extract data from web page

[thebadbad]

Yeah, my definition was a bit loose.

 

characters that are within the pattern need to be escaped (for the most part this is true.. but there are oddball exceptions.. but I digress)

 

I can't imagine a case where you don't have to escape delimiters in the pattern. Or were you talking about characters in general?

[nrg_alpha]

Nope, I am talking about delimiter characters in a pattern

 

Consider the following for example:

 

$str = 'Some text... <_FOO_> some more text <_BAR_> and yet again some more!';
$str = preg_replace('<<_[^_]+_>>', '<_XX_>', $str); // note there is no need to escape < and > inside the delimiters in this scenario...
echo $str; // Output: Some text... <_XX_> some more text <_XX_> and yet again some more!

 

This is admittedly a stupid (and rare) example. But it shows that under some circumstances, it might not require escaping. In this case, there is a caveat... there has to be matching opening and closing < and > characters inside the < > delimiters.. if there is a mismatch, then you'll probably need to escape (otherwise you'll get either 'Unknown modifier' or 'No ending matching delimiter' warnings).

 

I personally advise against using opening / closing matching punctuation as delimiters (such as <...>, {...}, (...) or [....]). Too much confusion / trouble for what it's worth IMO.

[thebadbad]

Ah, thanks for enlightening me! Didn't know you could have nonidentical delimiters like < and >. But I agree, seems more confusing than necessary.

 

I should probably just read a book, instead of keep bugging you

[nrg_alpha]

I should probably just read a book, instead of keep bugging you

 

lol no worries.

[nadeemshafi9]

thanks guys