UNIX for Dummies Questions & Answers

I am learning how to write shell scripts and have come across an issue. I'm trying to write a script that looks for a directory called public_html, and if it finds one, to print the number of lines that contain applet tags (containing '<applet') in all files that end in either .html or .htm that are under the public_html directory . Here's what I have so far:

[ -d public_html ] || echo no website here
[ -d public_html ] && `grep -e .*\<applet.* -e \.html$ -e \.htm$ public_html`

I know that the first line works, but I have doubts about the second. I'm sure that the problem is in my use of grep, but beyond that I'm stumped.

Any ideas?

Filename globbing is done in the shell, not in grep, and is not a regex. so just use simple shell globbing i.e. public_html/*.htm public_html/*.html

I think I see what you mean. How's this:

[ -d public_html ] && grep public_html/*\<applet* public_html/*.html$ public_html/*.htm$

Also, do I need that backslash before the '<'? I'm not sure if that character requires it.

There's no reason to put 'public_html/' in your regex unless your HTML files actually contain that anywyere, which I sincerely doubt. Search for what you're trying to match inside the file, paths to the file you're searching for don't belong there.

You've put more regex junk back into my shell globs, like an $ to match the end of line. Shell globs continue to not be regular expressions, that syntax is wrong. It's unnecessary anyway since shell globs never return incomplete matches -- the entire filename must always be matched to return a result, '*.htm' will match 'this.htm' but not 'this.html'. I suggest actually using what I posted, no more, no less.

Shell globbing is very simple compared to regular expressions. It does only the following things:No modifiers, no beginning or end of line, no backreferences. That's all you get.

OK, I think that I'm sterting to get it. How does this look?

[ -d public_html ] && grep public_html/*<applet* public_html/*.htm public_html/*.html public_html

I like to do piped, multiprocessing scripts in ksh:

find ... -type d -name public_html 2>/dev/null | while read d
do
cd $d
echo ===== $d
grep ... *.html *.htm 2>&1
echo
done >list_file

Forgive the endless replies for a simple subject - the text tat I'm working from isn't very clear about... well, anything.

In response to Corona688 - I see what you mean, but I also need to look for '<applet' within the files that grep looks for. Would this do that?

[ -d public_html ] && grep public_html/.*\<applet.* public_html/*.htm public_html/*.html

The more I look at this I can't help but think that I need to enclose part of this in backquotes (`). Just how wrong am I?