|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
namespace arc\html; |
|
4
|
|
|
|
|
5
|
|
|
class Parser |
|
6
|
|
|
{ |
|
7
|
|
|
public $options = [ |
|
8
|
|
|
'libxml_options' => 0 |
|
9
|
|
|
]; |
|
10
|
|
|
|
|
11
|
|
View Code Duplication |
public function __construct( $options = array() ) |
|
|
|
|
|
|
12
|
|
|
{ |
|
13
|
|
|
$optionList = [ 'libxml_options' ]; |
|
14
|
|
|
foreach( $options as $option => $optionValue ) { |
|
15
|
|
|
if ( in_array( $option, $optionList ) ) { |
|
16
|
|
|
$this->{$option} = $optionValue; |
|
17
|
|
|
} |
|
18
|
|
|
} |
|
19
|
|
|
} |
|
20
|
|
|
|
|
21
|
|
|
public function parse( $html, $encoding = 'UTF-8' ) |
|
22
|
|
|
{ |
|
23
|
|
|
if ( !$html ) { |
|
24
|
|
|
return \arc\html\Proxy( null ); |
|
25
|
|
|
} |
|
26
|
|
|
if ( $html instanceof Proxy ) { // already parsed |
|
27
|
|
|
return $html; |
|
28
|
|
|
} |
|
29
|
|
|
$html = (string) $html; |
|
30
|
|
|
if ( stripos($html, '<html>')!==false ) { |
|
31
|
|
|
return $this->parseFull( $html, $encoding ); |
|
32
|
|
|
} else { |
|
33
|
|
|
return $this->parsePartial( $html, $encoding ); |
|
34
|
|
|
} |
|
35
|
|
|
} |
|
36
|
|
|
|
|
37
|
|
|
private function parsePartial( $html, $encoding = 'UTF-8' ) |
|
38
|
|
|
{ |
|
39
|
|
|
$result = $this->parseFull( '<div id="ArcPartialHTML">'.$html.'</div>', $encoding ); |
|
40
|
|
|
if ( $result ) { |
|
41
|
|
|
$result = new \arc\html\Proxy( $result->find('#ArcPartialHTML')[0]->children(), $this ); |
|
42
|
|
|
} else { |
|
43
|
|
|
throw new \arc\Exception('parse error'); |
|
44
|
|
|
} |
|
45
|
|
|
return $result; |
|
46
|
|
|
} |
|
47
|
|
|
|
|
48
|
|
|
private function parseFull( $html, $encoding = 'UTF-8') |
|
49
|
|
|
{ |
|
50
|
|
|
$dom = new \DomDocument('1.0', $encoding); |
|
51
|
|
|
$prefix = ''; |
|
52
|
|
|
if ( isset($encoding) ) { |
|
53
|
|
|
$prefix = "<head id='ar_html_parser_encoding_header'><meta charset='$encoding'></head>"; |
|
54
|
|
|
} |
|
55
|
|
|
libxml_disable_entity_loader(); // prevents XXE attacks |
|
56
|
|
|
$prevErrorSetting = libxml_use_internal_errors(true); |
|
57
|
|
|
if ( $dom->loadHTML( $prefix . $html, $this->options['libxml_options'] ) ) { |
|
58
|
|
|
if ( isset($encoding) ) { |
|
59
|
|
|
$elm = $dom->getElementById('ar_html_parser_encoding_header'); |
|
60
|
|
|
if ( isset($elm) ) { |
|
61
|
|
|
$elm->parentNode->removeChild($elm); |
|
62
|
|
|
} |
|
63
|
|
|
} |
|
64
|
|
|
libxml_use_internal_errors( $prevErrorSetting ); |
|
65
|
|
|
return new \arc\html\Proxy( simplexml_import_dom( $dom ), $this ); |
|
66
|
|
|
} |
|
67
|
|
|
$errors = libxml_get_errors(); |
|
68
|
|
|
libxml_clear_errors(); |
|
69
|
|
|
libxml_use_internal_errors( $prevErrorSetting ); |
|
70
|
|
|
$message = 'Incorrect html passed.'; |
|
71
|
|
|
foreach ( $errors as $error ) { |
|
72
|
|
|
$message .= "\nline: ".$error->line."; column: ".$error->column."; ".$error->message; |
|
73
|
|
|
} |
|
74
|
|
|
throw new \arc\Exception( $message, \arc\exceptions::ILLEGAL_ARGUMENT ); |
|
75
|
|
|
} |
|
76
|
|
|
|
|
77
|
|
|
} |
|
78
|
|
|
|
Ariadne-CMS /
arc-html
Duplicated code is one of the most pungent code smells. If you need to duplicate the same code in three or more different places, we strongly encourage you to look into extracting the code into a single class or operation.
You can also find more detailed suggestions in the “Code” section of your repository.