Unexected result with preg_replace()

[leeming]

Im trying to create a copy of BBS codes.

 

i know i can do replace on these type of tags, but i have come across an error, and would like to understand it.

 

  //BOLD text
  $tag = '/(\[b\])(.*)(\[\/b\])/';
  $html = '<b>${2}</b>';
  $text = preg_replace($tag, $html, $text);

 

The problem i have with the above is...

 

[b]bold[/b] normal [b]bold[/b]

 

Shows as:

bold[ /b ] norma [ b ]bold

Just without the obvious spaces

 

 

Basicly, it is catching the 1st bold tag, and matching it with the last one it can find (thus the middle is missed out)

[MadTechie]

here

 

<?php
$string = 'Hello [b] how are you [/b] sorry';
$pattern[] = '/\[b\](.*)\[\/b\]/';
$replacement[] = '<B>${1}</B>';
echo preg_replace($pattern, $replacement, $string);
?> 

 

EDIT:

this will catch [b] how are you [/b] and replace with <B>(.*)</B> in this case <B>how are you</B>

no need to capture the [b]'s by them selfs as your replacing them..

i added an extended one below just for fun

 

 

i have been playing with RegEx and a may write my own BBCode when time permits sounds like good pratice

[MadTechie]

extended

 

<?php
$string = 'Hello [b] how are you [i] how are you [u] how are you [/u][/i][/b] sorry';
$pattern[] = '/\[b\](.*)\[\/b\]/';
$pattern[] = '/\[i\](.*)\[\/i\]/';
$pattern[] = '/\[u\](.*)\[\/u\]/';
$replacement[] = '<B>${1}</B>';
$replacement[] = '<I>${1}</I>';
$replacement[] = '<U>${1}</U>';
echo preg_replace($pattern, $replacement, $string);
?> 

[Glyde]

The way you want this to work requires (as far as I know) you to individually locate each one.  Lucky enough for you, I have already written a script that can do this, and it works pretty well as far as I can tell.

 

<?php
class bbcode {

/**
 * List of supported tags
 *
 */
var $_tagList = array(
	'b'		=> '<b>{TEXT}</b>',
	'u'		=> '<u>{TEXT}</u>',
	'i'		=> '<i>{TEXT}</i>',
	'color'	=> '<span style="color: {OPT}">{TEXT}</span>'
);

/**
 * List of tags with an optional attribute (so you can make [color=whatever]sdf[/color] for example
 *
 */
var $_attrList = array(
	'color'	=> "((#[0-9A-F]{3,6})|([A-Z\-]+))"
);

/**
 * Convert from a standard input to HTML
 *
 * @param string $input
 * @return string
 */
function toHTML($input) {

	// Loop through the taglist and convert all
	foreach ( $this->_tagList as $tagKey=>$tagHTML ) {
		$tagOpen = "[{$tagKey}]";
		$tagClose = "[/{$tagKey}]";

		// Find the first instance of the current tag item
		if (!$this->_attrList[$tagKey]) {
			while ((stripos($input, $tagOpen) !== FALSE) && (strripos($input, $tagClose) !== FALSE)) {
				$position = stripos($input, $tagOpen);
				if ($position === FALSE) continue;
				$endposition = strripos($input, $tagClose, $position);
				if ($endposition === FALSE) continue;

				// Replace the text
				$newText = substr($input, $position + strlen($tagOpen), $endposition - ($position + strlen($tagOpen)));
				$newText = str_replace("{TEXT}", $newText, $tagHTML);
				$input = substr($input, 0, $position) . $newText . substr($input, $endposition + strlen($tagClose));
			}
		} else {
			$tagOpen = "@\[{$tagKey}=".$this->_attrList[$tagKey]."\]@is";
			while (preg_match($tagOpen, $input) && (strripos($input, $tagClose) !== FALSE)) {
				preg_match($tagOpen, $input, $match);
				$position = stripos($input, $match[0]);
				if ($position === FALSE) continue;
				$endposition = strripos($input, $tagClose, $position);
				if ($endposition === FALSE) continue;

				// Replace the text
				$newText = substr($input, $position + strlen($match[0]), $endposition - ($position + strlen($match[0])));
				$tagHTML = str_replace("{OPT}", $match[1], $tagHTML);
				$newText = str_replace("{TEXT}", $newText, $tagHTML);
				$input = substr($input, 0, $position) . $newText . substr($input, $endposition + strlen($tagClose));
			}
		}
	}
}
}
?>

I haven't really optimized this code yet, but it works.  Anyways, I think you can figure out how it works, if not, let me know.

[leeming]

I thank you for the help, but... does not fix any thing...

 

my code still takes the 1st and last (ignoring middle tags) and applying formatting to the whole lot. But i may add, this seems to be only in sections (maybe lines?)

 

I am not fimilar with the syntax, which is why i was enclosing the [ b ] tags (i got hold of the bbs for [ url ] and changed it to do the rest of the tags for myself)

[leeming]

The way you want this to work requires (as far as I know) you to individually locate each one.  Lucky enough for you, I have already written a script that can do this, and it works pretty well as far as I can tell.

 

<?php
class bbcode {

/**
 * List of supported tags
 *
 */
var $_tagList = array(
	'b'		=> '<b>{TEXT}</b>',
	'u'		=> '<u>{TEXT}</u>',
	'i'		=> '<i>{TEXT}</i>',
	'color'	=> '<span style="color: {OPT}">{TEXT}</span>'
);

/**
 * List of tags with an optional attribute (so you can make [color=whatever]sdf[/color] for example
 *
 */
var $_attrList = array(
	'color'	=> "((#[0-9A-F]{3,6})|([A-Z\-]+))"
);

/**
 * Convert from a standard input to HTML
 *
 * @param string $input
 * @return string
 */
function toHTML($input) {

	// Loop through the taglist and convert all
	foreach ( $this->_tagList as $tagKey=>$tagHTML ) {
		$tagOpen = "[{$tagKey}]";
		$tagClose = "[/{$tagKey}]";

		// Find the first instance of the current tag item
		if (!$this->_attrList[$tagKey]) {
			while ((stripos($input, $tagOpen) !== FALSE) && (strripos($input, $tagClose) !== FALSE)) {
				$position = stripos($input, $tagOpen);
				if ($position === FALSE) continue;
				$endposition = strripos($input, $tagClose, $position);
				if ($endposition === FALSE) continue;

				// Replace the text
				$newText = substr($input, $position + strlen($tagOpen), $endposition - ($position + strlen($tagOpen)));
				$newText = str_replace("{TEXT}", $newText, $tagHTML);
				$input = substr($input, 0, $position) . $newText . substr($input, $endposition + strlen($tagClose));
			}
		} else {
			$tagOpen = "@\[{$tagKey}=".$this->_attrList[$tagKey]."\]@is";
			while (preg_match($tagOpen, $input) && (strripos($input, $tagClose) !== FALSE)) {
				preg_match($tagOpen, $input, $match);
				$position = stripos($input, $match[0]);
				if ($position === FALSE) continue;
				$endposition = strripos($input, $tagClose, $position);
				if ($endposition === FALSE) continue;

				// Replace the text
				$newText = substr($input, $position + strlen($match[0]), $endposition - ($position + strlen($match[0])));
				$tagHTML = str_replace("{OPT}", $match[1], $tagHTML);
				$newText = str_replace("{TEXT}", $newText, $tagHTML);
				$input = substr($input, 0, $position) . $newText . substr($input, $endposition + strlen($tagClose));
			}
		}
	}
}
}
?>

I haven't really optimized this code yet, but it works.  Anyways, I think you can figure out how it works, if not, let me know.

 

The class looks impressive but, i would like to add that this code i have custom tags, due it been made for some one (its part of my coursework, even though i use php quite a bit any way). And also i'd have to document all this external code. (like previously said, im wanting to understand why it is doing this... i can still use str_replace)

[Glyde]

Well you can add as many as you want to the array assuming the tags are formed such as: [b ].  While that code isn't exactly documented, you can feel free to use it if you want, or, you can just check out how I used stripos and strripos, but this might not give you the desired output.

 

This would take:

Some [b]sample[/b] [b]string[/b]

And make it:

Some <b>sample[/b] [b] string</b>

 

As stated, I haven't really fixed this function perfectly...But the reason it works as shown above is because that's how I intended it to work in order to deal with tags correctly.

[leeming]

...

This would take:

Some [b]sample[/b] [b]string[/b]

And make it:

Some <b>sample[/b] [b] string</b>

...

 

exactly what im trying to explain...

 

oh well, i will just use a simple string replace, and still wonder why i get that error

[Glyde]

str_replace works fine, just remember you won't have the functionality of tags such as [ /color]

[leeming]

Nope i am now forced to do this function for one of my custom tags

 

  
//References
  $tag = '/\[ref\](.*)\[\/ref\]/';
  $html = '<sup><a href="#ref${1}">[${1}]</a></sup>';
  $text = preg_replace($tag, $html, $text);

 

It is an page to add articles, and [ref] just anchors to the bottom where the references are all listed (by number)

 

i get the same annoying output as before...

 

text [ref]1[/ref] middle [ref]2[/ref] end

 

which outputs

 

text <sup><a href="#ref1[/ref] middle [ref]2">[1[/ref] middle [ref]2]</a></sup> end

 

is there no way to get preg_replace (or similar function) to go thru left to right instead of skipping to the last tag?

[MadTechie]

if its just numbers then you could use

 

<?php
$string = 'text [ref]1[/ref] middle [ref]2[/ref] end';
$pattern[] = '/\[ref\]([\d]*)\[/ref\]/';
$replacement[] = '<sup><a href="#ref${1}">[${1}]</a></sup>';
echo preg_replace($pattern, $replacement, $string);
?>

[MadTechie]

oh yeah if the ref's are single words (no spaces) this should work

 

 

 

<?php
$string = 'text [ref]1[/ref] middle [ref]2[/ref] end';
$pattern[] = '/\[ref\](\w*)\[\/ref\]/';
$replacement[] = '<sup><a href="#ref${1}">[${1}]</a></sup>';
echo preg_replace($pattern, $replacement, $string);
?>

[MadTechie]

if its just numbers then you could use

 

<?php
$string = 'text [ref]1[/ref] middle [ref]2[/ref] end';
$pattern[] = '/\[ref\](\d*)\[/ref\]/';
$replacement[] = '<sup><a href="#ref${1}">[${1}]</a></sup>';
echo preg_replace($pattern, $replacement, $string);
?>

[Guest prozente]

use the ungreedy modifier

 

<?php
$text = 'text [ref]1[/ref] middle [ref]2[/ref] end';
$tag = '/\[ref\](.*)\[\/ref\]/U';
$html = '<sup><a href="#ref${1}">[${1}]</a></sup>';

$text = preg_replace($tag, $html, $text);

?>