| 1 |
lars |
1 |
<?php
|
|
|
2 |
|
|
|
3 |
require_once(HTML2PS_DIR.'fetcher._interface.class.php');
|
|
|
4 |
|
|
|
5 |
/**
|
|
|
6 |
* This class handles fetching HTTP code using CURL extension
|
|
|
7 |
*/
|
|
|
8 |
class FetcherUrlCurl extends Fetcher {
|
|
|
9 |
/**
|
|
|
10 |
* @var String URL being fetched
|
|
|
11 |
* @access private
|
|
|
12 |
*/
|
|
|
13 |
var $url;
|
|
|
14 |
var $_proxy;
|
|
|
15 |
|
|
|
16 |
function FetcherUrlCurl() {
|
|
|
17 |
$this->url = "";
|
|
|
18 |
$this->set_proxy(null);
|
|
|
19 |
}
|
|
|
20 |
|
|
|
21 |
function _fix_url($url) {
|
|
|
22 |
// If only host name was specified, add trailing slash
|
|
|
23 |
// (e.g. replace http://www.google.com with http://www.google.com/
|
|
|
24 |
if (preg_match('#^.*://[^/]+$#', $url)) {
|
|
|
25 |
$url .= '/';
|
|
|
26 |
};
|
|
|
27 |
|
|
|
28 |
return $url;
|
|
|
29 |
}
|
|
|
30 |
|
|
|
31 |
function get_base_url() {
|
|
|
32 |
return $this->url;
|
|
|
33 |
}
|
|
|
34 |
|
|
|
35 |
function get_data($url) {
|
|
|
36 |
$this->url = $url;
|
|
|
37 |
|
|
|
38 |
// URL to be fetched
|
|
|
39 |
$curl = curl_init();
|
|
|
40 |
|
|
|
41 |
$fixed_url = $this->_fix_url($url);
|
|
|
42 |
|
|
|
43 |
curl_setopt($curl, CURLOPT_URL, $fixed_url);
|
|
|
44 |
curl_setopt($curl, CURLOPT_USERAGENT, DEFAULT_USER_AGENT);
|
|
|
45 |
|
|
|
46 |
if (!@curl_setopt($curl, CURLOPT_FOLLOWLOCATION, 1)) {
|
|
|
47 |
error_log('CURLOPT_FOLLOWLOCATION will not work in safe_mode; pages with redirects may be rendered incorrectly');
|
|
|
48 |
};
|
|
|
49 |
|
|
|
50 |
curl_setopt($curl, CURLOPT_HEADER, 1);
|
|
|
51 |
curl_setopt($curl, CURLOPT_RETURNTRANSFER, 1);
|
|
|
52 |
|
|
|
53 |
$proxy = $this->get_proxy();
|
|
|
54 |
if (!is_null($proxy)) {
|
|
|
55 |
curl_setopt($curl, CURLOPT_PROXY, $proxy);
|
|
|
56 |
};
|
|
|
57 |
|
|
|
58 |
/**
|
|
|
59 |
* Fetch headers and page content to the $response variable
|
|
|
60 |
* and close CURL session
|
|
|
61 |
*/
|
|
|
62 |
$response = curl_exec($curl);
|
|
|
63 |
|
|
|
64 |
if ($response === FALSE) {
|
|
|
65 |
error_log(sprintf('Cannot open %s, CURL error is: %s',
|
|
|
66 |
$url,
|
|
|
67 |
curl_error($curl)));
|
|
|
68 |
curl_close($curl);
|
|
|
69 |
return null;
|
|
|
70 |
}
|
|
|
71 |
|
|
|
72 |
curl_close($curl);
|
|
|
73 |
|
|
|
74 |
/**
|
|
|
75 |
* According to HTTP standard, headers block separated from
|
|
|
76 |
* body block with empty line - '\r\n\r\n' sequence. As body
|
|
|
77 |
* might contain this sequence too, we should use 'non-greedy'
|
|
|
78 |
* modifier on the first group in the regular expression.
|
|
|
79 |
* Of course, we should process the response as a whole using
|
|
|
80 |
* 's' modifier.
|
|
|
81 |
*/
|
|
|
82 |
preg_match('/^(.*?)\r\n\r\n(.*)$/s', $response, $matches);
|
|
|
83 |
|
|
|
84 |
/**
|
|
|
85 |
* Usually there's more than one line in a header block,
|
|
|
86 |
* separated with '\r\n' sequence.
|
|
|
87 |
*
|
|
|
88 |
* The very first line contains HTTP response code (e.g. HTTP/1.1 200 OK),
|
|
|
89 |
* so we may safely ignore it.
|
|
|
90 |
*/
|
|
|
91 |
$headers = array_slice(explode("\r\n", $matches[1]),1);
|
|
|
92 |
$content = $matches[2];
|
|
|
93 |
|
|
|
94 |
return new FetchedDataURL($content, $headers, $this->url);
|
|
|
95 |
}
|
|
|
96 |
|
|
|
97 |
function get_proxy() {
|
|
|
98 |
return $this->_proxy;
|
|
|
99 |
}
|
|
|
100 |
|
|
|
101 |
function set_proxy($proxy) {
|
|
|
102 |
$this->_proxy = $proxy;
|
|
|
103 |
}
|
|
|
104 |
}
|
|
|
105 |
?>
|