Understanding sed command syntax and sed commands

Question

Could someone explain what this sed command does here?

sed 's!^M$!!;s!\-!!g;s!\.!!g;s!\(..\)!\1:!g;s!:$!!'

It seems replacing/deleting some characters... But I couldn't figure it out... It's really complicated (I mean all of those s ; / g M ^ . and other characters)

thanx regards

Answer 1

You can split it up into a series of substitutions:

s!^M$!!
s!\-!!g
s!\.!!g
s!\(..\)!\1:!g
s!:$!!

Each one is using ! as the delimiter, so the patterns are s!match!replacement! . The g on the end means that some of them are global, so will happen as many times as possible rather than only once on each line.

^ matches the start of the line and $ matches the end, so the first one removes any M s that are found on a line by themselves.

The next two remove all . and - that are found. The . needs a slash before it so that it only matches a literal . rather than matching any character . The - doesn't need a slash before it but it doesn't do any harm either.

The fourth one adds a : after every 2 characters, using a capture group and back reference.

Hopefully you can work out what the last one does, based on my explanation of the first one!