awk, mawk, nawk, gawk... WHAT?

Question

I've just started learning awk and I'm a little confused about all those versions around. Is there any "version" which is found on all Unix-like systems? Like, you know, plain vi? Does the standard awk support the -F option?

Answer 1

awk - the most common and will be found on most Unix-like systems, oldest version and inferior to newer ones.

mawk - fast AWK implementation which it's code base is based on a byte-code interpreter.

nawk - while the AWK language was being developed the authors released a new version (hence the n - new awk) to avoid confusion. Think of it like the Python 3.0 of AWK.

gawk - abbreviated from GNU awk. The only version in which the developers attempted to add i18n support. Allowed users to write their own C shared libraries to extend it with their own "plug-ins". This version is the standard implementation for Linux, original AWK was written for Unix v7.

There are other versions like jawk (java implementation), bwk (Brian W. Kernighan's implementation) and so on.

Answer 2

You can just use awk. It is defined by POSIX and therefore has to exist on all POSIX-conformant systems.

The -F parameter is mandated by that as well.

Answer 3

awk will be on just about every *NIX based system, but the exact specifics of what it supports will shift slightly as it's entirely possible it will simply be symlinked to a different version akin to how /bin/sh is often linked to a specific shell, often bash or one of its derivatives. (For the record, I also know alot of machines where vi is symlinked to vim as well.)

Answer 4

The Wikipedia AWK page is a good starter reference for beginning to understand AWK.
The Field Separator option '-F' is supported in all variants of AWK -- afaik.