parser.php

来自「PHP 知识管理系统（基于树结构的知识管理系统）, 英文原版的PHP源码。」· PHP 代码 · 共 764 行 · 第 1/2 页

     *    leaves immediately afterwoods.
     *    @param string $mode    Mode to test.
     *    @return boolean        True if this is the exit mode.
     *    @access private
     */
    function _isSpecialMode($mode) {
        return (strncmp($mode, "_", 1) == 0);
    }
    
    /**
     *    Strips the magic underscore marking single token
     *    modes.
     *    @param string $mode    Mode to decode.
     *    @return string         Underlying mode name.
     *    @access private
     */
    function _decodeSpecial($mode) {
        return substr($mode, 1);
    }
    
    /**
     *    Calls the parser method named after the current
     *    mode. Empty content will be ignored. The lexer
     *    has a parser handler for each mode in the lexer.
     *    @param string $content        Text parsed.
     *    @param boolean $is_match      Token is recognised rather
     *                                  than unparsed data.
     *    @access private
     */
    function _invokeParser($content, $is_match) {
        if (($content === '') || ($content === false)) {
            return true;
        }
        $handler = $this->_mode_handlers[$this->_mode->getCurrent()];
        return $this->_parser->$handler($content, $is_match);
    }
    
    /**
     *    Tries to match a chunk of text and if successful
     *    removes the recognised chunk and any leading
     *    unparsed data. Empty strings will not be matched.
     *    @param string $raw         The subject to parse. This is the
     *                               content that will be eaten.
     *    @return array/boolean      Three item list of unparsed
     *                               content followed by the
     *                               recognised token and finally the
     *                               action the parser is to take.
     *                               True if no match, false if there
     *                               is a parsing error.
     *    @access private
     */
    function _reduce($raw) {
        if ($action = $this->_regexes[$this->_mode->getCurrent()]->match($raw, $match)) {
            $unparsed_character_count = strpos($raw, $match);
            $unparsed = substr($raw, 0, $unparsed_character_count);
            $raw = substr($raw, $unparsed_character_count + strlen($match));
            return array($raw, $unparsed, $match, $action);
        }
        return true;
    }
}

/**
 *    Breaks HTML into SAX events.
 *    @package SimpleTest
 *    @subpackage WebTester
 */
class SimpleHtmlLexer extends SimpleLexer {
    
    /**
     *    Sets up the lexer with case insensitive matching
     *    and adds the HTML handlers.
     *    @param SimpleSaxParser $parser  Handling strategy by
     *                                    reference.
     *    @access public
     */
    function SimpleHtmlLexer(&$parser) {
        $this->SimpleLexer($parser, 'text');
        $this->mapHandler('text', 'acceptTextToken');
        $this->_addSkipping();
        foreach ($this->_getParsedTags() as $tag) {
            $this->_addTag($tag);
        }
        $this->_addInTagTokens();
    }
    
    /**
     *    List of parsed tags. Others are ignored.
     *    @return array        List of searched for tags.
     *    @access private
     */
    function _getParsedTags() {
        return array('a', 'base', 'title', 'form', 'input', 'button', 'textarea', 'select',
                'option', 'frameset', 'frame', 'label');
    }
    
    /**
     *    The lexer has to skip certain sections such
     *    as server code, client code and styles.
     *    @access private
     */
    function _addSkipping() {
        $this->mapHandler('css', 'ignore');
        $this->addEntryPattern('<style', 'text', 'css');
        $this->addExitPattern('</style>', 'css');
        $this->mapHandler('js', 'ignore');
        $this->addEntryPattern('<script', 'text', 'js');
        $this->addExitPattern('</script>', 'js');
        $this->mapHandler('comment', 'ignore');
        $this->addEntryPattern('<!--', 'text', 'comment');
        $this->addExitPattern('-->', 'comment');
    }
    
    /**
     *    Pattern matches to start and end a tag.
     *    @param string $tag          Name of tag to scan for.
     *    @access private
     */
    function _addTag($tag) {
        $this->addSpecialPattern("</$tag>", 'text', 'acceptEndToken');
        $this->addEntryPattern("<$tag", 'text', 'tag');
    }
    
    /**
     *    Pattern matches to parse the inside of a tag
     *    including the attributes and their quoting.
     *    @access private
     */
    function _addInTagTokens() {
        $this->mapHandler('tag', 'acceptStartToken');
        $this->addSpecialPattern('\s+', 'tag', 'ignore');
        $this->_addAttributeTokens();
        $this->addExitPattern('/>', 'tag');
        $this->addExitPattern('>', 'tag');
    }
    
    /**
     *    Matches attributes that are either single quoted,
     *    double quoted or unquoted.
     *    @access private
     */
    function _addAttributeTokens() {
        $this->mapHandler('dq_attribute', 'acceptAttributeToken');
        $this->addEntryPattern('=\s*"', 'tag', 'dq_attribute');
        $this->addPattern("\\\\\"", 'dq_attribute');
        $this->addExitPattern('"', 'dq_attribute');
        $this->mapHandler('sq_attribute', 'acceptAttributeToken');
        $this->addEntryPattern("=\s*'", 'tag', 'sq_attribute');
        $this->addPattern("\\\\'", 'sq_attribute');
        $this->addExitPattern("'", 'sq_attribute');
        $this->mapHandler('uq_attribute', 'acceptAttributeToken');
        $this->addSpecialPattern('=\s*[^>\s]*', 'tag', 'uq_attribute');
    }
}

/**
 *    Converts HTML tokens into selected SAX events.
 *    @package SimpleTest
 *    @subpackage WebTester
 */
class SimpleHtmlSaxParser {
    var $_lexer;
    var $_listener;
    var $_tag;
    var $_attributes;
    var $_current_attribute;
    
    /**
     *    Sets the listener.
     *    @param SimpleSaxListener $listener    SAX event handler.
     *    @access public
     */
    function SimpleHtmlSaxParser(&$listener) {
        $this->_listener = &$listener;
        $this->_lexer = &$this->createLexer($this);
        $this->_tag = '';
        $this->_attributes = array();
        $this->_current_attribute = '';
    }
    
    /**
     *    Runs the content through the lexer which
     *    should call back to the acceptors.
     *    @param string $raw      Page text to parse.
     *    @return boolean         False if parse error.
     *    @access public
     */
    function parse($raw) {
        return $this->_lexer->parse($raw);
    }
    
    /**
     *    Sets up the matching lexer. Starts in 'text' mode.
     *    @param SimpleSaxParser $parser    Event generator, usually $self.
     *    @return SimpleLexer               Lexer suitable for this parser.
     *    @access public
     *    @static
     */
    function &createLexer(&$parser) {
        $lexer = &new SimpleHtmlLexer($parser);
        return $lexer;
    }
    
    /**
     *    Accepts a token from the tag mode. If the
     *    starting element completes then the element
     *    is dispatched and the current attributes
     *    set back to empty. The element or attribute
     *    name is converted to lower case.
     *    @param string $token     Incoming characters.
     *    @param integer $event    Lexer event type.
     *    @return boolean          False if parse error.
     *    @access public
     */
    function acceptStartToken($token, $event) {
        if ($event == LEXER_ENTER) {
            $this->_tag = strtolower(substr($token, 1));
            return true;
        }
        if ($event == LEXER_EXIT) {
            $success = $this->_listener->startElement(
                    $this->_tag,
                    $this->_attributes);
            $this->_tag = '';
            $this->_attributes = array();
            return $success;
        }
        if ($token != '=') {
            $this->_current_attribute = strtolower(SimpleHtmlSaxParser::decodeHtml($token));
            $this->_attributes[$this->_current_attribute] = '';
        }
        return true;
    }
    
    /**
     *    Accepts a token from the end tag mode.
     *    The element name is converted to lower case.
     *    @param string $token     Incoming characters.
     *    @param integer $event    Lexer event type.
     *    @return boolean          False if parse error.
     *    @access public
     */
    function acceptEndToken($token, $event) {
        if (! preg_match('/<\/(.*)>/', $token, $matches)) {
            return false;
        }
        return $this->_listener->endElement(strtolower($matches[1]));
    }
    
    /**
     *    Part of the tag data.
     *    @param string $token     Incoming characters.
     *    @param integer $event    Lexer event type.
     *    @return boolean          False if parse error.
     *    @access public
     */
    function acceptAttributeToken($token, $event) {
        if ($this->_current_attribute) {
            if ($event == LEXER_UNMATCHED) {
                $this->_attributes[$this->_current_attribute] .=
                        SimpleHtmlSaxParser::decodeHtml($token);
            }
            if ($event == LEXER_SPECIAL) {
                $this->_attributes[$this->_current_attribute] .=
                        preg_replace('/^=\s*/' , '', SimpleHtmlSaxParser::decodeHtml($token));
            }
        }
        return true;
    }
    
    /**
     *    A character entity.
     *    @param string $token    Incoming characters.
     *    @param integer $event   Lexer event type.
     *    @return boolean         False if parse error.
     *    @access public
     */
    function acceptEntityToken($token, $event) {
    }
    
    /**
     *    Character data between tags regarded as
     *    important.
     *    @param string $token     Incoming characters.
     *    @param integer $event    Lexer event type.
     *    @return boolean          False if parse error.
     *    @access public
     */
    function acceptTextToken($token, $event) {
        return $this->_listener->addContent($token);
    }
    
    /**
     *    Incoming data to be ignored.
     *    @param string $token     Incoming characters.
     *    @param integer $event    Lexer event type.
     *    @return boolean          False if parse error.
     *    @access public
     */
    function ignore($token, $event) {
        return true;
    }
    
    /**
     *    Decodes any HTML entities.
     *    @param string $html    Incoming HTML.
     *    @return string         Outgoing plain text.
     *    @access public
     *    @static
     */
    function decodeHtml($html) {
        return html_entity_decode($html, ENT_QUOTES);
    }
    
    /**
     *    Turns HTML into text browser visible text. Images
     *    are converted to their alt text and tags are supressed.
     *    Entities are converted to their visible representation.
     *    @param string $html        HTML to convert.
     *    @return string             Plain text.
     *    @access public
     *    @static
     */
    function normalise($html) {
        $text = preg_replace('|<!--.*?-->|', '', $html);
        $text = preg_replace('|<script[^>]*>.*?</script>|', '', $text);
        $text = preg_replace('|<img[^>]*alt\s*=\s*"([^"]*)"[^>]*>|', ' \1 ', $text);
        $text = preg_replace('|<img[^>]*alt\s*=\s*\'([^\']*)\'[^>]*>|', ' \1 ', $text);
        $text = preg_replace('|<img[^>]*alt\s*=\s*([a-zA-Z_]+)[^>]*>|', ' \1 ', $text);
        $text = preg_replace('|<[^>]*>|', '', $text);
        $text = SimpleHtmlSaxParser::decodeHtml($text);
        $text = preg_replace('|\s+|', ' ', $text);
        return trim(trim($text), "\xA0");        // TODO: The \xAO is a &nbsp;. Add a test for this.
    }
}

/**
 *    SAX event handler.
 *    @package SimpleTest
 *    @subpackage WebTester
 *    @abstract
 */
class SimpleSaxListener {
    
    /**
     *    Sets the document to write to.
     *    @access public
     */
    function SimpleSaxListener() {
    }
    
    /**
     *    Start of element event.
     *    @param string $name        Element name.
     *    @param hash $attributes    Name value pairs.
     *                               Attributes without content
     *                               are marked as true.
     *    @return boolean            False on parse error.
     *    @access public
     */
    function startElement($name, $attributes) {
    }
    
    /**
     *    End of element event.
     *    @param string $name        Element name.
     *    @return boolean            False on parse error.
     *    @access public
     */
    function endElement($name) {
    }
    
    /**
     *    Unparsed, but relevant data.
     *    @param string $text        May include unparsed tags.
     *    @return boolean            False on parse error.
     *    @access public
     */
    function addContent($text) {
    }
}
?>