On Mon, 2008-12-01 at 02:57 +0100, Jenda Krynicky wrote:
> OTOH, he can match the innermost [block]...[/block], remove it or
> replace by something that doesn't contain either "tag", match the
> next one ... It would not be pretty, but it would be doable.
>
> And actually if there really is just one tag and if we can assume
> that the text doesn't contain \x01 nor \x00 we might use a trick:
>
> $text = <<'*END*';
> [block]
> this is a block
> [block]
> this is another block
> [/block]
> first block ends
> [/block]
> *END*
>
> print "TEXT: $text\n\n";
> $text =~ s/\[block\]/\x01/g;
> print "TEXT: $text\n\n";
> $text =~ s/\[\/block\]/\x00/g;
>
> print "TEXT: $text\n\n";
> 1 while ($text =~ s{\x01([^\x00\x01]*)\x00}{
> print "found '$1'\n\n"; 'HERE WAS BLOCK'
> }ge);
> print "TEXT: $text\n\n";
>
> Now whether this is of anyuse to ZuLuuuu noone can tell.
>
If there were only two levels of nesting, then the data structure would
have bounded, nested contexts. These can be parsed with a FSA. A stack
is not needed.
--
Just my 0.00000002 million dollars worth,
Shawn
The key to success is being too stupid to realize you can fail.
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
http://learn.perl.org/
