Subversion Repositories Applications.papyrus

<?php

// $Id: ezmlm-parser.php,v 1.2.4.1 2007-03-12 11:22:12 alexandre_tb Exp $

//

// ezmlm-parser.php - ezmlm-php v2.0

// --------------------------------------------------------------

// Contains all the code for parsing messages.

// It handles all the nessesary decoding, attachments, etc...

// Note this does all the parsing itself now removing the dependancy

// on the mailparse library (as it looks like it will never make

// it into the official inclusion with PHP)...

// --------------------------------------------------------------

require_once("ezmlm.php");

require_once("Mail/mimeDecode.php") ;

// CLASS: ezmlm-parser

class ezmlm_parser extends ezmlm_php {

    var $headers;           // the full untouched headers of the message

    var $body;              // the full untouched (but decoded) body (this is not $this->parts[0]->body)

	var $parts;		// all the parts, if it is a multipart message. each part is an ezmlm_parser object...

    // Here's the most accessed headers, everything else can be

    // accessed from the $this->headers array.

    var $to;                // To:

    var $from;              // From:

    var $date;              // Date:

    var $subject;           // Subject:

    var $replyto;           // Reply-To:

    var $contenttype;       // Content-Type:

	var $multipart;		// TRUE if the message is a multipart message

	var $msgfile;		// if parsed from a file, this is the filename...

	// functions

	// recent_msgs - parses and returns an arbitrary number of the most recent messages

	function recent_msgs($show = 20, $month = "") {

		$repertoire_archive = opendir($this->listdir . "/archive/");

		$repertoire_message = array() ;

		$dernier_repertoire = 0 ;

		while (false !== ($item = readdir($repertoire_archive))) {

			// $item contient les noms des repertoires

			// on ne garde que ceux qui sont des chiffres

			if (preg_match('/[0-9]+/', $item)) {

				// on ouvre le fichier d index de chaque repertoire

				if ((int) $item > $dernier_repertoire) $dernier_repertoire = (int) $item;

			}

		}

		$tableau_message = array() ;

		$compteur_message = 0 ;

		$fichier_index = fopen ($this->listdir.'/archive/'.$dernier_repertoire.'/index', 'r') ;

		while (!feof($fichier_index)) {

				// Recuperation du numero de message, du hash du sujet et du sujet

				$temp = fgets($fichier_index, 4096);

				preg_match('/([0-9]+): ([a-z]+) (.*)/', $temp, $match) ;

				// dans la seconde on recupere la date, hash auteur et auteur

				$temp = fgets($fichier_index, 4096);

				preg_match('/\t([0-9]+) ([a-zA-Z][a-zA-Z][a-zA-Z]) ([0-9][0-9][0-9][0-9]) ([^;]+);(.*) (.*)/', $temp, $match_deuxieme_ligne) ;

				if ($match[1] != '') {

				$tableau_message[$match[1]] = array ($match[2], $match[3],

									$match_deuxieme_ligne[1].' '.$match_deuxieme_ligne[2].' '.$match_deuxieme_ligne[3],

									$match_deuxieme_ligne[5],

									$match_deuxieme_ligne[6]);

				}

			}

			fclose ($fichier_index);

		// on renverse le tableau pour afficher les derniers messages en premier

		$tableau_message = array_reverse($tableau_message, true);

		// On compte le nombre de message, s il est inferieur $show et que l on est

		// pas dans le premier index, on ouvre le fichier precedent et recupere

		// le n dernier message

		if (count ($tableau_message) < $show && $dernier_repertoire != '0') {

			$avant_dernier_repertoire = $dernier_repertoire - 1 ;

			// On utilise file_get_contents pour renverser le fichier

			$fichier_index = array_reverse(

									explode ("\n",

										preg_replace ('/\n$/', '',

											file_get_contents ($this->listdir.'/archive/'.$avant_dernier_repertoire.'/index')) ), true) ;

			reset ($fichier_index);

			//var_dump ($fichier_index);

			for ($i = count ($tableau_message); $i <= $show; $i++) {

				// Recuperation du numero de message, du hash du sujet et du sujet

				// dans la seconde on recupere la date, hash auteur et auteur

				preg_match('/\t([0-9]+) ([a-zA-Z][a-zA-Z][a-zA-Z]) ([0-9][0-9][0-9][0-9]) ([^;]+);(.*) (.*)/',

									current ($fichier_index), $match_deuxieme_ligne) ;

				preg_match('/([0-9]+): ([a-z]+) (.*)/', next($fichier_index), $match) ;

				next ($fichier_index);

				if ($match[1] != '') {

				$tableau_message[$match[1]] = array ($match[2], $match[3],

									$match_deuxieme_ligne[1].' '.$match_deuxieme_ligne[2].' '.$match_deuxieme_ligne[3],

									$match_deuxieme_ligne[5],

									$match_deuxieme_ligne[6]);

				}

			}

		}

		return ($tableau_message) ;

	}

	// parse_file - opens a file and feeds the data to parse, file can be relative to the listdir

	function parse_file($file,$simple = FALSE) {

		if (!is_file($file)) {

			if (is_file($this->listdir . "/" . $file)) { $file = $this->listdir . "/" . $file; }

			else if (is_file($this->listdir . "/archive/" . $file)) { $file = $this->listdir . "/archive/" . $file; }

			else { return FALSE; }

		}

		$this->msgfile = $file;

        $data = '' ;

		$fd = fopen($file, "r");

		while (!feof($fd)) { $data .= fgets($fd,4096); }

		fclose($fd);

		return $this->parse($data,$simple);

	}

    // parse_file_headers - ouvre un fichier et analyse les entêtes

	function parse_file_headers($file,$simple = FALSE) {

		if (!is_file($file)) {

			if (is_file($this->listdir . "/" . $file)) { $file = $this->listdir . "/" . $file; }

			else if (is_file($this->listdir . "/archive/" . $file)) { $file = $this->listdir . "/archive/" . $file; }

			else { return FALSE; }

		}

		$this->msgfile = $file;

        $data = file_get_contents ($file) ;

        $message = file_get_contents($file) ;

        $mimeDecode = new Mail_mimeDecode($message) ;

        $mailDecode = $mimeDecode->decode() ;

        return $mailDecode ;

		/*$fd = fopen($file, "r");

		while (!feof($fd)) { $data .= fgets($fd,4096); }

		fclose($fd);*/

        if ($this->_get_headers($data, $simple)) return true ;

		return false ;

	}

	// this does all of the work (well it calls two functions that do all the work :)

	// all the decoding a part breaking follows RFC2045 (http://www.faqs.org/rfcs/rfc2045.html)

	function parse($data,$simple = FALSE) {

		if (($this->_get_headers($data,$simple)) && $this->_get_body($data,$simple)) { return TRUE; }

		return FALSE;

	}

	// all of these are internal functions, you shouldn't call them directly...

	// _ct_parse: parse Content-Type headers -> $ct[0] = Full header, $ct[1] = Content-Type, $ct[2] ... $ct[n] = AP's

	function _ct_parse() {

		$instr = $this->headers['content-type'];

		preg_replace('/\(.*\)/','',$instr); // strip rfc822 comments

		if (preg_match('/: /', $instr)) {

			$ct = preg_split('/:/',trim($instr),2);

			$ct = preg_split('/;/',trim($ct[1]));

		} else {

			$ct = preg_split('/;/',trim($instr));

		}

		if (isset($ct[1])) $attrs = preg_split('/[\s\n]/',$ct[1]);

		$i = 2;

		$ct[1] = $ct[0];

		$ct[0] = $this->headers['content-type'];

        if (isset($attrs) && is_array($attrs)) {

            while (list($key, $val) = each($attrs)) {

                if ($val == '') continue;

                $ap = preg_split('/=/',$val,2);

                if (preg_match('/^"/',$ap[1])) { $ap[1] = substr($ap[1],1,strlen($ap[1])-2); }

                $ct[$i] = $ap;

                $i++;

            }

        }

		// are we a multipart message?

		if (preg_match('/^multipart/i', $ct[1])) { $this->multipart = TRUE; }

		return $ct;

	}

	// _get_headers: pulls the headers out of the data and builds the $this->headers array

	function _get_headers($data,$simple = FALSE) {

		$lines = preg_split('/\n/', $data);

		while (list($key, $val) = each($lines)) {

			$val = trim($val);

			if ($val == "") break;

			if (preg_match('/^From[^:].*$/', $val)) continue;	/* strips out any From lines added by the MTA */

			$hdr = preg_split('/: /', $val, 2);

			if (count($hdr) == 1) {

				// this is a continuation of the last header (like a recieved from line)

				$this->headers[$last] .= $val;

			} else {

				$this->headers[strtolower($hdr[0])] = $hdr[1];

                //echo htmlspecialchars($this->headers['from'])."<br />" ;

				$last = strtolower($hdr[0]);

			}

		}

        // ajout alex

        // pour supprimer le problème des ISO...

        // a déplacer ailleur, et appelé avant affichage

        if (preg_match ('/windows-[0-9][0-9][0-9][0-9]/', $this->headers['subject'], $nombre)) {

            $reg_exp = $nombre[0] ;

        } else {

            $reg_exp = 'ISO-8859-15?' ;

        }

        if (preg_match ('/UTF/i', $this->headers['subject'])) $reg_exp = 'UTF-8' ;

        preg_match_all ("/=\?$reg_exp\?(Q|B)\?(.*?)\?=/i", $this->headers['subject'], $match, PREG_PATTERN_ORDER)  ;

        for ($i = 0; $i < count ($match[0]); $i++ ) {

                if ($match[1][$i] == 'Q') {

                    $decode = quoted_printable_decode ($match[2][$i]) ;

                } elseif ($match[1][$i] == 'B') {

                    $decode = base64_decode ($match[2][$i]) ;

                }

                $decode = preg_replace ("/_/", " ", $decode) ;

            if ($reg_exp == 'UTF-8') {

                $decode = utf8_decode ($decode) ;

            }

            $this->headers['subject'] = str_replace ($match[0][$i], $decode, $this->headers['subject']) ;

        }

		// sanity anyone?

		if (!$this->headers['content-type']) { $this->headers['content-type'] = "text/plain; charset=us-ascii"; }

		if (!$simple) { $this->headers['content-type'] = $this->_ct_parse(); }

		return TRUE;

	}

	// _get_body: pulls the body out of the data and fills $this->body, decoding the data if nessesary.

	function _get_body($data,$simple = FALSE) {

		$lines = preg_split('/\n/', $data);

		$doneheaders = FALSE;

		$data = "";

		while (list($key,$val) = each($lines)) {

            //echo htmlspecialchars($val)."<br>";

			if (($val == '') and (!$doneheaders)) {

				$doneheaders = TRUE;

				continue;

			} else if ($doneheaders) {

				$data .= $val . "\n";

			}

		}

		// now here comes the fun part... decoding.

		switch($this->headers['content-transfer-encoding']) {

			case 'binary':

				$this->body = $this->_cte_8bit($this->_cte_qp($this->_cte_binary($data)),$simple);

				break;

			case 'base64':

				$this->body = $this->_cte_8bit($this->_cte_qp($this->_cte_base64($data)),$simple);

				break;

			case 'quoted-printable':

				$this->body = $this->_cte_8bit($this->_cte_qp($data),$simple);

				break;

			case '8bit':

				$this->body = $this->_cte_8bit($data,$simple);

				break;

			case '7bit':		// 7bit doesn't need to be decoded

			default:		// And the fall through as well...

				$this->body = $data;

				break;

		}

        //echo  $this->headers['content-type'][2][1];

        if (isset($this->headers['content-type'][2][1]) && $this->headers['content-type'][2][1] == 'UTF-8') {

                //$this->body = utf8_decode ($this->body) ;

                //echo quoted_printable_decode(utf8_decode ($this->body)) ;

        }

		if ($simple) { return TRUE; }

		// if we are a multipart message then break up the parts and decode, set the appropriate variables.

		// here comes the best part about making ezmlm-php OOP. since each part is just really a little message

		// in itself each part becomes a new parser object and all the wheels turn again... :)

		if ($this->multipart) {

			$boundary = '';

			for ($i = 2; $i <= count($this->headers['content-type']); $i++) {

				if (preg_match('/boundary/i', $this->headers['content-type'][$i][0])) {

					$boundary = $this->headers['content-type'][$i][1];

				}

			}

			if ($boundary != '') {

				$this->_get_parts($this->body,$boundary);

			} else {

				// whoopps... something's not right here. we were told that the message is supposed

				// to be a multipart message, yet the boundary wasn't set in the content type.

				// mark the message as non multipart and add a message to the top of the body.

				$this->multipart = FALSE;

				$this->body = "PARSER ERROR:\nWHILE PARSING THIS MESSAGE AS A MULTIPART MESSAGE AS DEFINED IN RFC2045 THE BOUNDARY IDENTIFIER WAS NOT FOUND!\nTHIS MESSAGE WILL NOT DISPLAY CORRECTLY!\n\n" . $this->body;

			}

		}

		return TRUE;

	}

	// _get_parts: breaks up $data into parts based on $boundary following the rfc specs

	// detailed in section 5 of RFC2046 (http://www.faqs.org/rfcs/rfc2046.html)

	// After the parts are broken up they are then turned into parser objects and the

	// resulting array of parts is set to $this->parts;

	function _get_parts($data,$boundary) {

		$inpart = -1;

		$lines = preg_split('/\n/', $data);

        // La première partie contient l'avertissement pour les client mail ne supportant pas

        // multipart, elle est stocké dans parts[-1]

		while(list($key,$val) = each($lines)) {

			if ($val == "--" . $boundary) { $inpart++; continue; } // start of a part

			else if ($val == "--" . $boundary . "--") { break; } // the end of the last part

			else { $parts[$inpart] .= $val . "\n"; }

		}

		for ($i = 0; $i < count($parts) - 1; $i++) {    // On saute la première partie

			$part[$i] = new ezmlm_parser();

			$part[$i]->parse($parts[$i]);

			$this->parts[$i] = $part[$i];

            //echo $this->parts[$i]."<br>" ;

		}

	}

	// _cte_8bit: decode a content transfer encoding of 8bit

	// NOTE: this function is a little bit special. Since the end result will be displayed in

	// a web browser _cte_8bit decodes ASCII characters > 127 (the US-ASCII table) into the

	// html ordinal equivilant, it also ensures that the messages content-type is changed

	// to include text/html if it changes anything...

	function _cte_8bit($data,$simple = FALSE) {

		if ($simple) { return $data; }

		$changed = FALSE;

		$chars = preg_split('//',$data);

		while (list($key,$val) = each($chars)) {

			if (ord($val) > 127) { $out .= '&#' . ord($val) . ';'; $changed = TRUE; }

			else { $out .= $val; }

		}

		if ($changed) { $this->headers['content-type'][1] = 'text/html'; }

		return $out;

	}

	// _cte_binary: decode a content transfer encoding of binary

	function _cte_binary($data) { return $data; }

	// _cte_base64: decode a content transfer encoding of base64

	function _cte_base64($data) { return base64_decode($data); }

	// _cte_qp: decode a content transfer encoding of quoted_printable

	function _cte_qp($data) {

		// For the time being we'll use PHP's function, it seems to work well enough.

		return quoted_printable_decode($data);

	}

}