|
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
|
|
|
$parentNode = $this->article()->getTopNode()->parent(); |
|
75
|
|
|
|
|
76
|
|
|
if ($parentNode instanceof Element) { |
|
77
|
|
|
$nodes = $parentNode->find('embed, object, iframe'); |
|
78
|
|
|
|
|
79
|
|
|
foreach ($nodes as $node) { |
|
80
|
|
|
if ($node->hasAttribute('src')) { |
|
81
|
|
|
$src = $node->attr('src'); |
|
82
|
|
|
} else { |
|
83
|
|
|
$src = $node->attr('data'); |
|
84
|
|
|
} |
|
85
|
|
|
|
|
86
|
|
|
$match = array_reduce(self::$VIDEO_PROVIDERS, function($match, $domain) use ($src) { |
|
87
|
|
|
$srcHost = parse_url($src, PHP_URL_HOST); |
|
88
|
|
|
$srcScheme = parse_url($src, PHP_URL_SCHEME); |
|
89
|
|
|
|
|
90
|
|
|
return $match || preg_match('@' . $domain . '$@i', $srcHost) && in_array($srcScheme, ['http', 'https']); |
|
91
|
|
|
}); |
|
92
|
|
|
|
|
93
|
|
|
if ($match) { |
|
94
|
|
|
$videos[] = $src; |
|
95
|
|
|
} |
|
96
|
|
|
} |
|
97
|
|
|
} |
|
98
|
|
|
|
|
99
|
|
|
return $videos; |
|
100
|
|
|
} |
|
101
|
|
|
|
|
102
|
|
|
/** |
|
103
|
|
|
* Pulls out links we like |
|
104
|
|
|
* |
|
105
|
|
|
* @return string[] |
|
106
|
|
|
*/ |
|
107
|
|
|
private function getLinks() { |
|
108
|
|
|
$goodLinks = []; |
|
109
|
|
|
|
|
110
|
|
|
$parentNode = $this->article()->getTopNode()->parent(); |
|
111
|
|
|
|
|
112
|
|
|
if ($parentNode instanceof Element) { |
|
113
|
|
|
$candidates = $parentNode->find('a[href]'); |
|
114
|
|
|
|
|
115
|
|
|
foreach ($candidates as $el) { |
|
116
|
|
|
if ($el->attr('href') != '#' && trim($el->attr('href')) != '') { |
|
117
|
|
|
$goodLinks[] = [ |
|
118
|
|
|
'url' => $el->attr('href'), |
|
119
|
|
|
'text' => Helper::textNormalise($el->text()), |
|
120
|
|
|
]; |
|
121
|
|
|
} |
|
122
|
|
|
} |
|
123
|
|
|
} |
|
124
|
|
|
|
|
125
|
|
|
return $goodLinks; |
|
126
|
|
|
} |
|
127
|
|
|
|
|
128
|
|
|
/** |
|
129
|
|
|
* @return string[] |
|
130
|
|
|
*/ |
|
131
|
|
|
private function getPopularWords() { |
|
132
|
|
|
$limit = 5; |
|
133
|
|
|
$minimumFrequency = 1; |
|
134
|
|
|
$stopWords = $this->config()->getStopWords()->getCurrentStopWords(); |
|
135
|
|
|
|
|
136
|
|
|
$text = $this->article()->getTitle(); |
|
137
|
|
|
$text .= ' ' . $this->article()->getMetaDescription(); |
|
138
|
|
|
|
|
139
|
|
|
if ($this->article()->getTopNode()) { |
|
140
|
|
|
$text .= ' ' . $this->article()->getCleanedArticleText(); |
|
141
|
|
|
} |
|
142
|
|
|
|
|
143
|
|
|
// Decode and split words by white-space |
|
144
|
|
|
$text = html_entity_decode($text, ENT_COMPAT | ENT_HTML5, 'UTF-8'); |
|
145
|
|
|
$words = preg_split('@[\s]+@iu', $text, -1, PREG_SPLIT_NO_EMPTY); |
|
146
|
|
|
|
|
147
|
|
|
// Determine stop words currently in $words |
|
148
|
|
|
$ignoreWords = array_intersect($words, $stopWords); |
|
149
|
|
|
// Remove ignored words from $words |
|
150
|
|
|
$words = array_udiff($words, $ignoreWords, 'strcasecmp'); |
|
151
|
|
|
|
|
152
|
|
|
// Count and sort $words |
|
153
|
|
|
$words = array_count_values($words); |
|
154
|
|
|
arsort($words); |
|
155
|
|
|
|
|
156
|
|
|
// Limit and filter $words |
|
157
|
|
|
$words = array_slice($words, 0, $limit); |
|
158
|
|
|
$words = array_filter($words, function($value) use ($minimumFrequency){ |
|
159
|
|
|
return !($value < $minimumFrequency); |
|
160
|
|
|
}); |
|
161
|
|
|
|
|
162
|
|
|
return $words; |
|
163
|
|
|
} |
|
164
|
|
|
} |
|
165
|
|
|
|
scotteh /
php-goose
