preg match all pattern help.

how to figure it out..

[achshar]

hello,
i am making a replacment system.. i dont really know what it is called...
its like when i enter into the database [bold]foo[/bold]
the scripts changes [bold] to <strong> i can change single words like strong, em, u, p, div etc but i want it to manipulate variables like

[url=LINK title=TITLE]link to foo[/url]

OR something like

[div class=CLASSNAME]div contenet[/ div]

i know preg_replace_all is what i need.. i can also do php. but i am having a hard tome figuring what the pattern is? like any tutorial or anything would be appreciated..
like this is i found..

/\[url=(.*?)\](.*?)\[\/url\]/i

what i coud get is..
(.*?) is any variable
and \ is used to escape plain words, just like usual php
but what is the i at end for?
or those extra /s..?

i have done php all by my self, so even if it had been something basic, i wouldn't be knowing it :(

Thanks
achshar

[Alcoholico]

Something like this should work, at least help you understand how it works:

function parseBBCode($txt) { 
 $patterns = array(
'#\[b\](.+)\[\/b\]#iUs',
'#\[url\=(.+)\](.+)\[\/url\]#iUs',
'#\[img\](.+)\[\/img\]#iUs',
'#\[quote\=(.+)\](.+)\[\/quote]#iUs'
);
 $replacements = array(
'<b>$1</b>', 
'<a href="$1">$2</a>',
'<img src="$1" />',
'<div class="quote_owner">$1 Said:</div><blockquote>$2</blockquote>'
);
 return preg_replace($patterns, $replacements, $txt); 
}

However, you're better off taking a look at PCRE documentation on php.net. You could also search the net for "php bbcode parser", that's what you're buiding.

[achshar]

hello Alcoholico.. thanks alot for reply... yes i'v now been able to figur out most of it by my self

but yeah php bbcode parser.. that is! thats what i needed.. i simply didnt knew what to search for.. :) thanks.. and that code
will try that too... thanks alot again.. :) :D

[achshar]

hey hello..
i now have one question.. in the code you gave me above the pattern for [b ][/b ] what is 'iUs' in the end for? it is there for every enty.. it works even if i remove it..

also this is what i've come up with.. this works.. just want to confirm if this is a legitimate way of doing it :)

$string = html_entity_decode($string, ENT_QUOTES);
$markup = array(
'#\[b\](.+)\[\/b\]#iUs' => '<b>$1</b>'
);

$from = array();
$to = array();

foreach ($markup as $a => $b){
$from[] = $a;
$to[] = $b;
}
echo preg_replace($from, $to, $string);

[Matthew1980]

Hi there achshar,

Those chars (i, U and s)are pattern delimiters (I hope ;)) they set conditions of use for the preceding pattern. In this case i = case insensitve, and U and s I am not sure about, however, if you check this link out it can explain it better than I ever could:-

[php.net ]

There are probably better explanations out there, but I'm sure a quick google will do the trick. But from what I can see 's' enables the engine to read everthing including the newline chars, without it, \n & \r would be ignored... Have a read, and hopefully all will become clear.

Hope that helps you,

Cheers,
MRb

[rocknbil]

# = delimiters, which define "the pattern." Those thar' be modifiers, as described. :-)

it works even if i remove it..

try it with [B ] or [b
]

or multiple instances within your test block of text, and it won't. :-)

As mentioned, i is a case insensitive modifer, which is why [B ] would fail without the modifiers.

U is a modifier to stop the "greediness" of a match. One of the problems with regexps is sometimes you have a pattern, like

<strong>bold</strong> more text more text <strong>more bold</strong>

So if you, say, want to match on the strong tags and take out everything in between, some regular expressions might go all the way to the SECOND ending </strong> and nuke the entire line when you just want to take out the two instances. This is to stop this "greediness." This is why they say, in the documentation,

(e.g. .*?)

The question mark, in other languages, is what is called a quantifier to limit greedy pattern matching. The dot . means any character, * means zero or more, so without the quantifer,

preg_replace('/<[^>]+>.*<\/[^>]+>/','',$string)

might indeed nuke the entire string, because there ARE zero or more characters between, and those include the first closing </strong> and the second opening <strong>. The quantifer or the U modifier will limit this and will only nuke the tags and what's between them.

s - as mentioned the dot character means "any character." When dots are used in the regexp and you include the s modifier, it will include newlines in the match and it otherwise wouldn't, which is why

[b
]

would fail without it.

[achshar]

ohkzz i believe i got most of it now:)
and yeah i have completed the bbcode parser.. and it works perfectly. Apparently i had been able to figure out that multiple instances or opposite case code does not work without those pattern delimiters :)

thanks alot :D

[Matthew1980]

rocknbil: Gah, I always get those backwards (delimiters/modifiers), thanks for the pointer ;)

Achshar: Glad you are all sorted out now.

Cheers,
MRb