Don't mention it :)
I thought it only meant "NOT" when it was inside the square brackets []...
Yes, you are right.. it was my fault.
It negates the character class if it is the first character
[^abc] //Matches any character except a,b,c
The circumflex ^ can also be outside of character class to indicate that the match must be at start of subject, also can be at start of alternative subpatterns, or after new line character "\n" in conjunction with the multiline modifier "m". In all these cases act as anchor to indicate matching in the start of subject or where new lines occur.
Example:
"/^[A-Z]+.+/" // Match only when the subject starts from one or more capital letters.
If you escape it, means just the ^ char. Example:
"/\d\^\d/" // Will match a string like "5^2"
I didn't use it because my aim was to match the headings and replace them..
This seems to be an easy way to exclude the headings..
(since you need everything except headings, you can temporarily replace them with an empty char "")
$Matches=array();
$Subject="Before Heading* <h4>THIS IS HEADING TEXT</h4> *After Heading";
$Pattern="/<h\d.+<\/h\d>/i"; // The pattern to match the heading tags ( note the caseless modifier "i" )
$Subject=preg_replace($Pattern,"", $Subject); // Replace the headings with empty string
echo $Subject; // prints: Before Heading* *After Heading (All the heading replaced with "")
Also the code with a couple of improvements is:
<?php
$keywords =array("sharing", "file", "program", "php"); // Keywords to search and replace
$links =array( "dictionary", "http://www.domain.com/data", "wiki", "wiki/web"); // Keywords links locations - URIs
$keywordList=array( "backup"); // Keywords to not match, example the: "backup file" will be not replaced
/* -------------------------------------------- SAMPLE HTML TEXT FOR TEST ------------------------------------------- */
// Our text as a long string (here a short HTML file for demonstration).
$Subject='
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html> <head>
<title>Preg Repace Test</title>
<META http-equiv="content-type" content="text/html; charset=iso-8859-1">
<META name="Keywords" content=" sharing, file, program, php, backup ">
<script language="JavaScript" type="text/javascript" src="Home/ScriptFile.js"></script></head>
<style type="text/css">
<!--
h3 { color:#991111; }
h4 { color:#991111; }
-->
</style>
<!-- NOTE: This is just a comment, to test if replaced (or no), keywords that are located before body tag (normally no), keywords: sharing file php program.
The previous keywords must stay intact (not replaced by anchor <a href=.. > tags) since they are before the body tag end excluded from replacement -->
<body topmargin="30" leftmargin="50" bgColor="#cecece" style="font:14px/18px verdana">
<center>
<h3> HEADING TEXT 1: Test sharing file program php </h3>
Regular Expression Test in PHP (preg_replace example)<br>
(automatically creates links to files for matched keywords)
<H3>*HEADING TEXT 2: sdfjsdflsdjf backup file sharing sdfsdf sldk*</H3>
<p>
php best file sharing <b> backup program </b> dslfks dfsj sdf lksjd building a web page sdlfjsdfkjsdflsdf creating a web page sdfsdldsfk jsdfk sdlksj df file share
program sdflksjdf lsdkfjsdflksdfj lksdfj file sharing application sdf sdfsf sdf df <b> backup file </b> sharing ds sdfs dfsdf PHP sdfs df <b> backup sharing </b> fsdsd f
</p>
<br>
<h4>HEADING TEXT 3: This test tags and keywords sharing file program php</h4>
</center></body>
</html>
'; // The end single quot of our $Subject var.
$OriginalSubject=$Subject;
$count=0;
echo "<center>\n<h2>".'REGULAR EXPRESSIONS: TEXT REPLACEMENT IN PHP'."</h2>\n";
echo '( Example by K.Zotos   -10/2007- )'."\n</center>\n<br><hr><br>\n";
for ($count2=0; $count2<count($keywords); $count2++) {
$Subject=$OriginalSubject; // Initialize our text after each replacement.
/* --------------------------------------- REMOVE EVERYTHING BEFORE THE BODY TAG -------------------------------------- */
// To avoid replace keywords in header.. we temporarily remove everything from document start until the body start tag ( <body )
// in other word removes anything that precedes the "<body" tag.
$Matches = array(0 => "");
//$Pattern = "/.+\/head\s*>/ims"; // This matches anything from document start until (including) the end of header
$Pattern = "/.+(?=<body)/ims"; // Matches anything until the "<body" start tag
$Result = preg_match($Pattern, $Subject, $Matches); // If text found from document start until the "<body" tag, $Matches[0] will contains the whole matched string
if ($count2==0) echo "<br>\n\n<!-- ------ TEXT EXCLUDED FROM REPLACEMENT (START) ------- -->\n".$Matches[0]."<!-- ------ TEXT EXCLUDED FROM REPLACEMENT
(END) ------- -->\n\n"; // To see the removed text (everything before the body start tag)
if ($Result >0) $Subject=preg_replace($Pattern, "", $Subject); // Temporarily replace the contents before the body tag with "" empty string
// we have stored them in $Matches array..
/* ------------------------------------------- REPLACE ALL HEADING (<H1 - H6>) TAGS ----------------------------------------- */
// Replace all heading tags with a custom text - place holder. After keyword matching, will restore the heading texts in their initial positions
$Pattern="/<h\d.+<\/h\d>/im"; // The pattern to match the heading tags
$HeadingSets=array();
preg_match_All($Pattern, $Subject, $HeadingSets, PREG_SET_ORDER);
$Subject=preg_replace($Pattern,"*PLACE_HOLDER*", $Subject); // Replace the headings
// echo 'Headings Found: '.count($HeadingSets)."<br>\n";
$HeadingTags=array();
$HeadingElement="";
for($a=0;$a<count($HeadingSets);$a++){ // Each element of $HeadingSets array, is also an array with its first element to be the matchin heading
$HeadingElement=$HeadingSets[$a][0];
array_push($HeadingTags,$HeadingElement);
//echo $HeadingElement."<br>\n";
}
/* ---------------------------------------------- REPLACE KEYWORDS WITH LINKS ----------------------------------------------- */
$Extension='.php'; // File extension
$File='/'.$keywords[$count2].$Extension; // File name with a forward slash in fron of it (will used to compose keywords URLs)
$Replacement='<a href="'.$links[$count2].$File.'">'.$keywords[$count2].'</a>';
// $Pattern="/(^(<\s*head.+\/head\s*>))|(?<!$keywordList[$count])$keywords[$count2]/ims";
// The following pattern matches a keyword that not preceded by words in $keywordList + one space.
// Eg. if search for the keyword "program" and the "not match" word is "backup" then the "backup program" will not replaced..
$Pattern = "/(?<!$keywordList[$count]\s)$keywords[$count2]/ims"; // i: case insensitive search, m:multiline, s:dot-all (. dot char can be anything, even new line)
$Subject = @$Matches[0].preg_replace($Pattern, $Replacement, $Subject,1); // Concatenates the previously removed init text (header) with the replaced one.
/* ------------------------------------------- RESTORE ALL HEADING (<H1 - H6>) TAGS ----------------------------------------- */
// Replace the placeholder texts with the original heading tags as stored in the $HeadingTags array. For this we are going to supply the peg_replace
// with 2 arrays; the saved headings and another one with an equivalent number of pattern elements (the same string used previously as place holder)
$PlaceHolders=array();
$Pat="/\*PLACE_HOLDER\*/";
for($a=0;$a<count($HeadingTags);$a++) {
array_push($PlaceHolders, $Pat); // The $PlaceHolders array requires to have equal number of elements (patterns) to achieve a one by one replacement
}
// echo 'Patterns Count: '.count($PlaceHolders)."<br>\n";
$Subject=preg_replace($PlaceHolders, $HeadingTags, $Subject,1);
/* ------------------------------------------------------- PRINT OUT RESULTS ------------------------------------------------------ */
// Prepare and Output the results in HTML format.
$Text="\n<div style='color:#cc2211'><b><span style='color:#115566'>".' -------------- TEXT REPLACEMENT   Array Element: '.$count2.' => '.$keywords[$count2]."
------------------- </span></b> ";
$Text.= "\n\n<br><br>\n".'Headings Found: '.count($HeadingSets)."<br>";
$Text.= "\n".'Keyword Replaced: <b>'.$keywords[$count2].'</b><br>';
$Text.= "\n".'Not Replaced: <b>'.$keywordList[$count].' '.$keywords[$count2].'</b>';
$Text.= "<br>\n".'File Link: <b>'.$links[$count2].$File.'</b></div>';
$Text.="<br>\n\n".$Subject. "\n\n\n<br><hr><br><br>";
echo $Text;
}
?>
Eventually, if you want to replace a certain keyword with link, only if that keyword-link not exist; I think is better to compare this keyword with the existing links in your array and proceed with the replacement only if not exist rather than trying to test this condition with a regex which has to "play" a double role, to match from one point and to avoid matching at the same time.. since this make things more complex.. This is my opinion of cource :)
Good luck with the project!