use EmailAddressValidator class for email validation FS#1503

darcs-hash:20081011173245-7ad00-578e03ca701fc4443ba91329be6ac1ca9344c7a7.gz

3 files changed, 151 insertions, 16 deletions

diff --git a/_test/cases/inc/mail_isvalid.test.php b/_test/cases/inc/mail_isvalid.test.php
index 120764e08..374204679 100644
--- a/_test/cases/inc/mail_isvalid.test.php
+++ b/_test/cases/inc/mail_isvalid.test.php
@@ -15,7 +15,7 @@ class mail_isvalid extends UnitTestCase {
         $tests[] = array('username+tag@domain.com',true); // FS#1447
         $tests[] = array("rfc2822+allthesechars_#*!'`/-={}are.legal@somewhere.com.au",true);
         $tests[] = array('_foo@test.com',true); // FS#1049
-        $tests[] = array('bugs@php.net1',false);
+        $tests[] = array('bugs@php.net1',true); // new ICAN rulez seem to allow this
         $tests[] = array('.bugs@php.net1',false);
         $tests[] = array('bu..gs@php.net',false);
         $tests[] = array('bugs@php..net',false);
diff --git a/inc/EmailAddressValidator.php b/inc/EmailAddressValidator.php
new file mode 100644
index 000000000..2ce2093e2
--- /dev/null
+++ b/inc/EmailAddressValidator.php
@@ -0,0 +1,146 @@
+<?php
+/**
+ * EmailAddressValidator Class
+ *
+ * @author  Dave Child <dave@addedbytes.com>
+ * @link    http://code.google.com/p/php-email-address-validation/
+ * @license http://www.opensource.org/licenses/bsd-license.php
+ */
+class EmailAddressValidator {
+
+    /**
+     * Check email address validity
+     * @param   strEmailAddress     Email address to be checked
+     * @return  True if email is valid, false if not
+     */
+    function check_email_address($strEmailAddress) {
+
+        // Control characters are not allowed
+        if (preg_match('/[\x00-\x1F\x7F-\xFF]/', $strEmailAddress)) {
+            return false;
+        }
+
+        // Split it into sections using last instance of "@"
+        $intAtSymbol = strrpos($strEmailAddress, '@');
+        if ($intAtSymbol === false) {
+            // No "@" symbol in email.
+            return false;
+        }
+        $arrEmailAddress[0] = substr($strEmailAddress, 0, $intAtSymbol);
+        $arrEmailAddress[1] = substr($strEmailAddress, $intAtSymbol + 1);
+
+        // Count the "@" symbols. Only one is allowed, except where
+        // contained in quote marks in the local part. Quickest way to
+        // check this is to remove anything in quotes.
+        $arrTempAddress[0] = preg_replace('/"[^"]+"/'
+                                         ,''
+                                         ,$arrEmailAddress[0]);
+        $arrTempAddress[1] = $arrEmailAddress[1];
+        $strTempAddress = $arrTempAddress[0] . $arrTempAddress[1];
+        // Then check - should be no "@" symbols.
+        if (strrpos($strTempAddress, '@') !== false) {
+            // "@" symbol found
+            return false;
+        }
+
+        // Check local portion
+        if (!$this->check_local_portion($arrEmailAddress[0])) {
+            return false;
+        }
+
+        // Check domain portion
+        if (!$this->check_domain_portion($arrEmailAddress[1])) {
+            return false;
+        }
+
+        // If we're still here, all checks above passed. Email is valid.
+        return true;
+
+    }
+
+    /**
+     * Checks email section before "@" symbol for validity
+     * @param   strLocalPortion     Text to be checked
+     * @return  True if local portion is valid, false if not
+     */
+    function check_local_portion($strLocalPortion) {
+        // Local portion can only be from 1 to 64 characters, inclusive.
+        // Please note that servers are encouraged to accept longer local
+        // parts than 64 characters.
+        if (!$this->check_text_length($strLocalPortion, 1, 64)) {
+            return false;
+        }
+        // Local portion must be:
+        // 1) a dot-atom (strings separated by periods)
+        // 2) a quoted string
+        // 3) an obsolete format string (combination of the above)
+        $arrLocalPortion = explode('.', $strLocalPortion);
+        for ($i = 0, $max = sizeof($arrLocalPortion); $i < $max; $i++) {
+             if (!preg_match('.^('
+                            .    '([A-Za-z0-9!#$%&\'*+/=?^_`{|}~-]'
+                            .    '[A-Za-z0-9!#$%&\'*+/=?^_`{|}~-]{0,63})'
+                            .'|'
+                            .    '("[^\\\"]{0,62}")'
+                            .')$.'
+                            ,$arrLocalPortion[$i])) {
+                return false;
+            }
+        }
+        return true;
+    }
+
+    /**
+     * Checks email section after "@" symbol for validity
+     * @param   strDomainPortion     Text to be checked
+     * @return  True if domain portion is valid, false if not
+     */
+    function check_domain_portion($strDomainPortion) {
+        // Total domain can only be from 1 to 255 characters, inclusive
+        if (!$this->check_text_length($strDomainPortion, 1, 255)) {
+            return false;
+        }
+        // Check if domain is IP, possibly enclosed in square brackets.
+        if (preg_match('/^(25[0-5]|2[0-4][0-9]|1[0-9][0-9]|[1-9]?[0-9])'
+           .'(\.(25[0-5]|2[0-4][0-9]|1[0-9][0-9]|[1-9]?[0-9])){3}$/'
+           ,$strDomainPortion) ||
+            preg_match('/^\[(25[0-5]|2[0-4][0-9]|1[0-9][0-9]|[1-9]?[0-9])'
+           .'(\.(25[0-5]|2[0-4][0-9]|1[0-9][0-9]|[1-9]?[0-9])){3}\]$/'
+           ,$strDomainPortion)) {
+            return true;
+        } else {
+            $arrDomainPortion = explode('.', $strDomainPortion);
+            if (sizeof($arrDomainPortion) < 2) {
+                return false; // Not enough parts to domain
+            }
+            for ($i = 0, $max = sizeof($arrDomainPortion); $i < $max; $i++) {
+                // Each portion must be between 1 and 63 characters, inclusive
+                if (!$this->check_text_length($arrDomainPortion[$i], 1, 63)) {
+                    return false;
+                }
+                if (!preg_match('/^(([A-Za-z0-9][A-Za-z0-9-]{0,61}[A-Za-z0-9])|'
+                   .'([A-Za-z0-9]+))$/', $arrDomainPortion[$i])) {
+                    return false;
+                }
+            }
+        }
+        return true;
+    }
+
+    /**
+     * Check given text length is between defined bounds
+     * @param   strText     Text to be checked
+     * @param   intMinimum  Minimum acceptable length
+     * @param   intMaximum  Maximum acceptable length
+     * @return  True if string is within bounds (inclusive), false if not
+     */
+    function check_text_length($strText, $intMinimum, $intMaximum) {
+        // Minimum and maximum are both inclusive
+        $intTextLength = strlen($strText);
+        if (($intTextLength < $intMinimum) || ($intTextLength > $intMaximum)) {
+            return false;
+        } else {
+            return true;
+        }
+    }
+}
+
diff --git a/inc/mail.php b/inc/mail.php
index 4b27963ee..a7a878c2d 100644
--- a/inc/mail.php
+++ b/inc/mail.php
@@ -8,6 +8,7 @@
 
   if(!defined('DOKU_INC')) define('DOKU_INC',fullpath(dirname(__FILE__).'/../').'/');
   require_once(DOKU_INC.'inc/utf8.php');
+  require_once(DOKU_INC.'inc/EmailAddressValidator.php');
 
   // end of line for mail lines - RFC822 says CRLF but postfix (and other MTAs?)
   // think different
@@ -174,26 +175,14 @@ function mail_encode_address($string,$header='',$names=true){
 }
 
 /**
- * Uses a regular expresion to check if a given mail address is valid
- *
- * May not be completly RFC conform!
- * @link    http://www.faqs.org/rfcs/rfc2822.html    (paras 3.4.1 & 3.2.4)
- *
- * @author  Chris Smith <chris@jalakai.co.uk>
+ * Check if a given mail address is valid
  *
  * @param   string $email the address to check
  * @return  bool          true if address is valid
  */
-
-// patterns for use in email detection and validation
-// NOTE: there is an unquoted '/' in RFC2822_ATEXT, it must remain unquoted to be used in the parser
-//       the pattern uses non-capturing groups as captured groups aren't allowed in the parser
-//       select pattern delimiters with care!
-if (!defined('RFC2822_ATEXT')) define('RFC2822_ATEXT',"0-9a-zA-Z!#$%&'*+/=?^_`{|}~-");
-if (!defined('PREG_PATTERN_VALID_EMAIL')) define('PREG_PATTERN_VALID_EMAIL', '['.RFC2822_ATEXT.']+(?:\.['.RFC2822_ATEXT.']+)*@(?i:[0-9a-z][0-9a-z-]*\.)+(?i:[a-z]{2,4}|museum|travel)');
-
 function mail_isvalid($email){
-  return preg_match('<^'.PREG_PATTERN_VALID_EMAIL.'$>i', $email);
+    $validator = new EmailAddressValidator;
+    return $validator->check_email_address($email);
 }
 
 /**