<?php

/*
 * This file is part of SwiftMailer.
 * (c) 2004-2009 Chris Corbyn
 *
 * For the full copyright and license information, please view the LICENSE
 * file that was distributed with this source code.
 */

/**
 * Defines the grammar to use for validation, implements the RFC 2822 (and friends) ABNF grammar definitions.
 * @package Swift
 * @subpackage Mime
 * @author Fabien Potencier
 * @author Chris Corbyn
 */
class Swift_Mime_Grammar
{
  /**
   * Special characters used in the syntax which need to be escaped.
   * @var string[]
   * @access private
   */
  private $_specials = array();

  /**
   * Tokens defined in RFC 2822 (and some related RFCs).
   * @var string[]
   * @access private
   */
  private $_grammar = array();
  
  /**
   * Initialize some RFC 2822 (and friends) ABNF grammar definitions.
   * @access protected
   */
  public function __construct()
  {
    $this->_specials = array(
      '(', ')', '<', '>', '[', ']',
      ':', ';', '@', ',', '.', '"'
      );
    
    /*** Refer to RFC 2822 for ABNF grammar ***/
    
    //All basic building blocks
    $this->_grammar['NO-WS-CTL'] = '[\x01-\x08\x0B\x0C\x0E-\x19\x7F]';
    $this->_grammar['WSP'] = '[ \t]';
    $this->_grammar['CRLF'] = '(?:\r\n)';
    $this->_grammar['FWS'] = '(?:(?:' . $this->_grammar['WSP'] . '*' .
        $this->_grammar['CRLF'] . ')?' . $this->_grammar['WSP'] . ')';
    $this->_grammar['text'] = '[\x00-\x08\x0B\x0C\x0E-\x7F]';
    $this->_grammar['quoted-pair'] = '(?:\\\\' . $this->_grammar['text'] . ')';
    $this->_grammar['ctext'] = '(?:' . $this->_grammar['NO-WS-CTL'] .
        '|[\x21-\x27\x2A-\x5B\x5D-\x7E])';
    //Uses recursive PCRE (?1) -- could be a weak point??
    $this->_grammar['ccontent'] = '(?:' . $this->_grammar['ctext'] . '|' .
        $this->_grammar['quoted-pair'] . '|(?1))';
    $this->_grammar['comment'] = '(\((?:' . $this->_grammar['FWS'] . '|' .
        $this->_grammar['ccontent']. ')*' . $this->_grammar['FWS'] . '?\))';
    $this->_grammar['CFWS'] = '(?:(?:' . $this->_grammar['FWS'] . '?' .
        $this->_grammar['comment'] . ')*(?:(?:' . $this->_grammar['FWS'] . '?' .
        $this->_grammar['comment'] . ')|' . $this->_grammar['FWS'] . '))';
    $this->_grammar['qtext'] = '(?:' . $this->_grammar['NO-WS-CTL'] .
        '|[\x21\x23-\x5B\x5D-\x7E])';
    $this->_grammar['qcontent'] = '(?:' . $this->_grammar['qtext'] . '|' .
        $this->_grammar['quoted-pair'] . ')';
    $this->_grammar['quoted-string'] = '(?:' . $this->_grammar['CFWS'] . '?"' .
        '(' . $this->_grammar['FWS'] . '?' . $this->_grammar['qcontent'] . ')*' .
        $this->_grammar['FWS'] . '?"' . $this->_grammar['CFWS'] . '?)';
    $this->_grammar['atext'] = '[a-zA-Z0-9!#\$%&\'\*\+\-\/=\?\^_`\{\}\|~]';
    $this->_grammar['atom'] = '(?:' . $this->_grammar['CFWS'] . '?' .
        $this->_grammar['atext'] . '+' . $this->_grammar['CFWS'] . '?)';
    $this->_grammar['dot-atom-text'] = '(?:' . $this->_grammar['atext'] . '+' .
        '(\.' . $this->_grammar['atext'] . '+)*)';
    $this->_grammar['dot-atom'] = '(?:' . $this->_grammar['CFWS'] . '?' .
        $this->_grammar['dot-atom-text'] . '+' . $this->_grammar['CFWS'] . '?)';
    $this->_grammar['word'] = '(?:' . $this->_grammar['atom'] . '|' .
        $this->_grammar['quoted-string'] . ')';
    $this->_grammar['phrase'] = '(?:' . $this->_grammar['word'] . '+?)';
    $this->_grammar['no-fold-quote'] = '(?:"(?:' . $this->_grammar['qtext'] .
        '|' . $this->_grammar['quoted-pair'] . ')*")';
    $this->_grammar['dtext'] = '(?:' . $this->_grammar['NO-WS-CTL'] .
        '|[\x21-\x5A\x5E-\x7E])';
    $this->_grammar['no-fold-literal'] = '(?:\[(?:' . $this->_grammar['dtext'] .
        '|' . $this->_grammar['quoted-pair'] . ')*\])';
    
    //Message IDs
    $this->_grammar['id-left'] = '(?:' . $this->_grammar['dot-atom-text'] . '|' .
        $this->_grammar['no-fold-quote'] . ')';
    $this->_grammar['id-right'] = '(?:' . $this->_grammar['dot-atom-text'] . '|' .
        $this->_grammar['no-fold-literal'] . ')';
    
    //Addresses, mailboxes and paths
    $this->_grammar['local-part'] = '(?:' . $this->_grammar['dot-atom'] . '|' .
        $this->_grammar['quoted-string'] . ')';
    $this->_grammar['dcontent'] = '(?:' . $this->_grammar['dtext'] . '|' .
        $this->_grammar['quoted-pair'] . ')';
    $this->_grammar['domain-literal'] = '(?:' . $this->_grammar['CFWS'] . '?\[(' .
        $this->_grammar['FWS'] . '?' . $this->_grammar['dcontent'] . ')*?' .
        $this->_grammar['FWS'] . '?\]' . $this->_grammar['CFWS'] . '?)';
    $this->_grammar['domain'] = '(?:' . $this->_grammar['dot-atom'] . '|' .
        $this->_grammar['domain-literal'] . ')';
    $this->_grammar['addr-spec'] = '(?:' . $this->_grammar['local-part'] . '@' .
        $this->_grammar['domain'] . ')';
  }
  
  /**
   * Get the grammar defined for $name token.
   * @param string $name execatly as written in the RFC
   * @return string
   */
  public function getDefinition($name)
  {
    if (array_key_exists($name, $this->_grammar))
    {
      return $this->_grammar[$name];
    }
    else
    {
      throw new Swift_RfcComplianceException(
        "No such grammar '" . $name . "' defined."
        );
    }
  }
  
  /**
   * Returns the tokens defined in RFC 2822 (and some related RFCs).
   * @return array
   */
  public function getGrammarDefinitions()
  {
    return $this->_grammar;
  }
  
  /**
   * Returns the current special characters used in the syntax which need to be escaped.
   * @return array
   */
  public function getSpecials()
  {
    return $this->_specials;
  }
  
  /**
   * Escape special characters in a string (convert to quoted-pairs).
   * @param string $token
   * @param string[] $include additonal chars to escape
   * @param string[] $exclude chars from escaping
   * @return string
   */
  public function escapeSpecials($token, $include = array(),
    $exclude = array())
  {
    foreach (
      array_merge(array('\\'), array_diff($this->_specials, $exclude), $include) as $char)
    {
      $token = str_replace($char, '\\' . $char, $token);
    }
    return $token;
  }
}