1
|
|
|
<?php |
2
|
|
|
|
3
|
|
|
namespace Goose\Modules\Extractors; |
4
|
|
|
|
5
|
|
|
use Goose\Article; |
6
|
|
|
use Goose\Utils\Helper; |
7
|
|
|
use Goose\Traits\ArticleMutatorTrait; |
8
|
|
|
use Goose\Modules\AbstractModule; |
9
|
|
|
use Goose\Modules\ModuleInterface; |
10
|
|
|
use DOMWrap\Element; |
11
|
|
|
|
12
|
|
|
/** |
13
|
|
|
* Additional Data Extractor |
14
|
|
|
* |
15
|
|
|
* @package Goose\Modules\Extractors |
16
|
|
|
* @license http://www.apache.org/licenses/LICENSE-2.0 Apache License 2.0 |
17
|
|
|
*/ |
18
|
|
|
class AdditionalDataExtractor extends AbstractModule implements ModuleInterface { |
19
|
|
|
use ArticleMutatorTrait; |
20
|
|
|
|
21
|
|
|
/** @var string */ |
22
|
|
|
protected static $A_REL_TAG_SELECTOR = "a[rel='tag'], a[href*='/tag/']"; |
23
|
|
|
|
24
|
|
|
/** @var string[] */ |
25
|
|
|
protected static $VIDEO_PROVIDERS = [ |
26
|
|
|
'youtube\.com', |
27
|
|
|
'youtu\.be', |
28
|
|
|
'vimeo\.com', |
29
|
|
|
'blip\.tv', |
30
|
|
|
'dailymotion\.com', |
31
|
|
|
'dai\.ly', |
32
|
|
|
'flickr\.com', |
33
|
|
|
'flic\.kr', |
34
|
|
|
]; |
35
|
|
|
|
36
|
|
|
/** |
37
|
|
|
* @param Article $article |
38
|
|
|
*/ |
39
|
|
|
public function run(Article $article) { |
40
|
|
|
$this->article($article); |
41
|
|
|
|
42
|
|
|
$article->setTags($this->getTags()); |
43
|
|
|
|
44
|
|
|
if ($this->article()->getTopNode() instanceof Element) { |
45
|
|
|
$article->setVideos($this->getVideos()); |
46
|
|
|
$article->setLinks($this->getLinks()); |
47
|
|
|
$article->setPopularWords($this->getPopularWords()); |
48
|
|
|
} |
49
|
|
|
} |
50
|
|
|
|
51
|
|
|
/** |
52
|
|
|
* @return string[] |
53
|
|
|
*/ |
54
|
|
|
private function getTags() { |
55
|
|
|
$nodes = $this->article()->getDoc()->find(self::$A_REL_TAG_SELECTOR); |
56
|
|
|
|
57
|
|
|
$tags = []; |
58
|
|
|
|
59
|
|
|
foreach ($nodes as $node) { |
60
|
|
|
$tags[] = Helper::textNormalise($node->text()); |
61
|
|
|
} |
62
|
|
|
|
63
|
|
|
return $tags; |
64
|
|
|
} |
65
|
|
|
|
66
|
|
|
/** |
67
|
|
|
* Pulls out videos we like |
68
|
|
|
* |
69
|
|
|
* @return string[] |
70
|
|
|
*/ |
71
|
|
|
private function getVideos() { |
72
|
|
|
$videos = []; |
73
|
|
|
|
74
|
|
|
$nodes = $this->article()->getTopNode()->parent()->find('embed, object, iframe'); |
75
|
|
|
|
76
|
|
|
foreach ($nodes as $node) { |
77
|
|
|
if ($node->hasAttribute('src')) { |
78
|
|
|
$src = $node->attr('src'); |
79
|
|
|
} else { |
80
|
|
|
$src = $node->attr('data'); |
81
|
|
|
} |
82
|
|
|
|
83
|
|
|
$match = array_reduce(self::$VIDEO_PROVIDERS, function($match, $domain) use ($src) { |
84
|
|
|
$srcHost = parse_url($src, PHP_URL_HOST); |
85
|
|
|
$srcScheme = parse_url($src, PHP_URL_SCHEME); |
86
|
|
|
|
87
|
|
|
return $match || preg_match('@' . $domain . '$@i', $srcHost) && in_array($srcScheme, ['http', 'https']); |
88
|
|
|
}); |
89
|
|
|
|
90
|
|
|
if ($match) { |
91
|
|
|
$videos[] = $src; |
92
|
|
|
} |
93
|
|
|
} |
94
|
|
|
|
95
|
|
|
return $videos; |
96
|
|
|
} |
97
|
|
|
|
98
|
|
|
/** |
99
|
|
|
* Pulls out links we like |
100
|
|
|
* |
101
|
|
|
* @return string[] |
102
|
|
|
*/ |
103
|
|
|
private function getLinks() { |
104
|
|
|
$goodLinks = []; |
105
|
|
|
|
106
|
|
|
$candidates = $this->article()->getTopNode()->parent()->find('a[href]'); |
107
|
|
|
|
108
|
|
|
foreach ($candidates as $el) { |
109
|
|
|
if ($el->attr('href') != '#' && trim($el->attr('href')) != '') { |
110
|
|
|
$goodLinks[] = [ |
111
|
|
|
'url' => $el->attr('href'), |
112
|
|
|
'text' => Helper::textNormalise($el->text()), |
113
|
|
|
]; |
114
|
|
|
} |
115
|
|
|
} |
116
|
|
|
|
117
|
|
|
return $goodLinks; |
118
|
|
|
} |
119
|
|
|
|
120
|
|
|
/** |
121
|
|
|
* @return string[] |
122
|
|
|
*/ |
123
|
|
|
private function getPopularWords() { |
124
|
|
|
$limit = 5; |
125
|
|
|
$minimumFrequency = 1; |
126
|
|
|
$stopWords = $this->config()->getStopWords()->getCurrentStopWords(); |
127
|
|
|
|
128
|
|
|
$text = $this->article()->getTitle(); |
129
|
|
|
$text .= ' ' . $this->article()->getMetaDescription(); |
130
|
|
|
|
131
|
|
|
if ($this->article()->getTopNode()) { |
132
|
|
|
$text .= ' ' . $this->article()->getCleanedArticleText(); |
133
|
|
|
} |
134
|
|
|
|
135
|
|
|
// Decode and split words by white-space |
136
|
|
|
$text = html_entity_decode($text, ENT_COMPAT | ENT_HTML5, 'UTF-8'); |
137
|
|
|
$words = preg_split('@[\s]+@iu', $text, -1, PREG_SPLIT_NO_EMPTY); |
138
|
|
|
|
139
|
|
|
// Determine stop words currently in $words |
140
|
|
|
$ignoreWords = array_intersect($words, $stopWords); |
141
|
|
|
// Remove ignored words from $words |
142
|
|
|
$words = array_diff($words, $ignoreWords); |
143
|
|
|
|
144
|
|
|
// Count and sort $words |
145
|
|
|
$words = array_count_values($words); |
146
|
|
|
arsort($words); |
147
|
|
|
|
148
|
|
|
// Limit and filter $words |
149
|
|
|
$words = array_slice($words, 0, $limit); |
150
|
|
|
$words = array_filter($words, function($value) use ($minimumFrequency){ |
151
|
|
|
return !($value < $minimumFrequency); |
152
|
|
|
}); |
153
|
|
|
|
154
|
|
|
return $words; |
155
|
|
|
} |
156
|
|
|
} |
157
|
|
|
|