Damian Conway wrote:
> > now, what if the double quoted range had a \E in it? either directly or
> > via interpolation? maybe the end escape should be another char than \E?
>
> Make \E significant only where it's explicit.
None of the \ escapes are significant via interpolation, even today. \E
shouldn't be any different. So I agree with this.
> Also make \E signifcant
> only when \I is in effect.
I'm not so sure about this one: \E already has semantics for when \Q, \L, \U
are in effect, too. So at minimum we'd have to be careful to document and
understand what is meant by:
$bar = 'abc';
$x = '$$$\Ifoo\U$bar\E\E$$$';
print $x;
Does this produce:
$$$fooABC\E$$$
or
$$$fooABC$$$
So I tried to see how " handles multiple \E things, and discovered that while \E
ends any of \Q \U \L, the semantics are individualized: \U will end a \L and I
assume vice versa, but \Q doesn't affect \U or \L (whichever might be in
effect). And \E will end the nearest prior \Q without disabling \U or \L (if
either is also in effect), needing another \E to turn off a \U or \L that might
be in effect. But a \E that has nothing in effect will still be elided from the
" string.
So we need to make it explicit how \I which would, I assume, also turn on \L,
\U, and \Q for use, would interact with them.... probably that they would all
have to be turned off before a \E would turn off \I.
Or, if my other email using \D in double quotes is accepted, similar issues
arise.
--
Glenn
=====
There are two kinds of people, those
who finish what they start, and so
on... -- Robert Byrne
_____NetZero Free Internet Access and Email______
http://www.netzero.net/download/index.html
