Use sed to replace line of text with new line including variable

Question

I was looking at this solution: Using sed to edit lines in a file with a variable and tried to apply it to my situation, but I have only recently discovered sed and haven't got the hang of it yet.

If I have a file located /tmp/mary.txt and it looks like this:

Mary had a little lamb
with fleece as white as
snow. Everywhere that
Mary went, the lamb was
sure to go.

I want to replace a line of text in this file, specifically:

Mary had a little lamb -> Mary had 2 little lambs

I used this:

sed -i 's/*.had a little lamb.*/Mary had $(nproc) little lambs/' /etc/mary.txt

I am using -i for inplace and s/ for substitute and it throws error:

sed: 1: "mary.txt": extra characters at the end of g command

---

The other thing I am trying to do, is replace the word "lamb" on the first line with the result of echo $PATH

sed -i 's/*.had a little lamb.*/ Mary had a little '$(echo $PATH)'/' /etc/mary.txt

This throws the same error as above. I know these are two different scenarios, but can it be shown to me what I did wrong, so I can use this in future.

Answer 1

Use double quotes to have the var evaluated properly:

$ var=43
$ sed "s/.*had a little lamb.*/Mary had $var little lambs/" file
Mary had 43 little lambs
with fleece as white as
snow. Everywhere that
Mary went, the lamb was
sure to go.

See Jacob Minshall's answer regarding the .* instead of *.. I had written the sed command from scratch, so I did not notice the error in the original code.

Answer 2

I think it is because you have *. at the beginning, but you want to have .*

The . means any character, while the * means zero or more of the preceding character, which there isn't one of