introduced a simplified mime decoder class

Signed-off-by: Janos SUTO <sj@acts.hu>

webui/model/search/message.php

@@ -1,7 +1,7 @@
 <?php
 
-require 'Zend/Mime/Decode.php';
+require DIR_SYSTEM . 'helper/mime.php';
-require 'Zend/Exception.php';
+
 
 class ModelSearchMessage extends Model {
 
@@ -12,24 +12,6 @@ class ModelSearchMessage extends Model {
    public $message;
    private $verification = 0;
 
-   public function get_boundary($line='') {
-      $parts = explode(";", $line);
-
-      for($i=0; $i<count($parts); $i++) {
-         if(stristr($parts[$i], "boundary")) {
-            $parts[$i] = preg_replace("/boundary\s{0,}=\s{0,}/i", "boundary=", $parts[$i]);
-            $parts[$i] = preg_replace("/\"\;{0,1}/", "", $parts[$i]);
-            $parts[$i] = preg_replace("/\'/", "", $parts[$i]);
-
-            $b = explode("boundary=", $parts[$i]);
-
-            return rtrim($b[count($b)-1]);
-         }
-      }
-
-      return "";
-   }
-
 
    public function verify_message($piler_id = '', $data = '') {
       if($piler_id == '') { return 0; }
@@ -122,11 +104,11 @@ class ModelSearchMessage extends Model {
 
       $msg = $this->get_raw_message($id);
 
-      Zend_Mime_Decode::splitMessageRaw($msg, $headers, $body);
+      Piler_Mime_Decode::splitMessageRaw($msg, $headers, $body);
 
       $has_journal = $this->remove_journal($headers);
 
-      $headers = $this->escape_lt_gt_symbols($headers);
+      $headers = Piler_Mime_Decode::escape_lt_gt_symbols($headers);
 
       return array('headers' => $headers, 'has_journal' => $has_journal);
    }
@@ -170,7 +152,7 @@ class ModelSearchMessage extends Model {
 
          $p = '';
 
-         $data = $this->escape_lt_gt_symbols($data);
+         $data = Piler_Mime_Decode::escape_lt_gt_symbols($data);
       }
 
       return $data;
@@ -228,53 +210,22 @@ class ModelSearchMessage extends Model {
 
       $has_journal = $this->remove_journal($msg);
 
-      Zend_Mime_Decode::splitMessage($msg, $headers, $body);
+      Piler_Mime_Decode::splitMessage($msg, $headers, $body);
-      $boundary = $this->get_boundary($headers['content-type']);
 
-      if(is_array($headers['from'])) { $headers['from'] = $headers['from'][0]; }
+      for($i=0; $i<count(Piler_Mime_Decode::HEADER_FIELDS); $i++) {
-      if(is_array($headers['to'])) { $headers['to'] = $headers['to'][0]; }
+         if(isset($headers[Piler_Mime_Decode::HEADER_FIELDS[$i]]) && is_array($headers[Piler_Mime_Decode::HEADER_FIELDS[$i]])) {
-      if(is_array($headers['cc'])) { $headers['cc'] = $headers['cc'][0]; }
+            $headers[Piler_Mime_Decode::HEADER_FIELDS[$i]] = $headers[Piler_Mime_Decode::HEADER_FIELDS[$i][0]];
-      if(is_array($headers['subject'])) { $headers['subject'] = $headers['subject'][0]; }
-      if(is_array($headers['date'])) { $headers['date'] = $headers['date'][0]; }
-
-      if(isset($headers['from'])) $from .= $this->escape_lt_gt_symbols($headers['from']);
-      if(isset($headers['to'])) $to .= $this->escape_lt_gt_symbols($headers['to']);
-      if(isset($headers['cc'])) $cc .= $this->escape_lt_gt_symbols($headers['cc']);
-      if(isset($headers['subject'])) $subject .= $this->escape_lt_gt_symbols($headers['subject']);
-      if(isset($headers['date'])) $date .= $headers['date'];
-
-      $this->message = array(
-                             'text/plain' => '',
-                             'text/html' => ''
-                            );
-
-      $this->extract_textuals_from_mime_parts($headers, $body, $boundary);
-
-      return array('from' => $from,
-                   'to' => $to,
-                   'cc' => $cc,
-                   'subject' => $this->highlight_search_terms($subject, $terms),
-                   'date' => $date,
-                   'message' => $this->message['text/html'] ? $this->message['text/html'] : $this->message['text/plain'],
-                   'has_journal' => $has_journal,
-                   'verification' => $this->verification
-            );
          }
 
-
+         if(Piler_Mime_Decode::HEADER_FIELDS[$i] == 'date') {
-   private function extract_textuals_from_mime_parts($headers = array(), $body = '', $boundary = '') {
+            ${Piler_Mime_Decode::HEADER_FIELDS[$i]} .= $headers[Piler_Mime_Decode::HEADER_FIELDS[$i]];
-      $mime_parts = array();
-
-      if($boundary) {
-         try {
-            $mime_parts = Zend_Mime_Decode::splitMessageStruct($body, $boundary);
-         }
-         catch (Exception $e) {
-            syslog(LOG_INFO, "Caught exception: " . $e->getMessage());
-         }
          } else {
-         $mime_parts[] = array('header' => $headers, 'body' => $body);
+            ${Piler_Mime_Decode::HEADER_FIELDS[$i]} .= Piler_Mime_Decode::escape_lt_gt_symbols($headers[Piler_Mime_Decode::HEADER_FIELDS[$i]]);
          }
+      }
+
+
+      Piler_Mime_Decode::parseMessage($msg, $parts);
 
       require_once DIR_SYSTEM . 'helper/HTMLPurifier.standalone.php';
 
@@ -285,69 +236,33 @@ class ModelSearchMessage extends Model {
 
       $purifier = new HTMLPurifier($config);
 
-      for($i=0; $i<count($mime_parts); $i++) {
+      $this->message = array(
-         $mime = array(
+                             'text/plain' => '',
-                       'content-type' => '',
+                             'text/html' => ''
-                       'encoding' => ''
                             );
 
-         if(isset($mime_parts[$i]['header']['content-type'])) {
+      for($i=0; $i<count($parts); $i++) {
-            $mime['content-type'] = Zend_Mime_Decode::splitContentType($mime_parts[$i]['header']['content-type']);
+
+         $body = Piler_Mime_Decode::fixMimeBodyPart($parts[$i]['headers'], $parts[$i]['body']);
+
+         if($parts[$i]['headers']['content-type']['type'] == 'text/html') {
+            $this->message['text/html'] = $purifier->purify($body);
          }
-         /*
+         else {
-           Fix the mime type for some emails having a single textual body part
+            $this->message['text/plain'] = $body;
-           without the Content-type header.
-          */
-         else if (count($mime_parts) == 1) {
-            $mime['content-type']['type'] = 'text/plain';
          }
 
-         $mime['content-type']['type'] = strtolower($mime['content-type']['type']);
-
-         if(in_array($mime['content-type']['type'], array('multipart/mixed', 'multipart/related', 'multipart/alternative')))
-            $this->extract_textuals_from_mime_parts($mime_parts[$i]['header'], $mime_parts[$i]['body'], $mime['content-type']['boundary']);
-
-         if(isset($mime_parts[$i]['header']['content-transfer-encoding']))
-            $mime['encoding'] = $mime_parts[$i]['header']['content-transfer-encoding'];
-
-         if(in_array($mime['content-type']['type'], array('text/plain', 'text/html')))
-            $this->message[$mime['content-type']['type']] .= $this->fix_mime_body_part($purifier, $mime, $mime_parts[$i]['body']);
-      }
       }
 
-
+      return array('from' => $from,
-   private function fix_mime_body_part($purifier, $mime = array(), $body = '') {
+                   'to' => $to,
-      if($mime['encoding'] == 'quoted-printable')
+                   'cc' => $cc,
-         $body = Zend_Mime_Decode::decodeQuotedPrintable($body);
+                   'subject' => $this->highlight_search_terms($subject, $terms),
-
+                   'date' => $date,
-      if($mime['encoding'] == 'base64')
+                   'message' => $this->message['text/html'] ? $this->message['text/html'] : $this->message['text/plain'],
-         $body = base64_decode($body);
+                   'has_journal' => $has_journal,
-
+                   'verification' => $this->verification
-      if(strtolower($mime['content-type']['charset']) != 'utf-8')
+            );
-         $body = iconv($mime['content-type']['charset'], 'utf-8' . '//IGNORE', $body);
-
-
-      if(strtolower($mime['content-type']['type']) == 'text/plain') {
-
-         $body = $this->escape_lt_gt_symbols($body);
-
-         $body = preg_replace("/\n/", "<br />\n", $body);
-         $body = "\n" . $this->print_nicely($body);
-      }
-
-      if(strtolower($mime['content-type']['type']) == 'text/html') {
-         $body = $purifier->purify($body);
-      }
-
-      return $body;
-   }
-
-
-   private function escape_lt_gt_symbols($s = '') {
-      $s = preg_replace("/</", "&lt;", $s);
-      $s = preg_replace("/>/", "&gt;", $s);
-
-      return $s;
    }
 
 
@@ -405,25 +320,6 @@ class ModelSearchMessage extends Model {
    }
 
 
-   private function print_nicely($chunk) {
-      $k = 0;
-      $nice_chunk = "";
-
-      $x = explode(" ", $chunk);
-
-      for($i=0; $i<count($x); $i++){
-         $nice_chunk .= $x[$i] . " ";
-         $k += strlen($x[$i]);
-
-         if(strstr($x[$i], "\n")){ $k = 0; }
-
-         if($k > 70){ $nice_chunk .= "\n"; $k = 0; }
-      }
-
-      return $nice_chunk;
-   }
-
-
    public function NiceSize($size) {
       if($size < 1000) return "1k";
       if($size < 100000) return round($size/1000) . "k";

webui/system/helper/mime.php

@@ -0,0 +1,303 @@
+<?php
+
+
+class Piler_Mime_Decode {
+   const HEADER_FIELDS = ['from', 'to', 'cc', 'subject', 'date'];
+
+
+   public static function parseMessage($message, &$result) {
+
+      self::splitMessage($message, $headers, $body);
+
+      $boundary = self::getBoundary($headers);
+
+      // No boundary defined
+
+      if($boundary == '') {
+         if($headers['content-type']['type'] == "message/rfc822") {
+            self::parseMessage($body, $result);
+         }
+         else {
+            $result[] = array(
+               'headers' => $headers,
+               'body' => $body
+            );
+         }
+
+         return;
+      }
+
+      $parts = self::splitMime($body, $boundary);
+
+      for($i=0; $i<count($parts); $i++) {
+
+         self::splitMessage($parts[$i], $headers, $body);
+
+         $boundary = self::getBoundary($headers);
+         if($boundary) {
+            self::parseMessage($parts[$i], $result);
+         }
+         else {
+            if(in_array($headers['content-type']['type'], ["text/plain", "text/html"])) {
+               $result[] = array('headers' => $headers, 'body' => $body);
+            }
+            else if($headers['content-type']['type'] == "message/rfc822") {
+               self::parseMessage($body, $result);
+            }
+         }
+      }
+   }
+
+
+   public static function splitMime($body, $boundary) {
+      $start = 0;
+      $res = array();
+
+      $body = self::remove_LF($body);
+
+      // Extract the mime parts excluding the boundary itself
+
+      $p = strpos($body, '--' . $boundary . "\n", $start);
+      if($p === false) {
+         // no parts found!
+         return array();
+      }
+
+      // Position after first boundary line
+
+      $start = $p + 3 + strlen($boundary);
+
+      while(($p = strpos($body, '--' . $boundary . "\n", $start)) !== false) {
+         $res[] = substr($body, $start, $p-$start);
+         $start = $p + 3 + strlen($boundary);
+      }
+
+      // No more parts, find end boundary
+
+      $p = strpos($body, '--' . $boundary . '--', $start);
+      if($p === false) {
+         return array();
+      }
+
+      // The remaining part also needs to be parsed:
+      $res[] = substr($body, $start, $p - $start);
+
+      return $res;
+   }   
+
+
+   public static function splitMessage($message, &$headers, &$body, $EOL = "\n") {
+      self::splitMessageRaw($message, $headers, $body);
+      $headers = self::splitHeaders($headers);
+   }
+
+
+   public static function splitMessageRaw($message, &$headers, &$body, $EOL = "\n") {
+      $headers = [];
+      $body = '';
+
+      $message = self::remove_LF($message);
+
+      // Find an empty line between headers and body, otherwise we got a header-only message
+
+      if(strpos($message, $EOL . $EOL)) {
+         list($headers, $body) = explode($EOL . $EOL, $message, 2);
+      }
+      else {
+         $headers = $message;
+      }
+   }
+
+
+   public static function splitHeaders($headers) {
+      $headers = self::headersToArray($headers);
+
+      // normalize header names
+      foreach ($headers as $name => $header) {
+         $lower = strtolower($name);
+         if($lower == $name) {
+            continue;
+         }
+
+         unset($headers[$name]);
+
+         if(!isset($headers[$lower])) {
+            $headers[$lower] = $header;
+            continue;
+         }
+
+         if(is_array($headers[$lower])) {
+            $headers[$lower][] = $header;
+            continue;
+         }
+
+         $headers[$lower] = array($headers[$lower], $header);
+      }
+
+      // Add some default values, if they are missing
+
+      if(!isset($headers['content-type'])) { $headers['content-type'] = 'text/plain'; }
+
+      for($i=0; $i<count(self::HEADER_FIELDS); $i++) {
+         if(!isset($headers[self::HEADER_FIELDS[$i]])) { $headers[self::HEADER_FIELDS[$i]] = ''; }
+
+         $headers[self::HEADER_FIELDS[$i]] = iconv_mime_decode($headers[self::HEADER_FIELDS[$i]], ICONV_MIME_DECODE_CONTINUE_ON_ERROR);
+      }
+
+      $headers['content-type'] = self::splitContentType($headers['content-type']);
+
+      $headers['content-type']['type'] = strtolower($headers['content-type']['type']);
+
+      return $headers;
+   }
+
+
+   public static function headersToArray($headers = '') {
+      $token = '';
+      $last_token = '';
+      $result = array();
+
+      $headers = explode("\n", $headers);
+
+      foreach($headers as $h) {
+
+         // Handle cases when there's no whitespace between the header key and value
+         // eg. Subject:som
+
+         $h = preg_replace("/^([\S]+):(\S)/", '${1}: ${2}', $h);
+         $h = preg_replace("/\s{1,}/", " ", $h);
+
+         $line = preg_split("/\s/", $h);
+
+         // Skip line if it doesn't have a colon (:) and the 1st character is not a whitespace
+
+         if(!ctype_space($h[0]) && !strchr($h, ':')) { continue; }
+
+         if($line) {
+            if(substr($line[0], -1) == ':') {
+               $token = array_shift($line);
+               $token = rtrim($token, ':');
+
+               $last_token = $token;
+            }
+            else {
+               $token = '';
+            }
+
+            $line_str = implode(" ", $line);
+
+            if(!isset($result[$last_token])) {
+               $result[$last_token] = $line_str;
+            }
+            else {
+               if($token) {
+                  $result[$last_token] .= "\n";
+               }
+
+               $result[$last_token] .= ' ' . $line_str;
+            }
+         } 
+
+      }
+
+      while(list($k, $v) = each($result)) {
+         if(strchr($v, "\n")) {
+            $result[$k] = explode("\n", $v);
+         }
+      }
+
+      return $result;
+   }
+
+
+   public static function splitContentType($field = '') {
+      $split = array();
+      $what  = 'type';
+
+      $field = $what . '=' . $field;
+      if(!preg_match_all('%([^=\s]+)\s*=\s*("[^"]+"|[^;]+)(;\s*|$)%', $field, $matches)) {
+         return $split;
+      }
+
+      $split = array();
+      foreach ($matches[1] as $key => $name) {
+         $name = strtolower($name);
+         if($matches[2][$key][0] == '"') {
+            $split[$name] = substr($matches[2][$key], 1, -1);
+         } else {
+            $split[$name] = $matches[2][$key];
+         }
+      }
+
+      return $split;
+   }
+
+
+   public static function remove_LF($message = '') {
+      return str_replace("\r", "", $message);
+      //return preg_replace("/\r/", "", $message);
+   }
+
+
+   public static function getBoundary($headers = array()) {
+      if(isset($headers['content-type']['boundary'])) {
+         return $headers['content-type']['boundary'];
+      }
+
+      return '';
+   }
+
+
+   public static function fixMimeBodyPart($headers = array(), $body = '') {
+
+      if(isset($headers['content-transfer-encoding'])) {
+         if($headers['content-transfer-encoding'] == 'quoted-printable') {
+            $body = quoted_printable_decode($body);
+         }
+
+         if($headers['content-transfer-encoding'] == 'base64') {
+            $body = base64_decode($body);
+         }
+      }
+
+      if(isset($headers['content-type']['charset'])) {
+         $body = iconv($headers['content-type']['charset'], 'utf-8' . '//IGNORE', $body);
+      }
+
+      if(strtolower($headers['content-type']['type']) == 'text/plain') {
+         $body = self::escape_lt_gt_symbols($body);
+         $body = preg_replace("/\n/", "<br />\n", $body);
+         $body = "\n" . self::printNicely($body);
+      }
+
+      return $body;
+   }
+
+
+   public static function escape_lt_gt_symbols($s = '') {
+      $s = preg_replace("/</", "&lt;", $s);
+      $s = preg_replace("/>/", "&gt;", $s);
+
+      return $s;
+   }
+
+
+   public static function printNicely($s = '') {
+      $k = 0;
+      $nice = "";
+
+      $x = explode(" ", $s);
+
+      for($i=0; $i<count($x); $i++){
+         $nice .= $x[$i] . " ";
+         $k += strlen($x[$i]);
+
+         if(strstr($x[$i], "\n")){ $k = 0; }
+
+         if($k > 70){ $nice .= "\n"; $k = 0; }
+      }
+
+      return $nice;
+   }
+
+}