The Unbeatable Effectiveness of Sed and Grep

Using the shell effectively in a Unix-like environment must show a certain flow, which is comparable to poetry. According to former AT&T Bell Labs employee Brian Kernighan in his book “Unix: A History and a Memoir”, many of the hackers who developed and utilized Research Unix at Bell Labs in the early days were extremely fond of literature. Of course, the arts and sciences are typically viewed as polar opposites, but it is no where more prevalent than in computing that those who create must learn from the work of the past, and must adapt that past work to fit their present task. This is key because at the time, people were in the early stages (not that the world is anywhere close to finished today) of finding out how to effectively develop software, and how to use computers to help with everyday tasks and to solve problems that would be much more difficult to solve had computers not existed.

With these ideas in mind, it is important to think about the idea that is engrained in the minds of programmers since that era; an idea that has stuck with programmers, even if they fail to live up to it: the Unix philosophy of “do one thing, and do it well”. Grep and Sed, incredibly simple yet incredibly useful in an immense, seemingly infinite range of scenarios, are prime justifications for the Unix philosophy that have stood the test of time.

Variants of many Unix programs are still in wide use today, but why do we still use Grep and Sed so often if there are so many programs out there that we don’t use anymore that can do so much more? It’s a testament to the rock solid stability and the ability to put trust in a program that you know will do what you ask it to do promptly and with minimal effort. That is one reason. Another is the universal nature of these programs. Only when these are combined together, a program will continue to be useful for many years.

Grep is admittedly much simpler than Sed, and in fact, the basic functionality of Grep is easily replicable in Sed. However, Grep, which is named after the Ed command g/re/p (which does exactly what Grep does as well), is only Grep. It doesn’t try to do anything else but read a file and print the lines that match the given pattern to standard output. Sed, which stands for “Stream EDitor”, is the same way. All it does is edit the inputted stream of text. That’s all it is meant to do, and that is all it will do.

These useful programs can be applied in many different scripts as well as be used in interactive shells due to their simple nature. It is easy to find a case where you may want to “grep” a certain piece of text from a source file (maybe to find a function definition or to check if a header file is included). You could also want to print all the lines of a file until a certain pattern is found, which can be easily accomplished with sed /re/q. Learning Sed and Grep is a very common and beneficial investment for command line users and hackers in general. It really is no wonder they survived this long.