aboutsummaryrefslogtreecommitdiffstats
path: root/library/HTMLPurifier/Lexer/DOMLex.php
blob: 7207544545bf58fbc5ae626f4c6adba86f5b594f (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
<?php

/**
 * Parser that uses PHP 5's DOM extension (part of the core).
 *
 * In PHP 5, the DOM XML extension was revamped into DOM and added to the core.
 * It gives us a forgiving HTML parser, which we use to transform the HTML
 * into a DOM, and then into the tokens.  It is blazingly fast (for large
 * documents, it performs twenty times faster than
 * HTMLPurifier_Lexer_DirectLex,and is the default choice for PHP 5.
 *
 * @note Any empty elements will have empty tokens associated with them, even if
 * this is prohibited by the spec. This is cannot be fixed until the spec
 * comes into play.
 *
 * @note PHP's DOM extension does not actually parse any entities, we use
 *       our own function to do that.
 *
 * @warning DOM tends to drop whitespace, which may wreak havoc on indenting.
 *          If this is a huge problem, due to the fact that HTML is hand
 *          edited and you are unable to get a parser cache that caches the
 *          the output of HTML Purifier while keeping the original HTML lying
 *          around, you may want to run Tidy on the resulting output or use
 *          HTMLPurifier_DirectLex
 */

class HTMLPurifier_Lexer_DOMLex extends HTMLPurifier_Lexer
{

    /**
     * @type HTMLPurifier_TokenFactory
     */
    private $factory;

    public function __construct()
    {
        // setup the factory
        parent::__construct();
        $this->factory = new HTMLPurifier_TokenFactory();
    }

    /**
     * @param string $html
     * @param HTMLPurifier_Config $config
     * @param HTMLPurifier_Context $context
     * @return HTMLPurifier_Token[]
     */
    public function tokenizeHTML($html, $config, $context)
    {
        $html = $this->normalize($html, $config, $context);

        // attempt to armor stray angled brackets that cannot possibly
        // form tags and thus are probably being used as emoticons
        if ($config->get('Core.AggressivelyFixLt')) {
            $char = '[^a-z!\/]';
            $comment = "/<!--(.*?)(-->|\z)/is";
            $html = preg_replace_callback($comment, array($this, 'callbackArmorCommentEntities'), $html);
            do {
                $old = $html;
                $html = preg_replace("/<($char)/i", '&lt;\\1', $html);
            } while ($html !== $old);
            $html = preg_replace_callback($comment, array($this, 'callbackUndoCommentSubst'), $html); // fix comments
        }

        // preprocess html, essential for UTF-8
        $html = $this->wrapHTML($html, $config, $context);

        $doc = new DOMDocument();
        $doc->encoding = 'UTF-8'; // theoretically, the above has this covered

        set_error_handler(array($this, 'muteErrorHandler'));
        $doc->loadHTML($html);
        restore_error_handler();

        $tokens = array();
        $this->tokenizeDOM(
            $doc->getElementsByTagName('html')->item(0)-> // <html>
            getElementsByTagName('body')->item(0)-> //   <body>
            getElementsByTagName('div')->item(0), //     <div>
            $tokens
        );
        return $tokens;
    }

    /**
     * Iterative function that tokenizes a node, putting it into an accumulator.
     * To iterate is human, to recurse divine - L. Peter Deutsch
     * @param DOMNode $node DOMNode to be tokenized.
     * @param HTMLPurifier_Token[] $tokens   Array-list of already tokenized tokens.
     * @return HTMLPurifier_Token of node appended to previously passed tokens.
     */
    protected function tokenizeDOM($node, &$tokens)
    {
        $level = 0;
        $nodes = array($level => new HTMLPurifier_Queue(array($node)));
        $closingNodes = array();
        do {
            while (!$nodes[$level]->isEmpty()) {
                $node = $nodes[$level]->shift(); // FIFO
                $collect = $level > 0 ? true : false;
                $needEndingTag = $this->createStartNode($node, $tokens, $collect);
                if ($needEndingTag) {
                    $closingNodes[$level][] = $node;
                }
                if ($node->childNodes && $node->childNodes->length) {
                    $level++;
                    $nodes[$level] = new HTMLPurifier_Queue();
                    foreach ($node->childNodes as $childNode) {
                        $nodes[$level]->push($childNode);
                    }
                }
            }
            $level--;
            if ($level && isset($closingNodes[$level])) {
                while ($node = array_pop($closingNodes[$level])) {
                    $this->createEndNode($node, $tokens);
                }
            }
        } while ($level > 0);
    }

    /**
     * @param DOMNode $node DOMNode to be tokenized.
     * @param HTMLPurifier_Token[] $tokens   Array-list of already tokenized tokens.
     * @param bool $collect  Says whether or start and close are collected, set to
     *                    false at first recursion because it's the implicit DIV
     *                    tag you're dealing with.
     * @return bool if the token needs an endtoken
     * @todo data and tagName properties don't seem to exist in DOMNode?
     */
    protected function createStartNode($node, &$tokens, $collect)
    {
        // intercept non element nodes. WE MUST catch all of them,
        // but we're not getting the character reference nodes because
        // those should have been preprocessed
        if ($node->nodeType === XML_TEXT_NODE) {
            $tokens[] = $this->factory->createText($node->data);
            return false;
        } elseif ($node->nodeType === XML_CDATA_SECTION_NODE) {
            // undo libxml's special treatment of <script> and <style> tags
            $last = end($tokens);
            $data = $node->data;
            // (note $node->tagname is already normalized)
            if ($last instanceof HTMLPurifier_Token_Start && ($last->name == 'script' || $last->name == 'style')) {
                $new_data = trim($data);
                if (substr($new_data, 0, 4) === '<!--') {
                    $data = substr($new_data, 4);
                    if (substr($data, -3) === '-->') {
                        $data = substr($data, 0, -3);
                    } else {
                        // Highly suspicious! Not sure what to do...
                    }
                }
            }
            $tokens[] = $this->factory->createText($this->parseData($data));
            return false;
        } elseif ($node->nodeType === XML_COMMENT_NODE) {
            // this is code is only invoked for comments in script/style in versions
            // of libxml pre-2.6.28 (regular comments, of course, are still
            // handled regularly)
            $tokens[] = $this->factory->createComment($node->data);
            return false;
        } elseif ($node->nodeType !== XML_ELEMENT_NODE) {
            // not-well tested: there may be other nodes we have to grab
            return false;
        }

        $attr = $node->hasAttributes() ? $this->transformAttrToAssoc($node->attributes) : array();

        // We still have to make sure that the element actually IS empty
        if (!$node->childNodes->length) {
            if ($collect) {
                $tokens[] = $this->factory->createEmpty($node->tagName, $attr);
            }
            return false;
        } else {
            if ($collect) {
                $tokens[] = $this->factory->createStart(
                    $tag_name = $node->tagName, // somehow, it get's dropped
                    $attr
                );
            }
            return true;
        }
    }

    /**
     * @param DOMNode $node
     * @param HTMLPurifier_Token[] $tokens
     */
    protected function createEndNode($node, &$tokens)
    {
        $tokens[] = $this->factory->createEnd($node->tagName);
    }


    /**
     * Converts a DOMNamedNodeMap of DOMAttr objects into an assoc array.
     *
     * @param DOMNamedNodeMap $node_map DOMNamedNodeMap of DOMAttr objects.
     * @return array Associative array of attributes.
     */
    protected function transformAttrToAssoc($node_map)
    {
        // NamedNodeMap is documented very well, so we're using undocumented
        // features, namely, the fact that it implements Iterator and
        // has a ->length attribute
        if ($node_map->length === 0) {
            return array();
        }
        $array = array();
        foreach ($node_map as $attr) {
            $array[$attr->name] = $attr->value;
        }
        return $array;
    }

    /**
     * An error handler that mutes all errors
     * @param int $errno
     * @param string $errstr
     */
    public function muteErrorHandler($errno, $errstr)
    {
    }

    /**
     * Callback function for undoing escaping of stray angled brackets
     * in comments
     * @param array $matches
     * @return string
     */
    public function callbackUndoCommentSubst($matches)
    {
        return '<!--' . strtr($matches[1], array('&amp;' => '&', '&lt;' => '<')) . $matches[2];
    }

    /**
     * Callback function that entity-izes ampersands in comments so that
     * callbackUndoCommentSubst doesn't clobber them
     * @param array $matches
     * @return string
     */
    public function callbackArmorCommentEntities($matches)
    {
        return '<!--' . str_replace('&', '&amp;', $matches[1]) . $matches[2];
    }

    /**
     * Wraps an HTML fragment in the necessary HTML
     * @param string $html
     * @param HTMLPurifier_Config $config
     * @param HTMLPurifier_Context $context
     * @return string
     */
    protected function wrapHTML($html, $config, $context)
    {
        $def = $config->getDefinition('HTML');
        $ret = '';

        if (!empty($def->doctype->dtdPublic) || !empty($def->doctype->dtdSystem)) {
            $ret .= '<!DOCTYPE html ';
            if (!empty($def->doctype->dtdPublic)) {
                $ret .= 'PUBLIC "' . $def->doctype->dtdPublic . '" ';
            }
            if (!empty($def->doctype->dtdSystem)) {
                $ret .= '"' . $def->doctype->dtdSystem . '" ';
            }
            $ret .= '>';
        }

        $ret .= '<html><head>';
        $ret .= '<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />';
        // No protection if $html contains a stray </div>!
        $ret .= '</head><body><div>' . $html . '</div></body></html>';
        return $ret;
    }
}

// vim: et sw=4 sts=4