sed Command Reference

The sed (Stream Editor) command is a powerful tool for performing text transformations and manipulations on streams of text, such as files or input from pipelines. This reference provides commonly used sed commands and examples to help you efficiently edit and process text files.

Basic Syntax

The general syntax for using sed is:

sed [options] 'command(s)' file(s)

Or for inline commands:

sed [options] -e 'command1' -e 'command2' file(s)

Commonly Used Options

Option	Description
`-e 'command'`	Add the script to the commands to be executed.
`-f scriptfile`	Execute the script in the specified file.
`-i[SUFFIX]`	Edit files in-place (optionally creating a backup with the given suffix).
`-n`	Suppress automatic printing of pattern space.
`-r` or `-E`	Use extended regular expressions in the script.
`-s`	Treat files as separate rather than as a single continuous long stream.

Basic Commands

Command	Description
`s/old/new/flags`	Substitute `new` for the first occurrence of `old` in a line.
`p`	Print the current pattern space.
`d`	Delete the pattern space; start next cycle.
`q`	Quit `sed`; no further processing.
`a\ text`	Append `text` after the current line.
`i\ text`	Insert `text` before the current line.
`c\ text`	Replace the current line with `text`.
`n`	Read next line into pattern space.
`h`	Copy pattern space to hold buffer.
`g`	Get hold buffer content and replace pattern space.
`G`	Append hold buffer content to pattern space.
`x`	Exchange the pattern space with the hold buffer.
`=`	Print the current line number.

Flags for Substitution Command (`s///`)

Flag	Description
`g`	Global replacement in a line (replace all occurrences).
`p`	Print the line if a replacement was made.
`w file`	Write the line to `file` if a replacement was made.
`I`	Case-insensitive search.
`n`	Replace the `n`-th occurrence in a line.

Addressing

Addresses specify which lines of input to apply commands to.

number - Match only the specified line number.
$ - Match the last line.
/pattern/ - Match lines that contain the regular expression pattern.
first~step - Starting at line first, match every step lines.
addr1,addr2 - Match range from addr1 to addr2, inclusive.

Common Examples

Command	Description
`sed 's/old/new/' file`	Replace first occurrence of `old` with `new` in each line.
`sed 's/old/new/g' file`	Replace all occurrences of `old` with `new` in each line.
`sed -i 's/old/new/g' file`	Edit the file in-place, replacing all occurrences.
`sed '/pattern/d' file`	Delete lines matching `pattern`.
`sed -n '/pattern/p' file`	Print only lines matching `pattern`.
`sed -n '5,10p' file`	Print lines from 5 to 10.
`sed '1d' file`	Delete the first line.
`sed '$d' file`	Delete the last line.
`sed '1i\New Header' file`	Insert "New Header" at the beginning of the file.
`sed 's/^/prefix_/' file`	Add "prefix_" at the beginning of each line.
`sed 's/$/_suffix/' file`	Add "_suffix" at the end of each line.
`sed 's/^$.\{5\}$/\1.../' file`	After first 5 characters of each line, insert "...".
`sed -n '=' file`	Print line numbers.
`sed '/^$/d' file`	Delete empty lines.
`sed 's/$.*$/\U\1/' file`	Convert text to uppercase.
`sed 's/$.*$/\L\1/' file`	Convert text to lowercase.
`sed 's/^\s\+//' file`	Remove leading whitespace.
`sed 's/\s\+$//' file`	Remove trailing whitespace.

Multi-Line Operations

By default, sed operates on a per-line basis. To perform multi-line operations, use the following commands:

N - Read the next line into the pattern space (separated by a newline).
D - Delete up to the first newline in pattern space and start next cycle.
H - Append the pattern space to the hold buffer with a newline.
G - Append the hold buffer to the pattern space with a newline.

Example: Join Every Two Lines

sed 'N;s/\n/ /' file

This reads the next line into the pattern space and replaces the newline character with a space, effectively joining every two lines.

Regular Expressions in sed

sed supports regular expressions for pattern matching:

^ - Start of line
$ - End of line
. - Any single character
* - Zero or more of the preceding character
\+ - One or more of the preceding character (requires -r or -E option)
\? - Zero or one of the preceding character (requires -r or -E option)
\| - Alternation (logical OR, requires -r or -E)
 - Grouping
\1, \2, ... - Back-references to captured groups
[abc] - Character class matching a, b, or c
[^abc] - Negated character class

Hold Space Operations

sed provides a hold space (buffer) for advanced text processing:

h - Copy pattern space to hold space (overwrites hold space).
H - Append pattern space to hold space (with a newline).
g - Replace pattern space with hold space.
G - Append hold space to pattern space.
x - Exchange pattern space with hold space.

Example: Swap Two Lines

sed 'N;swap_lines' file

# Define the swap_lines script:
swap_lines='
    h       # Copy pattern space (line1\nline2) to hold space
    s/.*\n//    # Remove line1, keep line2 in pattern space
    x       # Exchange pattern space and hold space
    s/\n.*//    # Remove line2, keep line1 in pattern space
    G       # Append hold space (which now contains line2) to pattern space
'

Tips and Best Practices

Use single quotes around the sed script to prevent shell interpretation.
For complex scripts, use the -f option with a script file.
Be cautious with in-place editing (-i), especially without backups.
Test your sed commands on sample data before applying to important files.
Use extended regular expressions with the -r or -E option for more concise patterns.

Return to Home