Help with special character in preg_match

[Naito]

I'm just starting to get back into seriously working with php and already I've run into a problem... I have a poorly structured flat file database that I need to convert into something useful, my thought was that I would use preg_match to pull the data between each instance of [' and [' into an array, then pull each instance between , into another array and then finally use only the lines from that array that contain the good data. Problem is, preg_match does not like the [, characters and simply backslashing them doesn't work..

Basically I'm writing

preg_match("['(.*)['", $pagedata, $stuffs);

I've tried changing the double quotes to single quotes and I've tried backslashing stuff but it still gives me errors every time... Is there another special character I need to be using and if so is there a list somewhere of these characters? The PHP Manual pages for preg_match and ereg don't really give me much help in this way.

I know I'm probably going to have a similar problem when I try to create an array out of every piece of data between ,'s too so if you could in your explination cover that aswell it would be greatly appreciated. 😕

[NogDog]

preg_() functions need a delimiter around the regex. Traditionally, the "/" character is used, but others can be, too:

preg_match("/\['(.*)\['/", $pagedata, $stuffs);
// ...or...
preg_match("#\['(.*)\['#", $pagedata, $stuffs);

[Naito]

Thank you muchly, now my issue is slightly different, it's now reading all the data I need, but putting it into one string, so when I try to;

echo "$stuffs[5]";

Or any such number it just errors out with an undefined index error, and when I print_r the variable it seems to come up as only one array instead of having a seperate section for each instance between [' and [', I've spent the last hour or so trying to find a solution to this problem and can only assume that eaither preg_match isin't the right function or I'm not specifiying some option that I don't know about. =/

[NogDog]

Are you expecting to get multiple matches? If so, you'll need to use preg_match_all().

The last parameter will store anything that matches the entire regex in $stuffs[0], and the first sub-pattern (the parenthesized part of the regex) in $stuffs[1]. So the first three matches of the sub-pattern will be $stuffs[1][0], $stuffs[1][1], and $stuffs[1][2]. You may also want to use the "ungreedy" modifier (the "U" after closing regex delimeter bleow):

preg_match_all("/\['(.*)\['/U", $pagedata, $stuffs);
if(count($stuffs)) // we found something
{
   echo "<h3>Matches:</h3>\n<ul>\n";
   foreach($stuffs[1] as $text)
   {
      echo "<li>$text</li>\n";
   }
   echo "</ul>\n";
}
else
{
   echo "<h3>No matches found</h3>\n";
}

[Naito]

Okies.. =/ The only way I can see to fully explain what I'm trying to do is to show you, so here it is, basically I'm trying to pull the javascript database info from a website so that I can get a list of spells and what level I get them at without spending 4-5 hours typing them out manually for each profession.. My full code is currently.

<?php

$urlpage = 'http://www.wowhead.com/?spells=7.11';
$pagedata = "";
$spells = "";
$stuff = "";
$recover1 = "['(.*)['";
$filetext = fopen("$urlpage", "r");
while (!feof($filetext))
{
$theline = fgets($filetext, 2048);
$pagedata .= "$theline\r";
}
fclose($filetext);

$pagedata = ereg_replace("<!DOCTYPE(.)<div id=\"lv\"><i>Loading...</i></div>", "", $pagedata);
$pagedata = ereg_replace("<div class=\"clear\"></div>(.)</html>", "", $pagedata);
preg_match_all("/['(.*)['/", $pagedata, $spells);
echo "$pagedata";

//print_r ($spells);

//echo "$spells[0][5]";

foreach($spells[1] as $text)
{
echo "<li>$text</li>\n";
}

?>

And my objective in the end is to seperate junk like

<script type="text/javascript">//<![CDATA[
var lv_data = [[17,'Spells',[['Healing Touch','Rank 1',1,'Nature','Restoration',5185,'Spell_Nature_HealingTouch',['<table><tr><td><table width="100%"><tr><td><b>Healing Touch</b></td><th><b class="q0">Rank 1</b></th></tr></table><table width="100%"><tr><td>25 Mana</td><th>40 yd range</th></tr></table>1.5 sec cast</td></tr></table><table><tr><td><span class="q">Heals a friendly target for 37 to 51.</span></td></tr></table>'],0,''],['Mark of the Wild','Rank 1',1,'Nature','Restoration',1126,'Spell_Nature_Regeneration',['<table><tr><td><table width="100%"><tr><td><b>Mark of the Wild</b></td><th><b class="q0">Rank 1</b></th></tr></table><table width="100%"><tr><td>20 Mana</td><th>30 yd range</th></tr></table>Instant cast</td></tr></table><table><tr><td><span class="q">Increases the friendly target\'s armor by 25 for 30 min.</span></td></tr></table>'],0,''],['Wrath','Rank 1',1,'Nature','Balance',5176,'Spell_Nature_AbolishMagic',['<table><tr><td><table width="100%"><tr><td><b>Wrath</b></td><th><b class="q0">Rank 1</b></th></tr></table><table width="100%"><tr><td>20 Mana</td><th>30 yd range</th></tr></table>1.5 sec cast</td></tr></table><table><tr><td><span class="q">Causes 12 to 14 Nature damage to the target.</span></td></tr></table>'],0,''],['Moonfire','Rank 1',4,'Arcane','Balance',8921,'Spell_Nature_StarFall',['<table><tr><td><table width="100%"><tr><td><b>Moonfire</b></td><th><b class="q0">Rank 1</b></th></tr></table><table width="100%"><tr><td>25 Mana</td><th>30 yd range</th></tr></table>Instant cast</td></tr></table><table><tr><td><span class="q">Burns the enemy for 7 to 9 Arcane damage and then an additional 12 Arcane damage over 9 sec.</span></td></tr></table>'],0,''],['Rejuvenation','Rank 1',4,'Nature','Restoration',774,'Spell_Nature_Rejuvenation',['<table><tr><td><table width="100%"><tr><td><b>Rejuvenation</b></td><th><b class="q0">Rank 1</b></th></tr></table><table width="100%"><tr><td>25 Mana</td><th>40 yd range</th></tr></table>Instant cast</td></tr></table><table><tr><td><span class="q">Heals the target for 32 over 12 sec.</span></td></tr></table>'],0,''],['Thorns','Rank 1',6,'Nature','Balance',467,'Spell_Nature_Thorns',['<table><tr><td><table

Into a nice array of $spellfinal[0][1] = "Healing Touch"
$spellfinal[0][2] = "Rank 1
$spellfinal[0][3] = "1"

Now, my thought is that I'll first pull the stuff in sections seperated by the [' and [' which if it pulls only the data between those two and then moves onto the next two after the end of the first I'll be able to then reference that data by grabbing the text before , and then further sorting it based on the slot I'll know it ends up in for the annoying 1 that is not within '' brackets. I've just been running into problems since I started but I'm persistant so I would really like to get it running. But so far the references like php manual that I've found have merely steered me into more errors.

[HalfaBee]

I can't do this for you, but you need to try and work out identical seperators for the data you need.

Things like <table></table> or <td><tr> </td/tr> and split the data up using thos strings.

A very odd looking DB 😉

[Naito]

Yep, thats exactally my plan and the only way I can see to do it, my problem isin't the logic, it's the execution with the php language... Preg_match_all does not individually split each match, and I still can't find a function that does. Now, if it gave me the option of storing only the first one found and then let me use ereg_replace to delete it and only it then I might be able to get somewhere by creating my own function in a loop. But it doesn't that I can see do that.