From: Dave Thomas <Dave@...>
Date: 2001-01-31T06:09:43+09:00
Subject: [ruby-talk:10140] Re: Easy regex question

Neil Conway <nconway@klamath.dyndns.org> writes:

> In Ruby (Jan 18 snapshot), I get this:
> 
> $ irb
> irb(main):001:0> str = "it's me"
> "it's me"
> irb(main):002:0> str.gsub(/'/, "\\'")
> "its mes me"
> 
> Can someone tell me:
> 
> 1) What Ruby is doing? I can't figure out what it's doing to go from
> "it's me" -> "its mes me"

This is ugly, so brace yourself...

"\\'" is a normal string, so before being passed to gsub it undergoes
the normal string replacements (things like \n becomes newline, and \\ 
becomes a single backslash). So what gsub gets passed as a second
parameter is the two character string

     \'

Now... Regexp substitution parameters have a number of special purpose 
escape sequences. \1 matches the first parenthesized subexpression, \& 
is the whole match, and (wait for it) \' is everything _after_ the
match. So, in your case, you've got

   it<'>s me

so everything after the match is 's me', so the string becomes 'its mes 
me'.

> 2) The correct regex to do what I'm looking for

Well, there are several options

  str = "it's me"
  str.gsub(/'/, "\\\\'")

  str.gsub(/'/) { "\\'" }


Regards


Dave