> Well ... it's also possible to have a consistent, totally O-O parser
> that handles this case without any problem.  It just requires another
> string representation.  For example, in Python, there's an alternative
> string representation that looks like this:
> 
>   r'a string with a single \ in it'
> 
> The fact that it begins with "r'" tells Python's parser that
> everything until the final single quote character is to be taken
> literally, and therefore, backslashes don't need to be escaped.  The
> only exception is when you want a single-quote in the string, in
> which case you need to double it, as so:
> 
>   r'a string with a single '' in it'
> 
> The main place I see this usage in Python is in strings that are used
> as arguments to the regular-expression parser, which is exactly what
> we're talking about here.  And since the one character that we'd have
> to double (single-quote) is not a regular-expression meta-character,
> we don't get any quadrupling of characters.
> 
> I'm not trying to make any Ruby-Python comparisons here.  I only am
> bringing this up to give a widely used, real-world example of a
> rigidly O-O method of parsing strings that doesn't complicate the
> parsing of regular expressions.

Actually, I went looking at the %Q(...) syntax before I sent that e-mail,
since I was hoping Matz would have thought of that. Seems he didn't. I'm
half-tempted to submit an RCR for this if nobody else has. It is a very useful
feature, and as you pointed out, not entirely difficult (at least, I wouldn't
imagine it is) to implement.