sed` script file with `-f` #

Create a script file script.sed to replace “foo” with “bar” and delete lines containing “baz”:

s/foo/bar/
bazd

Run sed with the script:

sed -f script.sed file.txt

Using Semicolons within a single script #

Using a semicolon to replace “foo” with “bar” and delete lines containing “baz”:

sed 's/foo/bar/; /baz/d' file.txt

Regular Expressions #

Leverage powerful pattern matching for flexible text processing.

Basic Regex #

.: Any single character.
*: Zero or more of the preceding element.
^: Start of line.
$: End of line.

Extended Regex #

Enabled with -E or -r flags.

+: One or more.
?: Zero or one.
|: Alternation.
() and {}: Grouping and quantifiers.