Cheatsheet - sed

sed (Stream Editor) processes text line by line from an input stream or a file. By default, it prints the processed output to standard output without changing the original file.

Basic Syntax

sed [OPTIONS] 'script' [input-file]
sed [OPTIONS] -e 'script1' -e 'script2' [input-file]
sed [OPTIONS] -f script-file [input-file]

Command-Line Options

Option	Description
-n, --quiet	Suppress the default behavior of printing every line. Use with the p command to print only specific lines.
-i [suffix]	In-place edit. Modifies the file directly. If a suffix is provided (e.g., -i.bak), a backup of the original file is created.
-e script	Add the script (command) to be executed. Useful for running multiple commands.
-f file	Read sed commands from a specified file instead of the command line.
-E, -r	Use Extended Regular Expressions (ERE) instead of Basic Regular Expressions (BRE). Simplifies some patterns.

Addressing: Specifying Which Lines to Act On

If no address is specified, the command applies to every line.

Address	Description	Example
number	Act on a specific line number.	sed '3d' file (deletes line 3)
$	Act on the last line of the file.	sed '$d' file (deletes the last line)
/pattern/	Act on any line matching the regular expression pattern.	sed '/error/d' file (deletes lines with "error")
start,end	Act on a range of lines from start to end (inclusive).	sed '5,10d' file (deletes lines 5 through 10)
num,+N	Act on line num and the next N lines.	sed '5,+3d' file (deletes lines 5, 6, 7, 8)
/start/,/end/	Act on a range from the first line matching /start/ to the first line matching /end/.	sed '/BEGIN/,/END/d' file
addr!	Invert the address. Act on all lines that do not match the address.	sed '/^#/!d' file (deletes all lines not starting with #)

Core Commands

s - Substitute

This is the most common command in sed. Syntax: 's/pattern/replacement/flags'

Flag	Description
g	Global. Replace all occurrences of the pattern on the line, not just the first one.
i	Case-insensitive matching (GNU sed extension).
N (number)	Replace only the Nth occurrence of the pattern on the line.
p	Print the line if a substitution was made. Often used with -n.
w file	Write the line to file if a substitution was made.

Substitution Examples:

# Replace the first 'cat' with 'dog' on each line
sed 's/cat/dog/' file

# Replace ALL 'cat's with 'dog' on each line
sed 's/cat/dog/g' file

# Replace 'cat' with 'dog' only on line 10
sed '10s/cat/dog/' file

# Replace 'cat' with 'dog' only on lines containing 'animal'
sed '/animal/s/cat/dog/g' file

# Use a different delimiter (#) when the pattern contains slashes
sed 's#/var/log#/usr/share/logs#g' file

# & refers to the entire matched pattern
# Add parentheses around numbers
sed 's/[0-9][0-9]*/(&)/g' file

# Use backreferences (\1, \2) with capturing groups \(...\)
# Swap two words
sed 's/\(first\)\s\(second\)/\2 \1/' file

d - Delete

Deletes the entire line matching the address.

# Delete line 5
sed '5d' file

# Delete all lines containing 'debug'
sed '/debug/d' file

# Delete all blank lines
sed '/^$/d' file

p - Print

Prints the line. Almost always used with the -n option to avoid duplicate output.

# Print only lines 1 through 5 (like head)
sed -n '1,5p' file

# Print only lines containing 'CRITICAL' (like grep)
sed -n '/CRITICAL/p' file

i, a, c - Insert, Append, Change

These commands insert, append, or change entire lines.

Syntax: address [command] \ text to insert/append/change

# Insert text before line 4
sed '4i \
This line will be inserted before line 4.' file

# Append text after the last line
sed '$a \
-- End of File --' file

# Change (replace) the content of line 3
sed '3c \
This is the new content for line 3.' file

Practical Recipes

# In-place edit a file, creating a backup with .bak extension
sed -i.bak 's/old_value/new_value/g' config.conf

# Delete trailing whitespace from all lines
sed -i 's/\s*$//' file

# Comment out lines containing a specific pattern
sed -i '/sensitive_info/s/^/#/' file

# Chain multiple commands to first delete debug lines, then fix a typo
sed -i -e '/DEBUG/d' -e 's/eror/error/g' file

# Add a prefix to every line
sed 's/^/-> /' file

# Print a specific section of a file by patterns
sed -n '/start-section/,/end-section/p' file

# Delete all comments and blank lines
sed -e '/^#/d' -e '/^$/d' file

Add One Line To The Beginning

$ sed -i '1s/^/line to insert\n/' /path/to/file

Remove 1st line in place

$ sed -i 1d filename

Add comma to the end of the line

$ cat foo.txt | sed s/$/,/g

Remove https://

$ ... | sed s#https://##`