Re: Help with emacs regexp

[DaLoverhino]

On Aug 22, 2:32 pm, DaLoverhino <DaLoveRh...@hotmail.com> wrote:
> Hello.  I'm trying to get a little fancier with regexp, but I there's
> always this big gulf to the next plain which I can't seem to jump.  :)
>
> I've been trying to write a regexp where all lines that begin with a
> white space or opening and closing braces are removed.
>
> I have a C-module with function definitions, and I want to strip out
> everything in
> the file, but leave the function signature and return type.
>
> So the file looks like this:
>
> static returnType
> function1(blah, blah)
> {
>   if( blah)
>   {
>     blah;
>   }
>   blah;
>
> }
>
> returnType
> function2(blah, blah)
> {
>   blah;
>   blah;
>
> }
>
> I want it to look like the following after replace-regexp:
>
> static returnType
> function1(blah, blah)
> returnType
> function2(blah, blah)
>
> Here's what I have, it only works partly:
>
> "^\s-+.*
> " replace with <nothing>
>
> The above strips everything but the braces, but also removes return
> type of all functions but the first.
>
> "^\(\s-\|{\}\)+.*" replace with <nothing>
>
> The above strips everything but the function signature, but it removes
> the return type, and leaves a bunch of blank lines.
>
> This doesn't work either:
>
> "^[^a-zA-Z0-9]+?.*\n" replace with <nothing>
>
> Ofcourse, the easiest method would be to:
>
> shell-command-on-region
> egrep '^\w'
>
> Can anyone help me out?  Thanks.

Dang it!! I just figured it out:

"^\Sw+?.*
" replace with <nothing>

Anyways, I have to use C-q C-j.  Using \n for the new line just does
not seem to work.  How come?