Odd PHP error

[ljm]

I've coded a BBcode parser for a site I am involved with. There is nothing wrong with that, as far as I know, but it keeps spitting out an odd error:

Warning: preg_replace() [function.preg-replace]: Compilation failed: missing ) at offset 23 in **directory** on line 55

Everything seems to load fine - I think, and I've checked the code and it also seems fine. Any ideas as to what may be causing this odd error?

EDIT: Sorry about the title! I was too busy copying and pasting the post and forgot about it!

[Shadowfiend]

Don't suppose you could change the thread topic? ;-)

[filburt1]

Make sure you're using [phpfunction]preg_quote[/phpfunction].

[ljm]

I tried using preg_quote, and it completely garbled the output. Here's the script, anyway:

<?php

// Press Start Online SimpleCode translator


    // For backwards compatibility, remove trailing slashes from past articles
    $original_string = str_replace("/]","]",$original_string);
    $original_string = str_replace("’","'",$original_string);
    
    // Search for the following SimpleCodes
    $bbcodes = array(
                    "'\[b\](.*?)\[/b\]'i",
                    "'\[u\](.*?)\[/u\]'i",
                    "'\[i\](.*?)\[/i\]'i",
                    "'\[img\](.*?)\[/img\]'is",
                      "'\[url\](.*?)\[/url\]'is",
                      "'\[url=(.*?)\](.*?)\[/url\]'is",
                    "'\[il=(.*?)\](.*?)\[/il\]'i",
                    "'(\[1img\](.*?)\[/1img\]'is",
                    "'\[2img\](.*?)\[/2img\]'is",
                    "'\[3img\](.*?)\[/3img\]'is",
                    "'\[1img\](.*?)\[endimg\]'is",
                    "'\[2img\](.*?)\[endimg\]'is",
                    "'\[3img\](.*?)\[endimg\]'is",
                    "'\[quote\](.*?)\[/quote\]'is",
                    "'\[h2\](.*?)\[/h2\]'i",
                    "'\[h1\](.*?)\[/h1\]'i",
                    "'\[fl=(.*?)\](.*?)\[/fl\]'i",
                    "'\[hr\]'",
                    "'\[p\]'"
                    );

    // Replace the SimpleCodes with XHTML
    $replace = array( 
                    "<strong>\\1</strong>",
                     "<u>\\1</u>",
                     "<em>\\1</em>",
                     "<img src=\"\\1\" alt=\"image\" />",
                     "<a href=\"\\1\" title=\"\\1\">\\1</a>",
                     "<a href=\"\\1\" title=\"\\1\">\\2</a>",
                     "<a href=\"image.php?id=\\1\" title='click for large version'>\\2</a>",
                     "<div id='oneimage'>\\1</div><div id=\"endfloat\"></div>",
                     "<div id='twoimages'>\\1</div><div id=\"endfloat\"></div>",
                     "<div id='threeimages'>\\1</div><div id=\"endfloat\"></div>",
                     "<div id='oneimage'>\\1</div><div id=\"endfloat\"></div>",
                     "<div id='twoimages'>\\1</div><div id=\"endfloat\"></div>",
                     "<div id='threeimages'>\\1</div><div id=\"endfloat\"></div>",
                     "<div id='quote_box'>\\1</div>",
                     "<h2>\\1</h2>",
                     "<h1>\\1</h1>",
                     "<div id='img_\\1'>\\2</div>",
                     "<div id=\"hrfix\"><div class=\"sub\"> </div><hr size=\"1\" height=\"1\" width=\"100%\"><div class=\"sub\"> </div></div>",
                     "<br /><br />"                    
                    );
    
    $new_string = preg_replace($bbcodes,$replace,$original_string);


?>

The way this works is that I include it into another file, after the variable $original_string has been declared, and before I echo $new_string. Just so I don't have to edit any other files with it all in. Repeated bits in the array are for backwards compatibility with the existing database entries, which were worked around errors in the previous bbcode script.

[MikeWire]

missing ) at offset 23 in **directory** on line 55

This tells me that you have a ")" missing in your 23rd offset in your array. Line 55 is at the end of your array - so a string is being pushed into your array with a "(" and then PHP pukes when it it doesn't see a closing ")" - not sure it thats helps, or if I am just stating the obvious...

[Shadowfiend]

"'([1img](.*?)[/1img]'is",

Look for the missing closing paren in the pattern ;-)

[Steax]

I'm wondering why you're splitting the 1img stuff.... can't you just merge them all into one regexp?

EDIT: Btw, you shouldn't use IDs for things that might be multiple. Unless of course this has to do with your CMS.

[ljm]

Oh so that was it. Hehe, would never have spotted that.

The 1img/2img/3img thing is related to how the site works. I've left it that way so the staff don't have to adjust to something new, and so all the past articles still format correctly.

The ID thing was the past administrator's error. I'll be sorting that out once I've got to grips with how he's coded the site. I'll be giving it a standards workout, anyway.

Thanks for the help guys.