How to search and replace using grep

In the command line or Unix-based terminal, grep is a powerful utility primarily used for searching lines in files that match specific patterns. However, grep does not natively support replacement functionality. If you need to perform search and replace operations, you typically use sed (stream editor) or similar tools. I'll first demonstrate how to use grep for searching text, and then show how to combine it with sed for replacement.

Using grep for Search

Assume you have a file example.txt with the following content:

shell
Hello, this is a test file.
This line contains the word test.
Another line.

If you want to search for all lines containing the word 'test', you can use the following grep command:

bash
grep "test" example.txt

This will display:

shell
Hello, this is a test file.
This line contains the word test.

Using sed for Search and Replace

Now, if you want to replace 'test' with 'exam' in the file, you can use the following sed command:

bash
sed 's/test/exam/g' example.txt

This command displays the replaced content in the terminal without modifying the original file. To save the changes, you can use the -i option (in-place editing):

bash
sed -i 's/test/exam/g' example.txt

After this, the content of example.txt will be changed to:

shell
Hello, this is a exam file.
This line contains the word exam.
Another line.

Combining grep and sed

Sometimes, you may want to first search for specific lines and then perform replacements on them. This can be achieved by piping the output of grep into sed. For example, if you only want to replace 'test' with 'exam' in lines containing 'word', you can do the following:

bash
grep "word" example.txt | sed 's/test/exam/g'

This command first finds all lines containing 'word', then replaces 'test' with 'exam' in those lines.

By doing this, you can flexibly combine Unix command-line tools for text processing, which is highly useful for daily programming, scripting, or data handling.