DirectLex = new HTMLPurifier_Lexer_DirectLex();
$this->PEARSax3 = new HTMLPurifier_Lexer_PEARSax3();
$this->_has_dom = version_compare(PHP_VERSION, '5', '>=');
if ($this->_has_dom) {
require_once 'HTMLPurifier/Lexer/DOMLex.php';
$this->DOMLex = new HTMLPurifier_Lexer_DOMLex();
}
$this->_entity_lookup = HTMLPurifier_EntityLookup::instance();
}
function test_tokenizeHTML() {
$input = array();
$expect = array();
$sax_expect = array();
$input[0] = '';
$expect[0] = array();
$input[1] = 'This is regular text.';
$expect[1] = array(
new HTMLPurifier_Token_Text('This is regular text.')
);
$input[2] = 'This is bold text';
$expect[2] = array(
new HTMLPurifier_Token_Text('This is ')
,new HTMLPurifier_Token_Start('b', array())
,new HTMLPurifier_Token_Text('bold')
,new HTMLPurifier_Token_End('b')
,new HTMLPurifier_Token_Text(' text')
);
$input[3] = 'Totally rad dude. asdf
';
$expect[3] = array(
new HTMLPurifier_Token_Start('DIV', array())
,new HTMLPurifier_Token_Text('Totally rad dude. ')
,new HTMLPurifier_Token_Start('b', array())
,new HTMLPurifier_Token_Text('asdf')
,new HTMLPurifier_Token_End('b')
,new HTMLPurifier_Token_End('div')
);
// [XML-INVALID]
$input[4] = '';
$expect[4] = array(
new HTMLPurifier_Token_Start('asdf')
,new HTMLPurifier_Token_End('asdf')
,new HTMLPurifier_Token_Start('d')
,new HTMLPurifier_Token_End('d')
,new HTMLPurifier_Token_Start('poOloka')
,new HTMLPurifier_Token_Start('poolasdf')
,new HTMLPurifier_Token_Start('ds')
,new HTMLPurifier_Token_End('asdf')
,new HTMLPurifier_Token_End('ASDF')
);
// DOM is different because it condenses empty tags into REAL empty ones
// as well as makes it well-formed
$dom_expect[4] = array(
new HTMLPurifier_Token_Empty('asdf')
,new HTMLPurifier_Token_Empty('d')
,new HTMLPurifier_Token_Start('pooloka')
,new HTMLPurifier_Token_Start('poolasdf')
,new HTMLPurifier_Token_Empty('ds')
,new HTMLPurifier_Token_End('poolasdf')
,new HTMLPurifier_Token_End('pooloka')
);
$input[5] = 'Link to foobar';
$expect[5] = array(
new HTMLPurifier_Token_Start('a',array('href'=>'foobar.php','title'=>'foo!'))
,new HTMLPurifier_Token_Text('Link to ')
,new HTMLPurifier_Token_Start('b',array('id'=>'asdf'))
,new HTMLPurifier_Token_Text('foobar')
,new HTMLPurifier_Token_End('b')
,new HTMLPurifier_Token_End('a')
);
$input[6] = '
';
$expect[6] = array(
new HTMLPurifier_Token_Empty('br')
);
// [SGML-INVALID] [RECOVERABLE]
$input[7] = ' ';
$expect[7] = array(
new HTMLPurifier_Token_Comment(' Comment ')
,new HTMLPurifier_Token_Text(' ')
,new HTMLPurifier_Token_Comment(' not so well formed -')
);
$sax_expect[7] = false; // we need to figure out proper comment output
// [SGML-INVALID]
$input[8] = '''))
);
// DOM parses it into an empty tag
$dom_expect[8] = array(
new HTMLPurifier_Token_Empty('a', array('href'=>''))
);
$input[9] = '<b>';
$expect[9] = array(
new HTMLPurifier_Token_Text('')
);
$sax_expect[9] = array(
new HTMLPurifier_Token_Text('<')
,new HTMLPurifier_Token_Text('b')
,new HTMLPurifier_Token_Text('>')
);
// note that SAX can clump text nodes together. We won't be
// too picky though
// [SGML-INVALID]
$input[10] = '';
// We barf on this, aim for no attributes
$expect[10] = array(
new HTMLPurifier_Token_Start('a', array('"' => ''))
);
// DOM correctly has no attributes, but also closes the tag
$dom_expect[10] = array(
new HTMLPurifier_Token_Empty('a')
);
// SAX barfs on this
$sax_expect[10] = array(
new HTMLPurifier_Token_Start('a', array('"' => ''))
);
// [INVALID] [RECOVERABLE]
$input[11] = '"';
$expect[11] = array( new HTMLPurifier_Token_Text('"') );
// compare with this valid one:
$input[12] = '"';
$expect[12] = array( new HTMLPurifier_Token_Text('"') );
$sax_expect[12] = false; // choked!
// DOM and SAX choke on this
//$char_circ = $this->_entity_lookup->table['circ'];
//$input[13] = 'ˆ';
//$expect[13] = array( new HTMLPurifier_Token_Text($char_circ) );
foreach($input as $i => $discard) {
$result = $this->DirectLex->tokenizeHTML($input[$i]);
$this->assertEqual($expect[$i], $result, 'Test '.$i.': %s');
paintIf($result, $expect[$i] != $result);
// assert unless I say otherwise
$sax_result = $this->PEARSax3->tokenizeHTML($input[$i]);
if (!isset($sax_expect[$i])) {
// by default, assert with normal result
$this->assertEqual($expect[$i], $sax_result, 'Test '.$i.': %s');
paintIf($sax_result, $expect[$i] != $sax_result);
} elseif ($sax_expect[$i] === false) {
// assertions were turned off, optionally dump
// paintIf($sax_expect, $i == NUMBER);
} else {
// match with a custom SAX result array
$this->assertEqual($sax_expect[$i], $sax_result, 'Test '.$i.': %s');
paintIf($sax_result, $sax_expect[$i] != $sax_result);
}
if ($this->_has_dom) {
$dom_result = $this->DOMLex->tokenizeHTML($input[$i]);
// same structure as SAX
if (!isset($dom_expect[$i])) {
$this->assertEqual($expect[$i], $dom_result, 'Test '.$i.': %s');
paintIf($dom_result, $expect[$i] != $dom_result);
} elseif ($dom_expect[$i] === false) {
// paintIf($dom_result, $i == NUMBER);
} else {
$this->assertEqual($dom_expect[$i], $dom_result, 'Test '.$i.': %s');
paintIf($dom_result, $dom_expect[$i] != $dom_result);
}
}
}
}
}
?>