|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
namespace Zrashwani\NewsScrapper\Adapters; |
|
4
|
|
|
|
|
5
|
|
|
use Symfony\Component\DomCrawler\Crawler; |
|
6
|
|
|
use Zrashwani\NewsScrapper\Selector; |
|
7
|
|
|
|
|
8
|
|
|
/** |
|
9
|
|
|
* Base class that defines skeleton of the any adapter implemented |
|
10
|
|
|
* |
|
11
|
|
|
* @author Zeid Rashwani <zrashwani.com> |
|
12
|
|
|
*/ |
|
13
|
|
|
abstract class AbstractAdapter |
|
14
|
|
|
{ |
|
15
|
|
|
public $currentUrl; |
|
16
|
|
|
|
|
17
|
|
|
abstract public function extractTitle(Crawler $crawler); |
|
18
|
|
|
|
|
19
|
|
|
abstract public function extractImage(Crawler $crawler); |
|
20
|
|
|
|
|
21
|
|
|
abstract public function extractDescription(Crawler $crawler); |
|
22
|
|
|
|
|
23
|
|
|
abstract public function extractKeywords(Crawler $crawler); |
|
24
|
|
|
|
|
25
|
|
|
abstract public function extractBody(Crawler $crawler); |
|
26
|
|
|
|
|
27
|
|
|
abstract public function extractPublishDate(Crawler $crawler); |
|
28
|
|
|
|
|
29
|
|
|
abstract public function extractAuthor(Crawler $crawler); |
|
30
|
|
|
|
|
31
|
|
|
/** |
|
32
|
|
|
* normalize link and turn it into absolute format |
|
33
|
|
|
* @param string $link |
|
34
|
|
|
* @param boolean $remove_hashes if set true, hashes will be removed from url |
|
35
|
|
|
* @return string |
|
36
|
|
|
*/ |
|
37
|
|
|
public function normalizeLink($link, $remove_hashes = false) |
|
38
|
|
|
{ |
|
39
|
|
|
$baseUrl = $this->currentUrl; |
|
40
|
|
|
if (preg_match('@^http(s?)://.*$@', $baseUrl) === 0 && //local environment assumed here |
|
41
|
|
|
preg_match('@^http(s?)://.*$@', $link) === 0) { |
|
42
|
|
|
$link = pathinfo($baseUrl, PATHINFO_DIRNAME).'/'.$link; |
|
43
|
|
|
} elseif (preg_match('@^http(s?)://.*$@', $link) === 0) { //is not absolute |
|
44
|
|
|
$urlParts = parse_url($baseUrl); |
|
45
|
|
|
$scheme = isset($urlParts['scheme']) === true ? $urlParts['scheme'] : 'http'; |
|
46
|
|
|
$host = isset($urlParts['host']) === true ? $urlParts['host'] : ''; |
|
47
|
|
|
if (strpos($link, '//') === 0) { //begins with // |
|
48
|
|
|
$link = $scheme.':'.$link; |
|
49
|
|
|
} elseif (strpos($link, '/') === 0) { //begins with / |
|
50
|
|
|
$link = $scheme.'://'.$host.$link; |
|
51
|
|
|
} else { |
|
52
|
|
|
$link = $scheme.'://'.$host.'/'.$link; |
|
53
|
|
|
} |
|
54
|
|
|
} |
|
55
|
|
|
|
|
56
|
|
|
if ($remove_hashes === true) { |
|
57
|
|
|
$link = preg_replace('@#.*$@', '', $link); |
|
58
|
|
|
} |
|
59
|
|
|
return $link; |
|
60
|
|
|
} |
|
61
|
|
|
|
|
62
|
|
|
/** |
|
63
|
|
|
* normalizing html scrapped by removing unwanted tags (ex. script, css) |
|
64
|
|
|
* and amending external resources paths |
|
65
|
|
|
* @param string $raw_html |
|
66
|
|
|
* @return string |
|
67
|
|
|
*/ |
|
68
|
|
|
public function normalizeHtml($raw_html) |
|
69
|
|
|
{ |
|
70
|
|
|
if (empty($raw_html)) { |
|
71
|
|
|
return $raw_html; |
|
72
|
|
|
} |
|
73
|
|
|
|
|
74
|
|
|
$disallowed_tags = ['script', 'style', 'meta', 'form', 'aside']; |
|
75
|
|
|
|
|
76
|
|
|
$xmlDoc = new \DOMDocument(); |
|
77
|
|
|
libxml_use_internal_errors(true); |
|
78
|
|
|
$xmlDoc->loadHTML(mb_convert_encoding($raw_html, 'HTML-ENTITIES', 'UTF-8')); |
|
79
|
|
|
libxml_clear_errors(); |
|
80
|
|
|
|
|
81
|
|
|
$xpath = new \DOMXPath($xmlDoc); |
|
82
|
|
|
foreach ($disallowed_tags as $tag) { |
|
83
|
|
|
$unwanted_entries = $xpath->query('//'.$tag); |
|
84
|
|
|
foreach ($unwanted_entries as $unwanted_elem) { |
|
85
|
|
|
$unwanted_elem->parentNode->removeChild($unwanted_elem); |
|
86
|
|
|
} |
|
87
|
|
|
} |
|
88
|
|
|
|
|
89
|
|
|
$html = $this->normalizeBodyLinks($xmlDoc->saveHTML()); |
|
90
|
|
|
$html2 = preg_replace('@\s{2,}@', ' ', $html); //remove empty spaces from document |
|
91
|
|
|
|
|
92
|
|
|
return $html2; |
|
93
|
|
|
} |
|
94
|
|
|
|
|
95
|
|
|
/** |
|
96
|
|
|
* covert all relative paths in html to absolute ones |
|
97
|
|
|
* including: img src, a href ...etc. |
|
98
|
|
|
* @param string $html |
|
99
|
|
|
* @return string |
|
100
|
|
|
*/ |
|
101
|
|
|
public function normalizeBodyLinks($html) |
|
102
|
|
|
{ |
|
103
|
|
|
if (empty($html) === true) { //if html is empty, do nothing |
|
104
|
|
|
return $html; |
|
105
|
|
|
} |
|
106
|
|
|
|
|
107
|
|
|
$xmlDoc = new \DOMDocument(); |
|
108
|
|
|
libxml_use_internal_errors(true); |
|
109
|
|
|
$xmlDoc->loadHTML(mb_convert_encoding($html, 'HTML-ENTITIES', 'UTF-8')); |
|
110
|
|
|
libxml_clear_errors(); |
|
111
|
|
|
|
|
112
|
|
|
$xpath = new \DOMXPath($xmlDoc); |
|
113
|
|
|
$lnk_entries = $xpath->query('//a'); |
|
114
|
|
|
|
|
115
|
|
|
foreach ($lnk_entries as $entry) { |
|
116
|
|
|
$href = $entry->getAttribute('href'); |
|
117
|
|
|
$normalized_href = $this->normalizeLink($href); |
|
118
|
|
|
|
|
119
|
|
|
$entry->setAttribute('href', $normalized_href); |
|
120
|
|
|
$entry->setAttribute('target', '_blank'); |
|
121
|
|
|
} |
|
122
|
|
|
|
|
123
|
|
|
$img_entries = $xpath->query('//img'); |
|
124
|
|
|
|
|
125
|
|
|
foreach ($img_entries as $entry) { |
|
126
|
|
|
$src = $entry->getAttribute('src'); |
|
127
|
|
|
$normalized_src = $this->normalizeLink($src); |
|
128
|
|
|
$entry->setAttribute('src', $normalized_src); |
|
129
|
|
|
} |
|
130
|
|
|
|
|
131
|
|
|
$final_html = $xmlDoc->saveHTML(); |
|
132
|
|
|
|
|
133
|
|
|
return $this->getBodyHtml($final_html); |
|
134
|
|
|
} |
|
135
|
|
|
|
|
136
|
|
|
|
|
137
|
|
|
/** |
|
138
|
|
|
* normalize keywords by removing spaces from each |
|
139
|
|
|
* @param array $keywords |
|
140
|
|
|
* @return array |
|
141
|
|
|
*/ |
|
142
|
|
|
public function normalizeKeywords(array $keywords) |
|
143
|
|
|
{ |
|
144
|
|
|
foreach ($keywords as $k => $word) { |
|
145
|
|
|
$keywords[$k] = trim($word); |
|
146
|
|
|
} |
|
147
|
|
|
|
|
148
|
|
|
return $keywords; |
|
149
|
|
|
} |
|
150
|
|
|
|
|
151
|
|
|
/** |
|
152
|
|
|
* extract body content from html document |
|
153
|
|
|
* @param string $doc_html |
|
154
|
|
|
* @return string |
|
155
|
|
|
*/ |
|
156
|
|
|
protected function getBodyHtml($doc_html) |
|
157
|
|
|
{ |
|
158
|
|
|
$html_crawler = new Crawler($doc_html); |
|
159
|
|
|
|
|
160
|
|
|
$ret = ''; |
|
161
|
|
|
$html_crawler->filter('body')->each( |
|
162
|
|
|
function (Crawler $node) use (&$ret) { |
|
163
|
|
|
$ret = $node->html(); |
|
164
|
|
|
} |
|
165
|
|
|
); |
|
166
|
|
|
|
|
167
|
|
|
return $ret; |
|
168
|
|
|
} |
|
169
|
|
|
|
|
170
|
|
|
|
|
171
|
|
|
/** |
|
172
|
|
|
* extract image source by selector |
|
173
|
|
|
* @param Crawler $crawler |
|
174
|
|
|
* @param string $selector |
|
175
|
|
|
* @return string|NULL |
|
176
|
|
|
*/ |
|
177
|
|
|
protected function getSrcByImgSelector(Crawler $crawler, $selector) |
|
178
|
|
|
{ |
|
179
|
|
|
$ret = null; |
|
180
|
|
|
$imgExtractClosure = function (Crawler $node) use (&$ret) { |
|
181
|
|
|
$ret = $node->attr('src'); |
|
182
|
|
|
}; |
|
183
|
|
|
if (Selector::isXPath($selector)) { |
|
184
|
|
|
$crawler->filterXPath($selector) |
|
185
|
|
|
->each($imgExtractClosure); |
|
186
|
|
|
} else { |
|
187
|
|
|
$crawler->filter($selector) |
|
188
|
|
|
->each($imgExtractClosure); |
|
189
|
|
|
} |
|
190
|
|
|
|
|
191
|
|
|
if (empty($ret) === false) { |
|
192
|
|
|
return $this->normalizeLink($ret); |
|
193
|
|
|
} else { |
|
194
|
|
|
return null; |
|
195
|
|
|
} |
|
196
|
|
|
} |
|
197
|
|
|
} |
|
198
|
|
|
|
zrashwani /
news-scrapper
