Skip to Content

sed Command in Linux

sed is a powerful text editing tool. Short for Stream Editor, sed allows us to find, replace, delete, insert, and transform text in a file or stream - all without opening a text editor. It is particularly useful for automating repetitive text-based tasks, manipulating files without opening them in a text editor, and performing search-and-replace operations. 


What is sed Command ?

sed is a non-interactive stream editor. It reads text line-by-line from a file or input stream, applies a set of instructions, and writes the modified output to standard output (by default).

Basic Syntax of sed Command:

sed [options] 'command' file
  • options : Modify sed's behavior, such as -i for in-place editing or -n to suppress automatic output.
  • command : Contains editing instructions, like the substitute command s or the delete command d.
  • file : The file(s) sed will process; can be piped from another command.


Example of sed Command in Linux

1. Simple Find and Replace

sed 's/error/success/' file.txt

🔹 Replaces the first occurrence of "error" with "success" in each line.

2. Global Replacement

sed 's/error/success/g' file.txt

🔹 Replaces all occurrences of "error" in each line.


To replace a specific occurrence, like the second instance of a pattern, specify it after the final delimiter. For example : sed 's/unix/linux/2' configfile.txt 

Also, you can combine n with g to replace from the nth occurrence to all occurrences in a line.

3. In-Place File Editing

sed -i 's/http/https/g' index.html

🔹 Modifies the file in place (no need to redirect output).

📝 Tip: Add a backup extension if you want to preserve the original :

sed -i.bak 's/http/https/g' index.html

4. Delete Specific Lines

sed '2d' file.txt

🔹 Deletes the 2nd line.

sed '3,5d' file.txt

🔹 Deletes lines 3 to 5.

sed '/^$/d' filename.txt

🔹 Deletes empty lines

5. Print Specific Lines

sed -n '2p' file.txt

🔹 Prints only the 2nd line.

sed -n '3,6p' file.txt

🔹 Prints lines 3 to 6 only.

6. Insert a Line Before or After a Pattern

  • Insert before a match:
    sed '/pattern/i New line before pattern' file.txt
  • Insert after a match:
    sed '/pattern/a New line after pattern' file.txt

7. Replace Only on a Specific Line

sed '5s/foo/bar/' file.txt

🔹 Only replaces "foo" with "bar" on line 5.


Regular expressions with sed

sed supports regular expressions (regex) for powerful pattern matching and manipulation. 

  • ^ : Matches the beginning of a line.
  • $ : Matches the end of a line.
  • . : Matches any single character.
  • * : Matches zero or more occurrences of the preceding character.
  • + : Matches one or more occurrences of the preceding character (requires -E or -r).
  • [] : Matches any character within the brackets.
  • () : Used for grouping subexpressions and backreferences.


Using sed command in Pipelines

sed works seamlessly in Unix pipelines :

cat access.log | sed -n '/404/p'

🔹 Filters only lines with 404 status codes from a web log.

Or :

ps aux | sed -n '/apache2/p'

🔹 Show only processes related to apache2.

Tips and Tricks

  • Combine with find for recursive edits :
    find . -name "*.html" -exec sed -i 's/http/https/g' {} +


sed command Use cases

  • Log analysis : Filtering log files for errors or specific events.
  • Configuration management : Editing configuration files, such as updating paths or changing settings.
  • Batch processing : Applying the same operation to multiple files.
  • Data extraction and transformation : Extracting relevant data from structured files or transforming data formats. 

The sed command is a great tool for working with text in Linux environments.


Back

Check out these tutorials :