RegExp patterns - help

[mastabog]

hello guys

it may be a bit silly but my question is kinda simple:
how do you write a regexp pattern (for use in preg_* functions) to match a string only if it doesnt have a specified string ahead of it?

For example, if you try to highlight some keywords of a language, but you don't want to highlight them if they are located after some inline comment like:

1. highlight "if", "else", "end":
   if (a < b)
   ... some code here ...
   else
   ... some other code here ...
   end

2. do NOT highlight after "%"
   if (a > b) % now we are checking --if-- its something --else--
   ... some code here ...
   end

remember, i want this down only by using regexp patterns, for use in preg_replace()
- if not % ahead of keywords, then highlight, if there is % ahead of keywords, then highlight the whole string after % in some other way, but don't match the keywords anymore

this was an example of matlab code ... any can apply.
any help is appreciated

big 10q

[Weedpacket]

"Only if it doesn't have a specified string ahead of it" sounds like a forward negative assertion to me....

Matching "foo" only if it's not followed by "bar"... where "foo" and "bar" are regular expressions that match what you want to highlight and what the specified string is respectively ...

foo(?!(.bar)*)

(plus bits on either side for matching the rest of the string depending on context).
E.g., match "if", "else" and "elseif" only if they're not followed by a %...

(if|elseif|else)(?!(.%)*)

That's straight off the top of my head.

But the problem is a lot more complicated than that. What about

if(a==b) % we need to see if b<50% or not

? Both "if"s will be highlighted, because both are followed by %.

You'll find yourself looking towards a full-blown parser; and regular expressions are not as powerful as you'll be wanting.

One idea I showed on this forum in the past might be of use to you; in effect it can be used as a poor man's parse tree, but ultimately you might find yourself wanting to do a proper job of it. Good luck!

[mastabog]

thanks for your time, but i solved it 10 minutes right after i posted the help message (2 nights with no sleep CAN get you to no-code-intuition 🙂)

here's the code i used and it works v fine:

function HiliteML ($mfile)
{
if (!($code = @file ($mfile)))
return "File not found: <i>$mfile</i>";

$code = implode ("", $code);

$keywords = "break|case|for|continue|else|elseif|end|for|while";

$needle = array
(
	"/(%.*?)(?=\r|\n|$)/m",
	"/^([^%]*?)(?<=\s|^)(".$keywords.")(?=\W|$)/im",
	"/^([^%]*?)(?<=\s|^|\w|\W)(\'.*?(?<!\\\)\')/im",
);
$replace = array
(
	'<span style="color:green"><i>\1</i></span>',
	'\1<span style="color:blue">\2</span>',
	'\1<span style="color:darkred">\2</span>',
);
return = preg_replace ($needle, $replace, $code);

}

It works. Maybe this helps someone else.
Cheers