diff options
Diffstat (limited to 'application/bookmark/LinkUtils.php')
-rw-r--r-- | application/bookmark/LinkUtils.php | 106 |
1 files changed, 0 insertions, 106 deletions
diff --git a/application/bookmark/LinkUtils.php b/application/bookmark/LinkUtils.php index 98d9038a..68914fca 100644 --- a/application/bookmark/LinkUtils.php +++ b/application/bookmark/LinkUtils.php | |||
@@ -3,112 +3,6 @@ | |||
3 | use Shaarli\Bookmark\Bookmark; | 3 | use Shaarli\Bookmark\Bookmark; |
4 | 4 | ||
5 | /** | 5 | /** |
6 | * Get cURL callback function for CURLOPT_WRITEFUNCTION | ||
7 | * | ||
8 | * @param string $charset to extract from the downloaded page (reference) | ||
9 | * @param string $title to extract from the downloaded page (reference) | ||
10 | * @param string $description to extract from the downloaded page (reference) | ||
11 | * @param string $keywords to extract from the downloaded page (reference) | ||
12 | * @param bool $retrieveDescription Automatically tries to retrieve description and keywords from HTML content | ||
13 | * @param string $curlGetInfo Optionally overrides curl_getinfo function | ||
14 | * | ||
15 | * @return Closure | ||
16 | */ | ||
17 | function get_curl_download_callback( | ||
18 | &$charset, | ||
19 | &$title, | ||
20 | &$description, | ||
21 | &$keywords, | ||
22 | $retrieveDescription, | ||
23 | $curlGetInfo = 'curl_getinfo' | ||
24 | ) { | ||
25 | $isRedirected = false; | ||
26 | $currentChunk = 0; | ||
27 | $foundChunk = null; | ||
28 | |||
29 | /** | ||
30 | * cURL callback function for CURLOPT_WRITEFUNCTION (called during the download). | ||
31 | * | ||
32 | * While downloading the remote page, we check that the HTTP code is 200 and content type is 'html/text' | ||
33 | * Then we extract the title and the charset and stop the download when it's done. | ||
34 | * | ||
35 | * @param resource $ch cURL resource | ||
36 | * @param string $data chunk of data being downloaded | ||
37 | * | ||
38 | * @return int|bool length of $data or false if we need to stop the download | ||
39 | */ | ||
40 | return function (&$ch, $data) use ( | ||
41 | $retrieveDescription, | ||
42 | $curlGetInfo, | ||
43 | &$charset, | ||
44 | &$title, | ||
45 | &$description, | ||
46 | &$keywords, | ||
47 | &$isRedirected, | ||
48 | &$currentChunk, | ||
49 | &$foundChunk | ||
50 | ) { | ||
51 | $currentChunk++; | ||
52 | $responseCode = $curlGetInfo($ch, CURLINFO_RESPONSE_CODE); | ||
53 | if (!empty($responseCode) && in_array($responseCode, [301, 302])) { | ||
54 | $isRedirected = true; | ||
55 | return strlen($data); | ||
56 | } | ||
57 | if (!empty($responseCode) && $responseCode !== 200) { | ||
58 | return false; | ||
59 | } | ||
60 | // After a redirection, the content type will keep the previous request value | ||
61 | // until it finds the next content-type header. | ||
62 | if (! $isRedirected || strpos(strtolower($data), 'content-type') !== false) { | ||
63 | $contentType = $curlGetInfo($ch, CURLINFO_CONTENT_TYPE); | ||
64 | } | ||
65 | if (!empty($contentType) && strpos($contentType, 'text/html') === false) { | ||
66 | return false; | ||
67 | } | ||
68 | if (!empty($contentType) && empty($charset)) { | ||
69 | $charset = header_extract_charset($contentType); | ||
70 | } | ||
71 | if (empty($charset)) { | ||
72 | $charset = html_extract_charset($data); | ||
73 | } | ||
74 | if (empty($title)) { | ||
75 | $title = html_extract_title($data); | ||
76 | $foundChunk = ! empty($title) ? $currentChunk : $foundChunk; | ||
77 | } | ||
78 | if ($retrieveDescription && empty($description)) { | ||
79 | $description = html_extract_tag('description', $data); | ||
80 | $foundChunk = ! empty($description) ? $currentChunk : $foundChunk; | ||
81 | } | ||
82 | if ($retrieveDescription && empty($keywords)) { | ||
83 | $keywords = html_extract_tag('keywords', $data); | ||
84 | if (! empty($keywords)) { | ||
85 | $foundChunk = $currentChunk; | ||
86 | // Keywords use the format tag1, tag2 multiple words, tag | ||
87 | // So we format them to match Shaarli's separator and glue multiple words with '-' | ||
88 | $keywords = implode(' ', array_map(function($keyword) { | ||
89 | return implode('-', preg_split('/\s+/', trim($keyword))); | ||
90 | }, explode(',', $keywords))); | ||
91 | } | ||
92 | } | ||
93 | |||
94 | // We got everything we want, stop the download. | ||
95 | // If we already found either the title, description or keywords, | ||
96 | // it's highly unlikely that we'll found the other metas further than | ||
97 | // in the same chunk of data or the next one. So we also stop the download after that. | ||
98 | if ((!empty($responseCode) && !empty($contentType) && !empty($charset)) && $foundChunk !== null | ||
99 | && (! $retrieveDescription | ||
100 | || $foundChunk < $currentChunk | ||
101 | || (!empty($title) && !empty($description) && !empty($keywords)) | ||
102 | ) | ||
103 | ) { | ||
104 | return false; | ||
105 | } | ||
106 | |||
107 | return strlen($data); | ||
108 | }; | ||
109 | } | ||
110 | |||
111 | /** | ||
112 | * Extract title from an HTML document. | 6 | * Extract title from an HTML document. |
113 | * | 7 | * |
114 | * @param string $html HTML content where to look for a title. | 8 | * @param string $html HTML content where to look for a title. |