Browse Source
• Document::__construct now accepts \DOMDocument|string|null as its source. • Document::serialize has been removed in favor of Document::saveHTML, making it more in line with PHP's DOM. • Added Document::loadDOM for loading a vanilla PHP \DOMDocument into this library. As mentioned above Document::__constructor also accepts a \DOMDocument. • Document::importNode will correctly convert an imported template element as a HTMLTemplateElement. • TemplateElement is now HTMLTemplateElement. • DOM walker methods now correctly check if the output of the closure is true • Unnecessary test cases removed.element-classes
Dustin Wilson
3 years ago
16 changed files with 273 additions and 992 deletions
@ -1,107 +0,0 @@ |
|||
<?php |
|||
/** @license MIT |
|||
* Copyright 2017 , Dustin Wilson, J. King et al. |
|||
* See LICENSE and AUTHORS files for details */ |
|||
|
|||
declare(strict_types=1); |
|||
namespace MensBeam\HTML\DOM\TestCase; |
|||
|
|||
use MensBeam\HTML\DOM\Charset; |
|||
|
|||
/** |
|||
* @covers \MensBeam\HTML\DOM\Charset |
|||
*/ |
|||
class TestCharset extends \PHPUnit\Framework\TestCase { |
|||
/** @dataProvider provideCharsets */ |
|||
public function testDetermineEncodingFromEncodingLabel(string $in, ?string $exp) { |
|||
$this->assertSame($exp, Charset::fromCharset($in)); |
|||
} |
|||
|
|||
public function provideCharsets() { |
|||
return [ |
|||
["UTF-8", "UTF-8"], |
|||
[" utf8 ", "UTF-8"], |
|||
["ISO-8859-1", "windows-1252"], |
|||
["text/html; charset=utf8", null], |
|||
]; |
|||
} |
|||
|
|||
/** @dataProvider provideContentTypes */ |
|||
public function testDetermineEncodingFromContentType(string $input, ?string $exp) { |
|||
$this->assertSame($exp, Charset::fromTransport($input)); |
|||
} |
|||
|
|||
public function provideContentTypes() { |
|||
return [ |
|||
["UTF-8", null], |
|||
["charset=utf8", null], |
|||
["text/html", null], |
|||
["text/html charset=utf8", null], |
|||
["text/html; charset=utf8", "UTF-8"], |
|||
["text/html;charset=utf8", "UTF-8"], |
|||
["text/html; charset=\"utf8\"", "UTF-8"], |
|||
["image/svg+xml; param=value; charset=utf8", "UTF-8"], |
|||
["image/svg+xml; charset=utf8; charset=big5", "UTF-8"], |
|||
["image/svg+xml; charset=utf8;charset=big5", "UTF-8"], |
|||
["text/html; charset=not-valid; charset=big5", null], |
|||
["text/html; charset=not-valid", null], |
|||
["text/html; charsaaet=\"a \\\"fancy\\\" encoding\"", null], |
|||
]; |
|||
} |
|||
|
|||
/** @dataProvider provideBOMs */ |
|||
public function testDetermineEncodingFromByteOrderMark(string $input, ?string $exp) { |
|||
$this->assertSame($exp, Charset::fromBOM($input)); |
|||
} |
|||
|
|||
public function provideBOMs() { |
|||
return [ |
|||
'UTF-8' => ["\u{FEFF}Hello world!", "UTF-8"], |
|||
'UTF-16 (big-endian)' => ["\xFE\xFF\0H\0e\0l\0l\0o\0 \0w\0o\0r\0l\0d\0!", "UTF-16BE"], |
|||
'UTF-16 (little-endian)' => ["\xFF\xFEH\0e\0l\0l\0o\0 \0w\0o\0r\0l\0d\0!\0", "UTF-16LE"], |
|||
'No byte order mark' => ["Hello world!", null], |
|||
]; |
|||
} |
|||
|
|||
/** @dataProvider provideStandardEncodingTests */ |
|||
public function testStandardEncoderTests(string $input, string $exp) { |
|||
$exp = strtolower($exp); |
|||
$this->assertSame(strtolower($exp), strtolower(Charset::fromBOM($input)?? Charset::fromPrescan($input, \PHP_INT_MAX) ?? "Windows-1252")); |
|||
} |
|||
|
|||
public function provideStandardEncodingTests() { |
|||
$tests = []; |
|||
$blacklist = []; |
|||
$files = new \AppendIterator(); |
|||
$files->append(new \GlobIterator(\MensBeam\HTML\DOM\BASE."tests/html5lib-tests/encoding/*.dat", \FilesystemIterator::SKIP_DOTS | \FilesystemIterator::CURRENT_AS_PATHNAME)); |
|||
$files->append(new \GlobIterator(\MensBeam\HTML\DOM\BASE."tests/cases/encoding/*.dat", \FilesystemIterator::SKIP_DOTS | \FilesystemIterator::CURRENT_AS_PATHNAME)); |
|||
foreach ($files as $file) { |
|||
if (!in_array(basename($file), $blacklist)) { |
|||
$tests[] = $file; |
|||
} |
|||
} |
|||
return $this->makeEncodingTests(...$tests); |
|||
} |
|||
|
|||
protected function makeEncodingTests(string ...$file): iterable { |
|||
foreach ($file as $path) { |
|||
$f = basename($path); |
|||
$test = file($path); |
|||
$end = sizeof($test); |
|||
$l = 0; |
|||
$index = 0; |
|||
while ($l < $end) { |
|||
$testId = "$f #".$index++; |
|||
$data = ""; |
|||
while ($l < $end && !preg_match("/^#data\s+$/", @$test[$l++])); |
|||
while ($l < $end && !preg_match("/^#encoding\s+$/", ($line = @$test[$l++]))) { |
|||
$data .= $line; |
|||
} |
|||
if ($l >= $end) { |
|||
return; |
|||
} |
|||
yield $testId => [trim($data, "\r\n"), trim($test[$l++])]; |
|||
} |
|||
} |
|||
} |
|||
} |
@ -1,235 +0,0 @@ |
|||
<?php |
|||
/** @license MIT |
|||
* Copyright 2017 , Dustin Wilson, J. King et al. |
|||
* See LICENSE and AUTHORS files for details */ |
|||
|
|||
declare(strict_types=1); |
|||
namespace MensBeam\HTML\DOM\TestCase; |
|||
|
|||
use MensBeam\HTML\DOM\Data; |
|||
use MensBeam\HTML\DOM\EOFToken; |
|||
use MensBeam\HTML\DOM\OpenElementsStack; |
|||
use MensBeam\HTML\DOM\ParseError; |
|||
use MensBeam\HTML\DOM\Tokenizer; |
|||
use MensBeam\HTML\DOM\CharacterToken; |
|||
use MensBeam\HTML\DOM\CommentToken; |
|||
use MensBeam\HTML\DOM\DOCTYPEToken; |
|||
use MensBeam\HTML\DOM\EndTagToken; |
|||
use MensBeam\HTML\DOM\NullCharacterToken; |
|||
use MensBeam\HTML\DOM\StartTagToken; |
|||
use MensBeam\HTML\DOM\TokenAttr; |
|||
use MensBeam\HTML\DOM\WhitespaceToken; |
|||
|
|||
/** |
|||
* @covers \MensBeam\HTML\DOM\Data |
|||
* @covers \MensBeam\HTML\DOM\Tokenizer |
|||
* @covers \MensBeam\HTML\DOM\CharacterToken |
|||
* @covers \MensBeam\HTML\DOM\CommentToken |
|||
* @covers \MensBeam\HTML\DOM\DataToken |
|||
* @covers \MensBeam\HTML\DOM\TagToken |
|||
* @covers \MensBeam\HTML\DOM\DOCTYPEToken |
|||
* @covers \MensBeam\HTML\DOM\TokenAttr |
|||
*/ |
|||
class TestTokenizer extends \PHPUnit\Framework\TestCase { |
|||
const STATE_MAP = [ |
|||
'Data state' => Tokenizer::DATA_STATE, |
|||
'PLAINTEXT state' => Tokenizer::PLAINTEXT_STATE, |
|||
'RCDATA state' => Tokenizer::RCDATA_STATE, |
|||
'RAWTEXT state' => Tokenizer::RAWTEXT_STATE, |
|||
'Script data state' => Tokenizer::SCRIPT_DATA_STATE, |
|||
'CDATA section state' => Tokenizer::CDATA_SECTION_STATE, |
|||
]; |
|||
|
|||
/** @dataProvider provideStandardTokenizerTests */ |
|||
public function testStandardTokenizerTests(string $input, array $expected, int $state, string $open = null, array $expErrors) { |
|||
// convert parse error constants into standard symbols in specification |
|||
$errorMap = array_map(function($str) { |
|||
return strtolower(str_replace("_", "-", $str)); |
|||
}, array_flip(array_filter((new \ReflectionClass(ParseError::class))->getConstants(), function($v) { |
|||
return is_int($v); |
|||
}))); |
|||
// create a stub error handler which collects parse errors |
|||
$errors = []; |
|||
$errorHandler = $this->createStub(ParseError::class); |
|||
$errorHandler->method("emit")->willReturnCallback(function($file, $line, $col, $code) use (&$errors, $errorMap) { |
|||
$errors[] = ['code' => $errorMap[$code], 'line' => $line, 'col' => $col]; |
|||
return true; |
|||
}); |
|||
// initialize a stack of open elements, possibly with an open element |
|||
$stack = new OpenElementsStack(); |
|||
if ($open) { |
|||
$stack[] = (new \DOMDocument)->createElement($open); |
|||
} |
|||
// initialize the data stream and tokenizer |
|||
$data = new Data($input, "STDIN", $errorHandler, "UTF-8"); |
|||
$tokenizer = new Tokenizer($data, $stack, $errorHandler); |
|||
$tokenizer->state = $state; |
|||
// perform the test |
|||
$actual = []; |
|||
try { |
|||
foreach ($tokenizer->tokenize() as $t) { |
|||
assert( |
|||
(!$t instanceof CharacterToken) |
|||
|| ($t instanceof NullCharacterToken && $t->data === "\0") |
|||
|| ($t instanceof WhitespaceToken && strspn($t->data, Data::WHITESPACE) === strlen($t->data)) |
|||
|| ($t->data !== "\0" && strspn($t->data, Data::WHITESPACE) === 0) |
|||
, new \Exception("Character token must either consist of a single null character, consist only of whitespace, or start with other than whitespace: ".get_class($t)." ".var_export($t->data ?? "''", true))); |
|||
$actual[] = $t; |
|||
} |
|||
} finally { |
|||
$actual = $this->combineCharacterTokens($actual); |
|||
$this->assertEquals($expected, $actual, $tokenizer->debugLog); |
|||
$this->assertEquals($expErrors, $errors, $tokenizer->debugLog); |
|||
} |
|||
} |
|||
|
|||
public function provideStandardTokenizerTests() { |
|||
$tests = []; |
|||
$blacklist = ["xmlViolation.test"]; |
|||
$files = new \AppendIterator(); |
|||
$files->append(new \GlobIterator(\MensBeam\HTML\DOM\BASE."tests/html5lib-tests/tokenizer/*.test", \FilesystemIterator::SKIP_DOTS | \FilesystemIterator::CURRENT_AS_PATHNAME)); |
|||
$files->append(new \GlobIterator(\MensBeam\HTML\DOM\BASE."tests/cases/tokenizer/*.test", \FilesystemIterator::SKIP_DOTS | \FilesystemIterator::CURRENT_AS_PATHNAME)); |
|||
foreach ($files as $file) { |
|||
if (!in_array(basename($file), $blacklist)) { |
|||
$tests[] = $file; |
|||
} |
|||
} |
|||
return $this->makeTokenTests(...$tests); |
|||
} |
|||
|
|||
protected function reverseDoubleEscape(string $str): string { |
|||
if (preg_match_all("/\\\\u([0-9a-f]{4})/i", $str, $matches)) { |
|||
for ($a = 0; $a < sizeof($matches[0]); $a++) { |
|||
$esc = $matches[0][$a]; |
|||
$chr = \MensBeam\Intl\Encoding\UTF8::encode(hexdec($matches[1][$a])); |
|||
$str = str_replace($esc, $chr, $str); |
|||
} |
|||
} |
|||
return $str; |
|||
} |
|||
|
|||
protected function combineCharacterTokens(array $tokens) : array { |
|||
$out = []; |
|||
$pending = null; |
|||
foreach ($tokens as $t) { |
|||
if ($t instanceof CharacterToken) { |
|||
if (!$pending) { |
|||
if ($t instanceof WhitespaceToken || $t instanceof NullCharacterToken) { |
|||
$t = new CharacterToken($t->data); |
|||
} |
|||
$pending = $t; |
|||
} else { |
|||
$pending->data .= $t->data; |
|||
} |
|||
} else { |
|||
if ($pending) { |
|||
$out[] = $pending; |
|||
$pending = null; |
|||
} |
|||
$out[] = $t; |
|||
} |
|||
} |
|||
if ($pending) { |
|||
$out[] = $pending; |
|||
} |
|||
return $out; |
|||
} |
|||
|
|||
protected function makeTokenTests(string ...$file): iterable { |
|||
foreach ($file as $path) { |
|||
$f = basename($path); |
|||
$testSet = json_decode(file_get_contents($path), true); |
|||
foreach ($testSet['tests'] ?? $testSet['xmlViolationTests'] as $index => $test) { |
|||
$testId = "$f #$index"; |
|||
if ($test['doubleEscaped'] ?? false) { |
|||
$test['input'] = $this->reverseDoubleEscape($test['input']); |
|||
for ($a = 0; $a < sizeof($test['output']); $a++) { |
|||
for ($b = 0; $b < sizeof($test['output'][$a]); $b++) { |
|||
if (is_string($test['output'][$a][$b])) { |
|||
$test['output'][$a][$b] = $this->reverseDoubleEscape($test['output'][$a][$b]); |
|||
} |
|||
} |
|||
} |
|||
} |
|||
$test['initialStates'] = $test['initialStates'] ?? ["Data state"]; |
|||
// check if a test needs a patch due to trivial differences in implementation |
|||
$this->patchTest($test); |
|||
for ($a = 0; $a < sizeof($test['initialStates']); $a++) { |
|||
$tokens = []; |
|||
foreach ($test['output'] as $token) { |
|||
switch ($token[0]) { |
|||
case "DOCTYPE": |
|||
$t = new DOCTYPEToken((string) $token[1], (string) $token[2], (string) $token[3]); |
|||
$t->forceQuirks = !$token[4]; |
|||
$tokens[] = $t; |
|||
break; |
|||
case "StartTag": |
|||
$t = new StartTagToken($token[1], $token[3] ?? false); |
|||
foreach ($token[2] ?? [] as $name => $value) { |
|||
$t->attributes[] = new TokenAttr((string) $name, $value); |
|||
} |
|||
$tokens[] = $t; |
|||
break; |
|||
case "EndTag": |
|||
$tokens[] = new EndTagToken($token[1]); |
|||
break; |
|||
case "Character": |
|||
$tokens[] = new CharacterToken($token[1]); |
|||
break; |
|||
case "Comment": |
|||
$tokens[] = new CommentToken($token[1]); |
|||
break; |
|||
default: |
|||
throw new \Exception("Token type '{$token[0]}' not implemented in standard test interpreter"); |
|||
} |
|||
unset($t); |
|||
} |
|||
$tokens[] = new EOFToken; |
|||
yield "$testId: {$test['description']} ({$test['initialStates'][$a]})" => [ |
|||
$test['input'], // input |
|||
$tokens, // output |
|||
self::STATE_MAP[$test['initialStates'][$a]], // initial state |
|||
$test['lastStartTag'] ?? null, // open element, if any |
|||
$test['errors'] ?? [], // errors, if any |
|||
]; |
|||
} |
|||
} |
|||
} |
|||
} |
|||
|
|||
protected function patchTest(&$test): void { |
|||
$id = [$test['input'], $test['initialStates']]; |
|||
switch ($id) { |
|||
// test emits input stream error first despite peeking |
|||
case ["<!\u{B}", ["Data state"]]: |
|||
$test['errors'] = array_reverse($test['errors']); |
|||
break; |
|||
// eof-in-<whatever> positions in some tests don't make sense |
|||
// https://github.com/html5lib/html5lib-tests/issues/125 |
|||
case ["", ["CDATA section state"]]: |
|||
// there is no position 2 |
|||
$test['errors'][0]['col']--; |
|||
break; |
|||
case ["\u{A}", ["CDATA section state"]]: |
|||
// the line break is, for some reason, not counted in the test |
|||
$test['errors'][0]['line']++; |
|||
$test['errors'][0]['col'] = 1; |
|||
break; |
|||
case ["<!----!\r\n>", ["Data state"]]: |
|||
case ["<!----!\n>", ["Data state"]]: |
|||
case ["<!----!\r>", ["Data state"]]: |
|||
// the line break is, for some reason, not counted in the test |
|||
$test['errors'][0]['line']++; |
|||
$test['errors'][0]['col'] = 2; |
|||
break; |
|||
case ["<!----! >", ["Data state"]]: |
|||
$test['errors'][0]['col']++; |
|||
break; |
|||
case [hex2bin("f4808080"), ["CDATA section state"]]: |
|||
case [hex2bin("3bf4808080"), ["CDATA section state"]]: |
|||
// malpaired surrogates count as two characters |
|||
$test['errors'][0]['col']++; |
|||
break; |
|||
} |
|||
} |
|||
} |
@ -1,455 +0,0 @@ |
|||
<?php |
|||
/** @license MIT |
|||
* Copyright 2017 , Dustin Wilson, J. King et al. |
|||
* See LICENSE and AUTHORS files for details */ |
|||
|
|||
declare(strict_types=1); |
|||
namespace MensBeam\HTML\DOM\TestCase; |
|||
|
|||
use MensBeam\HTML\DOM\Data; |
|||
use MensBeam\HTML\DOM\LoopException; |
|||
use MensBeam\HTML\DOM\NotImplementedException; |
|||
use MensBeam\HTML\DOM\OpenElementsStack; |
|||
use MensBeam\HTML\DOM\ParseError; |
|||
use MensBeam\HTML\DOM\Parser; |
|||
use MensBeam\HTML\DOM\TemplateInsertionModesStack; |
|||
use MensBeam\HTML\DOM\Tokenizer; |
|||
use MensBeam\HTML\DOM\TreeBuilder; |
|||
|
|||
/** |
|||
* @covers \MensBeam\HTML\DOM\Document |
|||
* @covers \MensBeam\HTML\DOM\Element |
|||
* @covers \MensBeam\HTML\DOM\Tokenizer |
|||
* @covers \MensBeam\HTML\DOM\TreeBuilder |
|||
* @covers \MensBeam\HTML\DOM\ActiveFormattingElementsList |
|||
* @covers \MensBeam\HTML\DOM\TemplateInsertionModesStack |
|||
* @covers \MensBeam\HTML\DOM\OpenElementsStack |
|||
* @covers \MensBeam\HTML\DOM\Stack |
|||
* @covers \MensBeam\HTML\DOM\TagToken |
|||
*/ |
|||
class TestTreeConstructor extends \PHPUnit\Framework\TestCase { |
|||
use \MensBeam\HTML\DOM\EscapeString; |
|||
|
|||
protected $out; |
|||
protected $depth; |
|||
|
|||
/** @dataProvider provideStandardTreeTests */ |
|||
public function testStandardTreeTests(string $data, array $exp, array $errors, $fragment): void { |
|||
// certain tests need to be patched to ignore unavoidable limitations of PHP's DOM |
|||
[$exp, $errors, $patched, $skip] = $this->patchTest($data, $fragment, $errors, $exp); |
|||
if (strlen($skip)) { |
|||
$this->markTestSkipped($skip); |
|||
} elseif ($patched) { |
|||
$this->markAsRisky(); |
|||
} |
|||
// convert parse error constants into standard symbols in specification |
|||
$errorMap = array_map(function($str) { |
|||
return strtolower(str_replace("_", "-", $str)); |
|||
}, array_flip(array_filter((new \ReflectionClass(ParseError::class))->getConstants(), function($v) { |
|||
return is_int($v); |
|||
}))); |
|||
// create a stub error handler which collects parse errors |
|||
$actualErrors = []; |
|||
$errorHandler = $this->createStub(ParseError::class); |
|||
$errorHandler->method("emit")->willReturnCallback(function($file, $line, $col, $code) use (&$actualErrors, $errorMap) { |
|||
$actualErrors[] = ['code' => $errorMap[$code], 'line' => $line, 'col' => $col]; |
|||
return true; |
|||
}); |
|||
// initialize the output document |
|||
$doc = new \DOMDocument; |
|||
// prepare the fragment context, if any |
|||
if ($fragment) { |
|||
$fragment = explode(" ", $fragment); |
|||
assert(sizeof($fragment) < 3); |
|||
if (sizeof($fragment) === 1) { |
|||
$fragmentContext = $doc->createElement($fragment[0]); |
|||
} else { |
|||
$ns = array_flip(Parser::NAMESPACE_MAP)[$fragment[0]] ?? null; |
|||
assert(isset($ns)); |
|||
$fragmentContext = $doc->createElementNS($ns, $fragment[1]); |
|||
} |
|||
} else { |
|||
$fragmentContext = null; |
|||
} |
|||
// initialize the other classes we need |
|||
$decoder = new Data($data, "STDIN", $errorHandler, "UTF-8"); |
|||
$stack = new OpenElementsStack($fragmentContext); |
|||
$tokenizer = new Tokenizer($decoder, $stack, $errorHandler); |
|||
$tokenList = $tokenizer->tokenize(); |
|||
$treeBuilder = new TreeBuilder($doc, $decoder, $tokenizer, $tokenList, $errorHandler, $stack, new TemplateInsertionModesStack, $fragmentContext); |
|||
// run the tree builder |
|||
try { |
|||
$treeBuilder->constructTree(); |
|||
} catch (LoopException $e) { |
|||
$act = $this->balanceTree($this->serializeTree($doc, (bool) $fragmentContext), $exp); |
|||
$this->assertEquals($exp, $act, $e->getMessage()."\n".$treeBuilder->debugLog); |
|||
throw $e; |
|||
} catch (NotImplementedException $e) { |
|||
$this->markTestSkipped($e->getMessage()); |
|||
return; |
|||
} |
|||
$act = $this->balanceTree($this->serializeTree($doc, (bool) $fragmentContext), $exp); |
|||
$this->assertEquals($exp, $act, $treeBuilder->debugLog); |
|||
if ($errors !== false) { |
|||
// If $errors is false, the test does not include errors when there are in fact errors |
|||
$this->assertCount(sizeof($errors), $actualErrors, var_export($errors, true).var_export($actualErrors, true)); |
|||
} |
|||
} |
|||
|
|||
protected function patchTest(string $data, $fragment, array $errors, array $exp): array { |
|||
$patched = false; |
|||
$skip = ""; |
|||
// comments outside the root element are silently dropped by the PHP DOM |
|||
if (!$fragment) { |
|||
for ($a = 0; $a < sizeof($exp); $a++) { |
|||
if (strpos($exp[$a], "| <!--") === 0) { |
|||
array_splice($exp, $a--, 1); |
|||
$patched = true; |
|||
} |
|||
} |
|||
} |
|||
// some tests don't document errors when they should |
|||
if (!$errors && in_array($data, [ |
|||
// math.dat |
|||
'<math><tr><td><mo><tr>', |
|||
'<math><thead><mo><tbody>', |
|||
'<math><tfoot><mo><tbody>', |
|||
'<math><tbody><mo><tfoot>', |
|||
'<math><tbody><mo></table>', |
|||
'<math><thead><mo></table>', |
|||
'<math><tfoot><mo></table>', |
|||
// namespace-sensitivity.dat |
|||
'<body><table><tr><td><svg><td><foreignObject><span></td>Foo', |
|||
// svg.dat |
|||
'<svg><tr><td><title><tr>', |
|||
'<svg><thead><title><tbody>', |
|||
'<svg><tfoot><title><tbody>', |
|||
'<svg><tbody><title><tfoot>', |
|||
'<svg><tbody><title></table>', |
|||
'<svg><thead><title></table>', |
|||
'<svg><tfoot><title></table>', |
|||
// template.dat |
|||
'<template><a><table><a>', |
|||
// tests6.dat |
|||
'<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01//EN"><html></html>', |
|||
// tests8.dat |
|||
'<table><li><li></table>', |
|||
// webkit01.dat |
|||
'<table><tr><td><svg><desc><td></desc><circle>', |
|||
// webkit02.dat |
|||
'<legend>test</legend>', |
|||
'<table><input>', |
|||
'<b><em><foo><foo><aside></b>', |
|||
'<b><em><foo><foo><aside></b></em>', |
|||
'<b><em><foo><foo><foo><aside></b>', |
|||
'<b><em><foo><foo><foo><aside></b></em>', |
|||
'<b><em><foo><foo><foo><foo><foo><foo><foo><foo><foo><foo><aside></b></em>', |
|||
'<b><em><foo><foob><foob><foob><foob><fooc><fooc><fooc><fooc><food><aside></b></em>', |
|||
'<option><XH<optgroup></optgroup>', |
|||
'<svg><foreignObject><div>foo</div><plaintext></foreignObject></svg><div>bar</div>', |
|||
'<svg><foreignObject></foreignObject><title></svg>foo', |
|||
'</foreignObject><plaintext><div>foo</div>', |
|||
])) { |
|||
$errors = false; |
|||
} |
|||
// other tests do list errors, but they are plainly incorrect in missing some |
|||
if (in_array($data, [ |
|||
// doctype01.dat" |
|||
"<!DOCTYPE HTML PUBLIC \"-//W3C//DTD HTML 4.01//EN\"\n \"http://www.w3.org/TR/html4/strict.dtd\">Hello", |
|||
'<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN""http://www.w3.org/TR/html4/strict.dtd">', |
|||
'<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"\'http://www.w3.org/TR/html4/strict.dtd\'>', |
|||
'<!DOCTYPE HTML PUBLIC"-//W3C//DTD HTML 4.01//EN"\'http://www.w3.org/TR/html4/strict.dtd\'>', |
|||
"<!DOCTYPE HTML PUBLIC'-//W3C//DTD HTML 4.01//EN''http://www.w3.org/TR/html4/strict.dtd'>", |
|||
// entities02.dat |
|||
"<div>ZZ&prod=23</div>", |
|||
"<div>ZZÆ=</div>", |
|||
// foreign-fragment.dat |
|||
"<body><foo>", |
|||
"<p><foo>", |
|||
"<p></p><foo>", |
|||
// ruby.dat |
|||
"<html><ruby>a<rtc>b<span></ruby></html>", |
|||
// test1.dat |
|||
"<!-----><font><div>hello<table>excite!<b>me!<th><i>please!</tr><!--X-->", // this one is pretty hairy with buffered characters |
|||
])) { |
|||
$errors = false; |
|||
} |
|||
if ($errors) { |
|||
// some "old" errors are made redundant by "new" errors |
|||
$obsoleteSymbolList = implode("|", [ |
|||
"illegal-codepoint-for-numeric-entity", |
|||
"eof-in-attribute-value-double-quote", |
|||
"non-void-element-with-trailing-solidus", |
|||
"invalid-character-in-attribute-name", |
|||
"attributes-in-end-tag", |
|||
"expected-tag-name", |
|||
"unexpected-character-after-solidus-in-tag", |
|||
"expected-closing-tag-but-got-char", |
|||
"eof-in-tag-name", |
|||
"need-space-after-doctype", |
|||
"expected-doctype-name-but-got-right-bracket", |
|||
"expected-dashes-or-doctype", |
|||
"expected-space-or-right-bracket-in-doctype", |
|||
"unexpected-char-in-comment", |
|||
"eof-in-comment-double-dash", |
|||
"expected-named-entity", |
|||
"named-entity-without-semicolon", |
|||
"numeric-entity-without-semicolon", |
|||
"expected-numeric-entity", |
|||
"eof-in-attribute-name", |
|||
"unexpected-eof-in-text-mode", |
|||
"unexpected-EOF-after-solidus-in-tag", |
|||
"expected-attribute-name-but-got-eof", |
|||
"eof-in-script-in-script", |
|||
"expected-script-data-but-got-eof", |
|||
"unexpected-EOF-in-text-mode", |
|||
"expected-tag-name-but-got-question-mark", |
|||
"incorrect-comment", |
|||
"self-closing-flag-on-end-tag", |
|||
"invalid-codepoint", |
|||
"invalid-codepoint-in-body", |
|||
"invalid-codepoint-in-foreign-content", |
|||
"end-table-tag-in-caption", |
|||
"equals-in-unquoted-attribute-value", |
|||
"eof-in-numeric-entity", |
|||
"unexpected-char-in-doctype", |
|||
"unexpected-end-of-doctype", |
|||
"unexpected-dash-after-double-dash-in-comment", |
|||
"unexpected-bang-after-double-dash-in-comment", |
|||
]); |
|||
for ($a = 0, $stop = sizeof($errors); $a < $stop; $a++) { |
|||
if (preg_match("/^\(\d+,\d+\):? ($obsoleteSymbolList)$/", $errors[$a])) { |
|||
// these errors are redundant with "new" errors |
|||
unset($errors[$a]); |
|||
} |
|||
} |
|||
$errors = array_values($errors); |
|||
// some other errors appear to document implementation details |
|||
// rather than what the specificatioon dictates, or are |
|||
// simple duplicates |
|||
for ($a = 0, $stop = sizeof($errors); $a < $stop; $a++) { |
|||
if ( |
|||
preg_match("/^\(\d+,\d+\): unexpected-end-tag-in-special-element$/", $errors[$a]) |
|||
|| preg_match('/^\d+: Unclosed element “[^”]+”\.$/u', $errors[$a]) |
|||
|| ($data === '<!---x' && $errors[$a] === "(1:7) eof-in-comment") |
|||
|| ($data === "<!DOCTYPE html><body><table><caption><math><mi>foo</mi><mi>bar</mi>baz</table><p>quux" && $errors[$a] === "(1,78) expected-one-end-tag-but-got-another") |
|||
|| ($data === "<!DOCTYPE html><!-- XXX - XXX" && $errors[$a] === "(1,29): eof-in-comment") |
|||
|| ($data === "<!DOCTYPE html><!-- X" && $errors[$a] === "(1,21): eof-in-comment") |
|||
|| ($data === "<!doctype html><math></html>" && $errors[$a] === "(1,28): expected-one-end-tag-but-got-another") |
|||
|| ($data === "</" && $errors[$a] === "(1,2): expected-closing-tag-but-got-eof") |
|||
|| ($data === "<div foo=`bar`>" && $errors[$a] === "(1,14): unexpected-character-in-unquoted-attribute-value") |
|||
|| ( |
|||
$errors[$a] === "51: Self-closing syntax (“/>”) used on a non-void HTML element. Ignoring the slash and treating as a start tag." |
|||
&& ( |
|||
$data === "<b></b><mglyph/><i></i><malignmark/><u></u><mtext/>X" |
|||
|| $data === "<b></b><mglyph/><i></i><malignmark/><u></u><mi/>X" |
|||
|| $data === "<b></b><mglyph/><i></i><malignmark/><u></u><mo/>X" |
|||
|| $data === "<b></b><mglyph/><i></i><malignmark/><u></u><mn/>X" |
|||
|| $data === "<b></b><mglyph/><i></i><malignmark/><u></u><ms/>X" |
|||
) |
|||
) |
|||
|| ($data === "&ammmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmp;" && $errors[$a] === "(1,950): unknown-named-character-reference") |
|||
|| ($data === "&ammmp;" && $errors[$a] === "(1,7): unknown-named-character-reference") |
|||
|| ($data === "FOO<!-- BAR -- <QUX> -- MUX -- >BAZ" && $errors[$a] === "(1,35): eof-in-comment") |
|||
|| ($data === "FOO<!-- BAR -- >BAZ" && $errors[$a] === "(1,21): eof-in-comment") |
|||
) { |
|||
// these errors seems to simply be redundant |
|||
unset($errors[$a]); |
|||
} |
|||
} |
|||
$errors = array_values($errors); |
|||
// other errors are spurious, or are for runs of character tokens |
|||
for ($a = 0, $stop = sizeof($errors); $a < $stop; $a++) { |
|||
if (preg_match("/^\((\d+),(\d+)\):? (foster-parenting-character(?:-in-table)?|unexpected-character-in-colgroup|unexpected-char-after-frameset|unexpected-char-in-frameset|expected-eof-but-got-char)$/", $errors[$a], $m1) && preg_match("/^\((\d+),(\d+)\):? $m1[3]$/", $errors[$a + 1] ?? "", $m2)) { |
|||
// if the next error is also a character error at the next or same character position, this implies a run of characters where we only have one token |
|||
// technically we should be reporting each one, so this is properly a FIXME |
|||
if ($m1[1] == $m2[1] && ($m1[2] + 1 == $m2[2] || $m1[2] == $m2[2])) { |
|||
unset($errors[$a]); |
|||
$patched = true; |
|||
} |
|||
} elseif (preg_match("/^foster-parenting text /", $errors[$a]) && preg_match("/^foster-parenting text /", $errors[$a + 1] ?? "")) { |
|||
// template tests have a different format of error message |
|||
unset($errors[$a]); |
|||
$patched = true; |
|||
} elseif (preg_match("/^\((\d+,\d+)\):? unexpected-end-tag$/", $errors[$a], $m) && preg_match("/^\($m[1]\):? (unexpected-end-tag|end-tag-too-early|expected-one-end-tag-but-got-another|adoption-agency-1.3)$/", $errors[$a + 1] ?? "")) { |
|||
// unexpected-end-tag errors should only be reported once for a given tag |
|||
unset($errors[$a]); |
|||
} |
|||
} |
|||
$errors = array_values($errors); |
|||
} |
|||
return [$exp, $errors, $patched, $skip]; |
|||
} |
|||
|
|||
protected function balanceTree(array $act, array $exp): array { |
|||
// makes sure that the actual tree contain the same number of lines as the expected tree |
|||
// lines are inserted where the two trees diverge, until the end of the actual tree is reached |
|||
// this usually results in cleaner PHPUnit comparison failure output |
|||
for ($a = 0; $a < sizeof($act) && sizeof($act) < sizeof($exp); $a++) { |
|||
if (!isset($act[$a]) || $exp[$a] !== $act[$a]) { |
|||
array_splice($act, $a, 0, [""]); |
|||
} |
|||
} |
|||
return $act; |
|||
} |
|||
|
|||
protected function push(string $data): void { |
|||
$this->out[] = "| ".str_repeat(" ", $this->depth).$data; |
|||
} |
|||
|
|||
protected function serializeTree(\DOMDocument $d, bool $fragment): array { |
|||
$this->out = []; |
|||
$this->depth = 0; |
|||
if ($fragment){ |
|||
foreach ($d->documentElement->childNodes as $n) { |
|||
$this->serializeNode($n); |
|||
} |
|||
} else { |
|||
if ($d->doctype) { |
|||
$dt = "<!DOCTYPE "; |
|||
$dt .= ($d->doctype->name !== " ") ? $d->doctype->name : ""; |
|||
if (strlen($d->doctype->publicId) || strlen($d->doctype->systemId)) { |
|||
$dt .= ' "'.$d->doctype->publicId.'"'; |
|||
$dt .= ' "'.$d->doctype->systemId.'"'; |
|||
} |
|||
$dt .= ">"; |
|||
$this->push($dt); |
|||
} |
|||
if ($d->documentElement) { |
|||
$this->serializeElement($d->documentElement); |
|||
} |
|||
} |
|||
return $this->out; |
|||
} |
|||
|
|||
protected function serializeElement(\DOMElement $e): void { |
|||
if ($e->namespaceURI) { |
|||
$prefix = Parser::NAMESPACE_MAP[$e->namespaceURI]; |
|||
assert((bool) $prefix, new \Exception("Prefix for namespace {$e->namespaceURI} is not defined")); |
|||
$prefix .= " "; |
|||
} else { |
|||
$prefix = ""; |
|||
} |
|||
$localName = $this->uncoerceName($e->localName); |
|||
$this->push("<".$prefix.$localName.">"); |
|||
$this->depth++; |
|||
$attr = []; |
|||
foreach ($e->attributes as $a) { |
|||
$prefix = ""; |
|||
if ($a->namespaceURI) { |
|||
$prefix = Parser::NAMESPACE_MAP[$a->namespaceURI]; |
|||
assert((bool) $prefix, new \Exception("Prefix for namespace {$a->namespaceURI} is not defined")); |
|||
$prefix .= " "; |
|||
} |
|||
$attr[$prefix.$this->uncoerceName($a->name)] = $a->value; |
|||
} |
|||
ksort($attr, \SORT_STRING); |
|||
foreach ($attr as $k => $v) { |
|||
$this->push($k.'="'.$v.'"'); |
|||
} |
|||
if ($e->localName === "template" && $e->namespaceURI === null) { |
|||
$this->push("content"); |
|||
$this->depth++; |
|||
foreach ($e->childNodes as $n) { |
|||
$this->serializeNode($n); |
|||
} |
|||
$this->depth--; |
|||
} else { |
|||
foreach ($e->childNodes as $n) { |
|||
$this->serializeNode($n); |
|||
} |
|||
} |
|||
$this->depth--; |
|||
} |
|||
|
|||
protected function serializeNode(\DOMNode $n): void { |
|||
if ($n instanceof \DOMElement) { |
|||
$this->serializeElement($n); |
|||
} elseif ($n instanceof \DOMProcessingInstruction) { |
|||
$this->push("<?".$n->target." ".$n->data.">");
|
|||
} elseif ($n instanceof \DOMComment) { |
|||
$this->push("<!-- ".$n->data." -->"); |
|||
} elseif ($n instanceof \DOMCharacterData) { |
|||
$this->push('"'.$n->data.'"'); |
|||
} else { |
|||
throw new \Exception("Node type ".get_class($n)." not handled"); |
|||
} |
|||
} |
|||
|
|||
public function provideStandardTreeTests(): iterable { |
|||
$blacklist = []; |
|||
$files = new \AppendIterator(); |
|||
$files->append(new \GlobIterator(\MensBeam\HTML\DOM\BASE."tests/html5lib-tests/tree-construction/*.dat", \FilesystemIterator::SKIP_DOTS | \FilesystemIterator::CURRENT_AS_PATHNAME)); |
|||
$files->append(new \GlobIterator(\MensBeam\HTML\DOM\BASE."tests/cases/tree-construction/*.dat", \FilesystemIterator::SKIP_DOTS | \FilesystemIterator::CURRENT_AS_PATHNAME)); |
|||
foreach ($files as $file) { |
|||
$index = 0; |
|||
$l = 0; |
|||
if (!in_array(basename($file), $blacklist)) { |
|||
$lines = array_map(function($v) { |
|||
return rtrim($v, "\n"); |
|||
}, file($file)); |
|||
while ($l < sizeof($lines)) { |
|||
$pos = $l + 1; |
|||
assert($lines[$l] === "#data", new \Exception("Test $file #$index does not start with #data tag at line ".($l + 1))); |
|||
// collect the test input |
|||
$data = []; |
|||
for (++$l; $l < sizeof($lines); $l++) { |
|||
if ($lines[$l] === "#errors") { |
|||
break; |
|||
} |
|||
$data[] = $lines[$l]; |
|||
} |
|||
$data = implode("\n", $data); |
|||
// collect the test errors |
|||
$errors = []; |
|||
assert(($lines[$l] ?? "") === "#errors", new \Exception("Test $file #$index does not list errors at line ".($l + 1))); |
|||
for (++$l; $l < sizeof($lines); $l++) { |
|||
if ($lines[$l] === "#new-errors") { |
|||
continue; |
|||
} elseif (preg_match('/^#(document(-fragment)?|script-(on|off)|)$/', $lines[$l])) { |
|||
break; |
|||
} |
|||
$errors[] = $lines[$l]; |
|||
} |
|||
// set the script mode, if present |
|||
assert(preg_match('/^#(script-(on|off)|document(-fragment)?)$/', $lines[$l]) === 1, new \Exception("Test $file #$index follows errors with something other than script flag, document fragment, or document at line ".($l + 1))); |
|||
$script = null; |
|||
if ($lines[$l] === "#script-off") { |
|||
$script = false; |
|||
$l++; |
|||
} elseif ($lines[$l] === "#script-on") { |
|||
$script = true; |
|||
$l++; |
|||
} |
|||
// collect the document fragment, if present |
|||
assert(preg_match('/^#document(-fragment)?$/', $lines[$l]) === 1, new \Exception("Test $file #$index follows script flag with something other than document fragment or document at line ".($l + 1))); |
|||
$fragment = null; |
|||
if ($lines[$l] === "#document-fragment") { |
|||
$fragment = $lines[++$l]; |
|||
$l++; |
|||
} |
|||
// collect the output tree |
|||
$exp = []; |
|||
assert($lines[$l] === "#document", new \Exception("Test $file #$index follows document fragment with something other than document at line ".($l + 1))); |
|||
for (++$l; $l < sizeof($lines); $l++) { |
|||
if ($lines[$l] === "" && ($lines[$l + 1] ?? "") === "#data") { |
|||
break; |
|||
} elseif (($lines[$l][0] ?? "") !== "|") { |
|||
// apend the data to the previous token |
|||
$exp[sizeof($exp) - 1] .= "\n".$lines[$l]; |
|||
continue; |
|||
} |
|||
assert(preg_match('/^[^#]/', $lines[$l]) === 1, new \Exception("Test $file #$index contains unrecognized data after document at line ".($l + 1))); |
|||
$exp[] = $lines[$l]; |
|||
} |
|||
if (!$script) { |
|||
// scripting-dependent tests are skipped entirely since we will not support scripting |
|||
yield basename($file)." #$index (line $pos)" => [$data, $exp, $errors, $fragment]; |
|||
} |
|||
$l++; |
|||
$index++; |
|||
} |
|||
} |
|||
} |
|||
} |
|||
} |
@ -1,36 +0,0 @@ |
|||
#data |
|||
<!DOCTYPE HTML> |
|||
<meta charset="x-user-defined"> |
|||
#encoding |
|||
Windows-1252 |
|||
|
|||
#data |
|||
<!DOCTYPE HTML> |
|||
<meta charset="utf-8" charset="windows-1252"> |
|||
#encoding |
|||
UTF-8 |
|||
|
|||
#data |
|||
<!DOCTYPE HTML> |
|||
<!-- There is a space at the end of the next line. This must not be removed. --> |
|||
<meta charset |
|||
#encoding |
|||
Windows-1252 |
|||
|
|||
#data |
|||
<!DOCTYPE HTML> |
|||
<meta charset=> |
|||
#encoding |
|||
Windows-1252 |
|||
|
|||
#data |
|||
<!DOCTYPE HTML> |
|||
<meta http-equiv="Content-Type" content="text/html; charset charset="> |
|||
#encoding |
|||
Windows-1252 |
|||
|
|||
#data |
|||
<!DOCTYPE HTML> |
|||
<meta http-equiv="Content-Type" content="text/html; charset charset=utf-8"> |
|||
#encoding |
|||
UTF-8 |
@ -1,27 +0,0 @@ |
|||
{ |
|||
"tests": [ |
|||
{ |
|||
"description":"Whitespace character references", |
|||
"initialStates":["RCDATA state"], |
|||
"input":" ", |
|||
"output":[["Character", " "]] |
|||
}, |
|||
{ |
|||
"description":"Script end tag with whitespace", |
|||
"initialStates":["Script data state"], |
|||
"lastStartTag":"script", |
|||
"input":"</script >", |
|||
"output":[["EndTag", "script"]] |
|||
}, |
|||
{ |
|||
"description":"Self-closing script end tag", |
|||
"initialStates":["Script data state"], |
|||
"lastStartTag":"script", |
|||
"input":"</script/>", |
|||
"output":[["EndTag", "script"]], |
|||
"errors":[ |
|||
{ "code": "end-tag-with-trailing-solidus", "line": 1, "col": 10 } |
|||
] |
|||
} |
|||
] |
|||
} |
@ -1,68 +0,0 @@ |
|||
#data |
|||
<!DOCTYPE html><svg xmlns="http://www.w3.org/2000/svg"/> |
|||
#errors |
|||
#document |
|||
| <!DOCTYPE html> |
|||
| <html> |
|||
| <head> |
|||
| <body> |
|||
| <svg svg> |
|||
| xmlns xmlns="http://www.w3.org/2000/svg" |
|||
|
|||
#data |
|||
<!DOCTYPE html><svg xmlns="http://www.w3.org/1999/xlink"/> |
|||
#errors |
|||
(1,58): invalid-namespace-attribute-value |
|||
#document |
|||
| <!DOCTYPE html> |
|||
| <html> |
|||
| <head> |
|||
| <body> |
|||
| <svg svg> |
|||
|
|||
#data |
|||
<!DOCTYPE html><svg xmlns:xlink="http://www.w3.org/1999/xlink"/> |
|||
#errors |
|||
#document |
|||
| <!DOCTYPE html> |
|||
| <html> |
|||
| <head> |
|||
| <body> |
|||
| <svg svg> |
|||
| xmlns xlink="http://www.w3.org/1999/xlink" |
|||
|
|||
#data |
|||
<!DOCTYPE html><svg xlink:href="http://example.com/"/> |
|||
#errors |
|||
#document |
|||
| <!DOCTYPE html> |
|||
| <html> |
|||
| <head> |
|||
| <body> |
|||
| <svg svg> |
|||
| xlink href="http://example.com/" |
|||
|
|||
#data |
|||
<!DOCTYPE html><svg xmlns:xlink="http://www.w3.org/1999/xhtml" xlink:href="http://example.com/"/> |
|||
#errors |
|||
(1,97): invalid-namespace-attribute-value |
|||
#document |
|||
| <!DOCTYPE html> |
|||
| <html> |
|||
| <head> |
|||
| <body> |
|||
| <svg svg> |
|||
| xlink href="http://example.com/" |
|||
|
|||
#data |
|||
<!DOCTYPE html><svg xml:id="proper"/><p xml:id="bogus"> |
|||
#errors |
|||
#document |
|||
| <!DOCTYPE html> |
|||
| <html> |
|||
| <head> |
|||
| <body> |
|||
| <svg svg> |
|||
| xml id="proper" |
|||
| <p> |
|||
| xml:id="bogus" |
Loading…
Reference in new issue