That's the simple modification I told about:

$ perl -n -e '$count++ while /\bhello\b/g; END { print $count }' text.
+out
[download]

And to answer kyle at Re^2: grep command, my interpretation was that he was concerned about how many times the word happened with no concern for which particular lines. But then he tells about grepping — so if the issue is finding the lines and then telling how many times it appeared in total, this one may work:

$ perl -n -e 'print, $count++ while /\bhello\b/g; END { print $count }
+' text.out
[download]

Almost!

However, the last line in which "hello" is found is printed twice.

CLI: perl -n -e "print, $count++ while /\bhello\b/g; END { print $count }" hello.out
        (quotes revised for windows)

file hello.txt:
hello. hellow, world. bye
world says hello.
bye
hello, hellow, hello
not now.
done.


output
hello. hellow, world. bye
world says hello.
hello, hellow, hello
hello, hellow, hello
4

I don't see it as a *simple* spoiler, as people quite often try it in scalar context like

$cnt += m{pattern}g;

which doesn't work as they expect.


Enjoy, Have FUN! H.Merijn

I don't see it as a *simple* spoiler, as people quite often try it in scalar context like

$cnt += m{pattern}g;
[download]

But then the usual workaround is

$cnt += () = m{pattern}g;
[download]

which is somewhat simpler and more natural than your reference-and-dereference trick.