Lexer = new HTMLPurifier_Lexer();
$this->DirectLex = new HTMLPurifier_Lexer_DirectLex();
// E_STRICT = 2048, int used for PHP4 compat
if ( $GLOBALS['HTMLPurifierTest']['PEAR'] &&
((error_reporting() & 2048) != 2048)
) {
$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 test_create() {
$config = HTMLPurifier_Config::create(array('Core.MaintainLineNumbers' => true));
$lexer = HTMLPurifier_Lexer::create($config);
$this->assertIsA($lexer, 'HTMLPurifier_Lexer_DirectLex');
}
function assertExtractBody($text, $extract = true) {
$result = $this->Lexer->extractBody($text);
if ($extract === true) $extract = $text;
$this->assertIdentical($extract, $result);
}
function test_parseData() {
$HP =& $this->Lexer;
$this->assertIdentical('asdf', $HP->parseData('asdf'));
$this->assertIdentical('&', $HP->parseData('&'));
$this->assertIdentical('"', $HP->parseData('"'));
$this->assertIdentical("'", $HP->parseData('''));
$this->assertIdentical("'", $HP->parseData('''));
$this->assertIdentical('&&&', $HP->parseData('&&&'));
$this->assertIdentical('&&', $HP->parseData('&&')); // [INVALID]
$this->assertIdentical('Procter & Gamble',
$HP->parseData('Procter & Gamble')); // [INVALID]
// This is not special, thus not converted. Test of fault tolerance,
// realistically speaking, this should never happen
$this->assertIdentical('-', $HP->parseData('-'));
}
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')
);
// 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) );
// test weird characters in attributes
$input[18] = '
';
$expect[18] = array( new HTMLPurifier_Token_Empty('br', array('test' => 'x < 6')) );
// test emoticon protection
$input[19] = 'Whoa! >.< That\'s not good >.>';
$expect[19] = array(
new HTMLPurifier_Token_Start('b'),
new HTMLPurifier_Token_Text('Whoa! >.'),
new HTMLPurifier_Token_Text('< That\'s not good >'),
new HTMLPurifier_Token_Text('.>'),
new HTMLPurifier_Token_End('b'),
);
$sax_expect[19] = false; // SAX drops the < character
$dom_expect[19] = false; // DOM drops the entire pseudo-tag
// test comment parsing with funky characters inside
$input[20] = '
';
$expect[20] = array(
new HTMLPurifier_Token_Comment(' This >< comment '),
new HTMLPurifier_Token_Empty('br')
);
$sax_expect[20] = false;
// test comment parsing of missing end
$input[21] = '