[ruby-talk:16260] Re: line numbers in multiline regular expressions.

On Tue, Jun 05, 2001 at 09:36:00PM +0100, [e-mail] wrote:
> Wayne Blair wrote:
> > 
> > ----- Original Message -----
> > From: "jonas" <[e-mail]>
> > To: "ruby-talk ML" <[e-mail]>; <[e-mail]>
> > Sent: Tuesday, June 05, 2001 3:01 PM
> > Subject: [ruby-talk:16245] line numbers in multiline regular expressions.
> > 
> > > Hi,
> > >
> > > Here is my 5 minute attempt to create a program that extracts all
> > > comments in a C source file containing the string "TODO":
> > >
> > > todo_comments = File.
> > >   open(ARGV[0], "r").
> > >   read.
> > >   scan(/\/\*.*?\*\//m).
> > >   delete_if { |c| c !~ /TODO/ }
> > >
> > > It works fine, but I really would like to have access to the line number
> > > where the scan matches. Is that easy to accomplish or do I have to
> > > rethink the structure of the program from scratch?
> > >
> > > /j
> > >
> > The scan can match across multiple lines, though, right?  Can you give an
> > example of input and desired output?
> 
> 
> Input:
> 
> ------------------------8<-------------
> /* comment 1 */
> 
> /*
>  * comment 2
>  */
> 
> /* 
>  * TODO  foo bar
>  * flopp flerp plopp plerp blipp
>  * blopp boo.com
>  */
> ------------------------8<-------------
> 
> 
> 
> Output: 
> 
> I would like the line number of the line where "TODO" match and it's the
> comment's complete content, i.e the complete last comment above.

You could try something like this --- maybe a bit non-Rubyesque, but it'll
work

line_nr = 0
begin_ln = 0
while file.gets
	line_nr+=1
	if comment
		comment <<= $_
		if %r{\*/}
			print begin_ln, line_nr, comment if comment =~ /TODO/
			comment = nil
		end
	elsif %r{/\*}
		comment = $_.dup
		begin_ln = line_nr
	end
end

Renald