From c0f64d360fcb410192fa35d6353a2956bcc421c6 Mon Sep 17 00:00:00 2001
From: David M <davidmj@users.sourceforge.net>
Date: Sat, 23 Dec 2006 17:56:50 +0000
Subject: [PATCH] regex man strikes again!

git-svn-id: file:///svn/phpbb/trunk@6794 89ea8834-ac86-4346-8a33-228a782c2dd0
---
 phpBB/develop/regex.php      | 9 ++++-----
 phpBB/includes/functions.php | 8 ++++----
 2 files changed, 8 insertions(+), 9 deletions(-)

diff --git a/phpBB/develop/regex.php b/phpBB/develop/regex.php
index 4893a3d271..2b736afec2 100644
--- a/phpBB/develop/regex.php
+++ b/phpBB/develop/regex.php
@@ -43,12 +43,11 @@ echo 'IPv4: ' . $ipv4 . "<br />\nIPv6: " . $ipv6 . "<br />\n";
 $pct_encoded = "%[\dA-F]{2}";
 $unreserved = 'a-z0-9\-._~';
 $sub_delims = '!$&\'()*+,;=';
-$pchar = "(?:[$unreserved$sub_delims:@|]|$pct_encoded)"; // rfc: no "|"
+$pchar = "(?:[$unreserved$sub_delims:@|]+|$pct_encoded)"; // rfc: no "|"
 
 $scheme = '[a-z][a-z\d+\-.]*';
-$reg_name = "(?:[$unreserved$sub_delims|]|$pct_encoded)+"; // rfc: * instead of + and no "|"
-$authority = "(?:(?:[\w\-.~!$&'()*+,;=:]|$pct_encoded)*@){0,1}(?:$reg_name|$ipv4|\[$ipv6\])[:]?\d*";
-$userinfo = "(?:(?:[$unreserved$sub_delims:]|$pct_encoded))*";
+$reg_name = "(?:[$unreserved$sub_delims|]+|$pct_encoded)+"; // rfc: * instead of + and no "|"
+$userinfo = "(?:(?:[$unreserved$sub_delims:]+|$pct_encoded))*";
 $ipv4_simple = '[0-9.]+';
 $ipv6_simple = '\[[a-z0-9.:]+\]';
 $host = "(?:$reg_name|$ipv4_simple|$ipv6_simple)";
@@ -57,7 +56,7 @@ $authority = "(?:$userinfo@)?$host(?::$port)?";
 $segment = "$pchar*";
 $path_abempty = "(?:/$segment)*";
 $hier_part = "/{2}$authority$path_abempty";
-$query = "(?:[$unreserved$sub_delims:@/?|]|$pct_encoded)*"; // pchar | "/" | "?", rfc: no "|"
+$query = "(?:[$unreserved$sub_delims:@/?|]+|$pct_encoded)*"; // pchar | "/" | "?", rfc: no "|"
 $fragment = $query;
 
 $url =  "$scheme:$hier_part(?:\?$query)?(?:\#$fragment)?";
diff --git a/phpBB/includes/functions.php b/phpBB/includes/functions.php
index ea05d969a2..323590b269 100644
--- a/phpBB/includes/functions.php
+++ b/phpBB/includes/functions.php
@@ -2740,7 +2740,7 @@ function get_preg_expression($mode)
 		case 'bbcode_htm':
 			return array(
 				'#<!\-\- e \-\-><a href="mailto:(.*?)">.*?</a><!\-\- e \-\->#',
-				'#<!\-\- (l|m|w) \-\-><a href="(.*?)">.*?</a><!\-\- \1 \-\->#',
+				'#<!\-\- ([lmw]) \-\-><a href="(.*?)">.*?</a><!\-\- \1 \-\->#',
 				'#<!\-\- s(.*?) \-\-><img src="\{SMILIES_PATH\}\/.*? \/><!\-\- s\1 \-\->#',
 				'#<!\-\- .*? \-\->#s',
 				'#<.*?>#s',
@@ -2749,15 +2749,15 @@ function get_preg_expression($mode)
 
 		case 'url':
 			// generated with regex generation file in the develop folder
-			return "[a-z][a-z\d+\-.]*:/{2}(?:(?:(?:[a-z0-9\-._~!$&'()*+,;=:]|%[\dA-F]{2}))*@)?(?:(?:[a-z0-9\-._~!$&'()*+,;=|]|%[\dA-F]{2})+|[0-9.]+|\[[a-z0-9.:]+\])(?::\d*)?(?:/(?:[a-z0-9\-._~!$&'()*+,;=:@|]|%[\dA-F]{2})*)*(?:\?(?:[a-z0-9\-._~!$&'()*+,;=:@/?|]|%[\dA-F]{2})*)?(?:\#(?:[a-z0-9\-._~!$&'()*+,;=:@/?|]|%[\dA-F]{2})*)?";
+			return "[a-z][a-z\d+\-.]*:/{2}(?:(?:(?:[a-z0-9\-._~!$&'()*+,;=:]+|%[\dA-F]{2}))*@)?(?:(?:[a-z0-9\-._~!$&'()*+,;=|]+|%[\dA-F]{2})+|[0-9.]+|\[[a-z0-9.:]+\])(?::\d*)?(?:/(?:[a-z0-9\-._~!$&'()*+,;=:@|]+|%[\dA-F]{2})*)*(?:\?(?:[a-z0-9\-._~!$&'()*+,;=:@/?|]+|%[\dA-F]{2})*)?(?:\#(?:[a-z0-9\-._~!$&'()*+,;=:@/?|]+|%[\dA-F]{2})*)?";
 		break;
 
 		case 'www_url':
-			return "www\.(?:[a-z0-9\-._~!$&'()*+,;=|]|%[\dA-F]{2})+(?::\d*)?(?:/(?:[a-z0-9\-._~!$&'()*+,;=:@|]|%[\dA-F]{2})*)*(?:\?(?:[a-z0-9\-._~!$&'()*+,;=:@/?|]|%[\dA-F]{2})*)?(?:\#(?:[a-z0-9\-._~!$&'()*+,;=:@/?|]|%[\dA-F]{2})*)?";
+			return "www\.(?:[a-z0-9\-._~!$&'()*+,;=|]+|%[\dA-F]{2})+(?::\d*)?(?:/(?:[a-z0-9\-._~!$&'()*+,;=:@|]+|%[\dA-F]{2})*)*(?:\?(?:[a-z0-9\-._~!$&'()*+,;=:@/?|]+|%[\dA-F]{2})*)?(?:\#(?:[a-z0-9\-._~!$&'()*+,;=:@/?|]+|%[\dA-F]{2})*)?";
 		break;
 
 		case 'relative_url':
-			return "(?:[a-z0-9\-._~!$&'()*+,;=:@|]|%[\dA-F]{2})*(?:/(?:[a-z0-9\-._~!$&'()*+,;=:@|]|%[\dA-F]{2})*)*(?:\?(?:[a-z0-9\-._~!$&'()*+,;=:@/?|]|%[\dA-F]{2})*)?(?:\#(?:[a-z0-9\-._~!$&'()*+,;=:@/?|]|%[\dA-F]{2})*)?";
+			return "(?:[a-z0-9\-._~!$&'()*+,;=:@|]+|%[\dA-F]{2})*(?:/(?:[a-z0-9\-._~!$&'()*+,;=:@|]+|%[\dA-F]{2})*)*(?:\?(?:[a-z0-9\-._~!$&'()*+,;=:@/?|]+|%[\dA-F]{2})*)?(?:\#(?:[a-z0-9\-._~!$&'()*+,;=:@/?|]+|%[\dA-F]{2})*)?";
 		break;
 	}