Lexer = new HTMLPurifier_Lexer();
$this->DirectLex = new HTMLPurifier_Lexer_DirectLex();
if ( $GLOBALS['HTMLPurifierTest']['PEAR'] ) {
$this->_has_pear = true;
require_once 'HTMLPurifier/Lexer/PEARSax3.php';
$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();
}
$this->_entity_lookup = HTMLPurifier_EntityLookup::instance();
}
function assertCleanUTF8($string, $expect = null) {
if ($expect === null) $expect = $string;
$this->assertIdentical($this->Lexer->cleanUTF8($string), $expect);
}
function test_cleanUTF8() {
$this->assertCleanUTF8('Normal string.');
$this->assertCleanUTF8("Test\tAllowed\nControl\rCharacters");
$this->assertCleanUTF8("null byte: \0", 'null byte: ');
$this->assertCleanUTF8("\1\2\3\4\5\6\7", '');
$this->assertCleanUTF8("\x7F", ''); // one byte invalid SGML char
$this->assertCleanUTF8("\xC2\x80", ''); // two byte invalid SGML
$this->assertCleanUTF8("\xF3\xBF\xBF\xBF"); // valid four byte
$this->assertCleanUTF8("\xDF\xFF", ''); // malformed UTF8
}
function test_substituteNonSpecialEntities() {
$char_theta = $this->_entity_lookup->table['theta'];
$this->assertIdentical($char_theta,
$this->Lexer->substituteNonSpecialEntities('θ') );
$this->assertIdentical('"',
$this->Lexer->substituteNonSpecialEntities('"') );
// numeric tests, adapted from Feyd
$args = array();
$args[] = array(1114112,false );
$args[] = array(1114111,'F48FBFBF'); // 0x0010FFFF
$args[] = array(1048576,'F4808080'); // 0x00100000
$args[] = array(1048575,'F3BFBFBF'); // 0x000FFFFF
$args[] = array(262144, 'F1808080'); // 0x00040000
$args[] = array(262143, 'F0BFBFBF'); // 0x0003FFFF
$args[] = array(65536, 'F0908080'); // 0x00010000
$args[] = array(65535, 'EFBFBF' ); // 0x0000FFFF
$args[] = array(57344, 'EE8080' ); // 0x0000E000
$args[] = array(57343, false ); // 0x0000DFFF these are ill-formed
$args[] = array(56040, false ); // 0x0000DAE8 these are ill-formed
$args[] = array(55296, false ); // 0x0000D800 these are ill-formed
$args[] = array(55295, 'ED9FBF' ); // 0x0000D7FF
$args[] = array(53248, 'ED8080' ); // 0x0000D000
$args[] = array(53247, 'ECBFBF' ); // 0x0000CFFF
$args[] = array(4096, 'E18080' ); // 0x00001000
$args[] = array(4095, 'E0BFBF' ); // 0x00000FFF
$args[] = array(2048, 'E0A080' ); // 0x00000800
$args[] = array(2047, 'DFBF' ); // 0x000007FF
$args[] = array(128, 'C280' ); // 0x00000080 invalid SGML char
$args[] = array(127, '7F' ); // 0x0000007F invalid SGML char
$args[] = array(0, '00' ); // 0x00000000 invalid SGML char
$args[] = array(20108, 'E4BA8C' ); // 0x00004E8C
$args[] = array(77, '4D' ); // 0x0000004D
$args[] = array(66306, 'F0908C82'); // 0x00010302
$args[] = array(1072, 'D0B0' ); // 0x00000430
foreach ($args as $arg) {
$string = '' . $arg[0] . ';' . // decimal
'' . dechex($arg[0]) . ';'; // hex
$expect = '';
if ($arg[1] !== false) {
$chars = str_split($arg[1], 2);
foreach ($chars as $char) {
$expect .= chr(hexdec($char));
}
$expect .= $expect; // double it
}
$this->assertIdentical(
$this->Lexer->substituteNonSpecialEntities($string),
$expect,
$arg[0] . ': %s'
);
}
}
function assertExtractBody($text, $extract = true) {
$result = $this->Lexer->extractBody($text);
if ($extract === true) $extract = $text;
$this->assertIdentical($extract, $result);
}
function test_extractBody() {
$this->assertExtractBody('Bold');
$this->assertExtractBody('
Bold', 'Bold');
$this->assertExtractBody('Bold', 'Bold');
$this->assertExtractBody(
'
xyz
',
'
');
$this->assertExtractBody('Bold', 'Bold');
$this->assertExtractBody('asdf'); // not closed, don't accept
}
function test_tokenizeHTML() {
$input = array();
$expect = array();
$sax_expect = array();
$config = 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!
// CDATA sections!
$input[13] = 'can't get me!]]>';
$expect[13] = array( new HTMLPurifier_Token_Text(
'You can't get me!' // raw
) );
$sax_expect[13] = array( // SAX has a seperate call for each entity
new HTMLPurifier_Token_Text('You '),
new HTMLPurifier_Token_Text('<'),
new HTMLPurifier_Token_Text('b'),
new HTMLPurifier_Token_Text('>'),
new HTMLPurifier_Token_Text('can'),
new HTMLPurifier_Token_Text('&'),
new HTMLPurifier_Token_Text('#39;t'),
new HTMLPurifier_Token_Text('<'),
new HTMLPurifier_Token_Text('/b'),
new HTMLPurifier_Token_Text('>'),
new HTMLPurifier_Token_Text(' get me!')
);
$char_theta = $this->_entity_lookup->table['theta'];
$char_rarr = $this->_entity_lookup->table['rarr'];
// test entity replacement
$input[14] = 'θ';
$expect[14] = array( new HTMLPurifier_Token_Text($char_theta) );
// test that entities aren't replaced in CDATA sections
$input[15] = 'θ ';
$expect[15] = array( new HTMLPurifier_Token_Text($char_theta . ' →') );
$sax_expect[15] = array(
new HTMLPurifier_Token_Text($char_theta . ' '),
new HTMLPurifier_Token_Text('&'),
new HTMLPurifier_Token_Text('rarr;')
);
// test entity resolution in attributes
$input[16] = 'Link';
$expect[16] = array(
new HTMLPurifier_Token_Start('a',array('href' => 'index.php?title=foo&id=bar'))
,new HTMLPurifier_Token_Text('Link')
,new HTMLPurifier_Token_End('a')
);
$sax_expect[16] = false; // PEARSax doesn't support it!
// test that UTF-8 is preserved
$char_hearts = $this->_entity_lookup->table['hearts'];
$input[17] = $char_hearts;
$expect[17] = array( new HTMLPurifier_Token_Text($char_hearts) );
$default_config = HTMLPurifier_Config::createDefault();
foreach($input as $i => $discard) {
if (!isset($config[$i])) $config[$i] = $default_config;
$result = $this->DirectLex->tokenizeHTML($input[$i], $config[$i]);
$this->assertEqual($expect[$i], $result, 'DirectLexTest '.$i.': %s');
paintIf($result, $expect[$i] != $result);
if ($this->_has_pear) {
// assert unless I say otherwise
$sax_result = $this->PEARSax3->tokenizeHTML($input[$i], $config[$i]);
if (!isset($sax_expect[$i])) {
// by default, assert with normal result
$this->assertEqual($expect[$i], $sax_result, 'PEARSax3Test '.$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], $sax_result, 'PEARSax3Test (custom) '.$i.': %s');
paintIf($sax_result, $sax_expect[$i] != $sax_result);
}
}
if ($this->_has_dom) {
$dom_result = $this->DOMLex->tokenizeHTML($input[$i], $config[$i]);
// same structure as SAX
if (!isset($dom_expect[$i])) {
$this->assertEqual($expect[$i], $dom_result, 'DOMLexTest '.$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], $dom_result, 'DOMLexTest (custom) '.$i.': %s');
paintIf($dom_result, $dom_expect[$i] != $dom_result);
}
}
}
}
}
?>