|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
/* |
|
4
|
|
|
* This file is part of dispositif/wikibot application (@github) |
|
5
|
|
|
* 2019-2023 © Philippe M./Irønie <[email protected]> |
|
6
|
|
|
* For the full copyright and MIT license information, view the license file. |
|
7
|
|
|
*/ |
|
8
|
|
|
declare(strict_types=1); |
|
9
|
|
|
|
|
10
|
|
|
namespace App\Application\Http; |
|
11
|
|
|
|
|
12
|
|
|
use App\Domain\InfrastructurePorts\ExternHttpClientInterface; |
|
13
|
|
|
use DomainException; |
|
14
|
|
|
use GuzzleHttp\Client; |
|
15
|
|
|
use Normalizer; |
|
16
|
|
|
use Psr\Log\LoggerInterface; |
|
17
|
|
|
use Psr\Log\NullLogger; |
|
18
|
|
|
use Throwable; |
|
19
|
|
|
|
|
20
|
|
|
/** |
|
21
|
|
|
* Http client (Guzzle) configured for web crawling. |
|
22
|
|
|
*/ |
|
23
|
|
|
class ExternHttpClient implements ExternHttpClientInterface |
|
24
|
|
|
{ |
|
25
|
|
|
private readonly Client $client; |
|
26
|
|
|
|
|
27
|
|
|
public function __construct(private readonly LoggerInterface $log = new NullLogger()) |
|
28
|
|
|
{ |
|
29
|
|
|
$this->client = new Client( |
|
|
|
|
|
|
30
|
|
|
[ |
|
31
|
|
|
'timeout' => 30, |
|
32
|
|
|
'allow_redirects' => true, |
|
33
|
|
|
'headers' => ['User-Agent' => getenv('USER_AGENT')], |
|
34
|
|
|
'verify' => false, // CURLOPT_SSL_VERIFYHOST |
|
35
|
|
|
// 'proxy' => '192.192.192.192:10', |
|
36
|
|
|
] |
|
37
|
|
|
); |
|
38
|
|
|
} |
|
39
|
|
|
|
|
40
|
|
|
//hack for WikiwixAdapter todo : plutôt request() compatible ClientInterface ? |
|
41
|
|
|
public function getClient(): Client |
|
42
|
|
|
{ |
|
43
|
|
|
return $this->client; |
|
44
|
|
|
} |
|
45
|
|
|
|
|
46
|
|
|
/** |
|
47
|
|
|
* import source from URL with Guzzle. |
|
48
|
|
|
* todo abstract + refac async request |
|
49
|
|
|
*/ |
|
50
|
|
|
public function getHTML(string $url, ?bool $normalized = false): ?string |
|
51
|
|
|
{ |
|
52
|
|
|
// todo : check banned domains ? |
|
53
|
|
|
// todo : check DNS record => ban ? |
|
54
|
|
|
// todo : accept non-ascii URL ? |
|
55
|
|
|
// idn_to_ascii($url); |
|
56
|
|
|
// idn_to_ascii('teßt.com',IDNA_NONTRANSITIONAL_TO_ASCII,INTL_IDNA_VARIANT_UTS46) |
|
57
|
|
|
// checkdnsrr($string, "A") // check DNS record |
|
58
|
|
|
if (!self::isHttpURL($url)) { |
|
59
|
|
|
throw new DomainException('URL not compatible : ' . $url); |
|
60
|
|
|
} |
|
61
|
|
|
$response = $this->client->get($url); |
|
62
|
|
|
|
|
63
|
|
|
if (200 !== $response->getStatusCode()) { |
|
64
|
|
|
echo 'HTTP error ' . $response->getStatusCode(); |
|
65
|
|
|
$this->log->error('HTTP error ' . $response->getStatusCode() . ' ' . $response->getReasonPhrase()); |
|
66
|
|
|
|
|
67
|
|
|
return null; |
|
68
|
|
|
} |
|
69
|
|
|
|
|
70
|
|
|
$html = $response->getBody()->getContents() ?? ''; |
|
71
|
|
|
|
|
72
|
|
|
return ($normalized) ? $this->normalizeHtml($html, $url) : $html; |
|
73
|
6 |
|
} |
|
74
|
|
|
|
|
75
|
6 |
|
/** |
|
76
|
1 |
|
* Todo Move HttpUtils ? |
|
77
|
|
|
* Better than filter_var($url, FILTER_VALIDATE_URL) because it's not multibyte capable. |
|
78
|
|
|
* See for example .中国 domain name |
|
79
|
6 |
|
*/ |
|
80
|
|
|
public static function isHttpURL(string $url): bool |
|
81
|
|
|
{ |
|
82
|
|
|
return (bool)preg_match('#^https?://[^ ]+$#i', $url); |
|
83
|
|
|
} |
|
84
|
|
|
|
|
85
|
|
|
/** |
|
86
|
|
|
* Normalize and converting to UTF-8 encoding |
|
87
|
|
|
*/ |
|
88
|
|
|
private function normalizeHtml(string $html, ?string $url = ''): ?string |
|
89
|
|
|
{ |
|
90
|
|
|
$e = null; |
|
91
|
|
|
if (empty($html)) { |
|
92
|
|
|
return $html; |
|
93
|
|
|
} |
|
94
|
|
|
|
|
95
|
|
|
$html2 = Normalizer::normalize($html); |
|
96
|
|
|
|
|
97
|
|
|
if (is_string($html2) && !empty($html2)) { |
|
98
|
|
|
return $html2; |
|
99
|
|
|
} |
|
100
|
|
|
|
|
101
|
|
|
$charset = $this->extractCharset($html) ?? 'WINDOWS-1252'; |
|
102
|
|
|
|
|
103
|
|
|
if (empty($charset)) { |
|
104
|
|
|
throw new DomainException('normalized html error and no charset found : ' . $url); |
|
105
|
|
|
} |
|
106
|
|
|
try { |
|
107
|
|
|
$html2 = iconv($charset, 'UTF-8//TRANSLIT', $html); |
|
108
|
|
|
if (false === $html2) { |
|
109
|
|
|
throw new DomainException("error iconv : $charset to UTF-8 on " . $url); |
|
110
|
|
|
} |
|
111
|
|
|
$html2 = Normalizer::normalize($html2); |
|
112
|
|
|
if (!is_string($html2)) { |
|
|
|
|
|
|
113
|
|
|
throw new DomainException("error normalizer : $charset to UTF-8 on " . $url); |
|
114
|
|
|
} |
|
115
|
|
|
} catch (Throwable $e) { |
|
116
|
|
|
throw new DomainException("error converting : $charset to UTF-8 on " . $url, $e->getCode(), $e); |
|
117
|
|
|
} |
|
118
|
|
|
|
|
119
|
|
|
return $html2; |
|
120
|
|
|
} |
|
121
|
|
|
|
|
122
|
|
|
/** |
|
123
|
|
|
* Extract charset from HTML text |
|
124
|
|
|
*/ |
|
125
|
|
|
private function extractCharset(string $html): ?string |
|
126
|
|
|
{ |
|
127
|
|
|
if (preg_match( |
|
128
|
|
|
'#<meta(?!\s*(?:name|value)\s*=)(?:[^>]*?content\s*=[\s"\']*)?([^>]*?)[\s"\';]*charset\s*=[\s"\']*([^\s"\'/>]*)#', |
|
129
|
|
|
$html, |
|
130
|
|
|
$matches |
|
131
|
|
|
) |
|
132
|
|
|
) { |
|
133
|
|
|
$charset = $matches[2] ?? $matches[1] ?? null; |
|
134
|
|
|
} |
|
135
|
|
|
if (empty($charset)) { |
|
136
|
|
|
|
|
137
|
|
|
$encoding = mb_detect_encoding($html, null, true); |
|
138
|
|
|
$charset = is_string($encoding) ? strtoupper($encoding) : null; |
|
|
|
|
|
|
139
|
|
|
} |
|
140
|
|
|
|
|
141
|
|
|
return $charset; |
|
142
|
|
|
} |
|
143
|
|
|
} |
|
144
|
|
|
|
Dispositif /
Wikibot
Loading history...