3 if (!defined('GNUSOCIAL')) { exit(1); }
5 // FIXME: To support remote video/whatever files, this plugin needs reworking.
7 class StoreRemoteMediaPlugin extends Plugin
9 // settings which can be set in config.php with addPlugin('Oembed', array('param'=>'value', ...));
10 // WARNING, these are _regexps_ (slashes added later). Always escape your dots and end your strings
11 public $domain_whitelist = array( // hostname => service provider
12 '^i\d*\.ytimg\.com$' => 'YouTube',
13 '^i\d*\.vimeocdn\.com$' => 'Vimeo',
15 public $append_whitelist = array(); // fill this array as domain_whitelist to add more trusted sources
16 public $check_whitelist = false; // security/abuse precaution
18 public $domain_blacklist = array();
19 public $check_blacklist = false;
21 public $max_image_bytes = 10485760; // 10MiB max image size by default
23 protected $imgData = array();
25 // these should be declared protected everywhere
26 public function initialize()
30 $this->domain_whitelist = array_merge($this->domain_whitelist, $this->append_whitelist);
34 * Save embedding information for a File, if applicable.
36 * Normally this event is called through File::saveNew()
38 * @param File $file The abount-to-be-inserted File object.
40 * @return boolean success
42 public function onStartFileSaveNew(File &$file)
44 // save given URL as title if it's a media file this plugin understands
45 // which will make it shown in the AttachmentList widgets
47 if (isset($file->title) && strlen($file->title)>0) {
48 // Title is already set
51 if (!isset($file->mimetype)) {
52 // Unknown mimetype, it's not our job to figure out what it is.
55 switch (common_get_mime_media($file->mimetype)) {
57 // Just to set something for now at least...
58 //$file->title = $file->mimetype;
65 public function onCreateFileImageThumbnailSource(File $file, &$imgPath, $media=null)
67 // If we are on a private node, we won't do any remote calls (just as a precaution until
68 // we can configure this from config.php for the private nodes)
69 if (common_config('site', 'private')) {
73 if ($media !== 'image') {
77 // If there is a local filename, it is either a local file already or has already been downloaded.
78 if (!empty($file->filename)) {
82 $remoteUrl = $file->getUrl();
84 if (!$this->checkWhiteList($remoteUrl) ||
85 !$this->checkBlackList($remoteUrl)) {
91 $http = new HTTPClient();
92 common_debug(sprintf('Performing HEAD request for remote file id==%u to avoid unnecessarily downloading too large files. URL: %s', $file->getID(), $remoteUrl));
93 $head = $http->head($remoteUrl);
94 $remoteUrl = $head->getEffectiveUrl(); // to avoid going through redirects again
95 if (!$this->checkBlackList($remoteUrl)) {
96 common_log(LOG_WARN, sprintf('%s: Non-blacklisted URL %s redirected to blacklisted URL %s', __CLASS__, $file->getUrl(), $remoteUrl));
100 $headers = $head->getHeader();
101 $filesize = isset($headers['content-length']) ? $headers['content-length'] : null;
103 $filesize = $file->getSize();
104 if (empty($filesize)) {
105 // file size not specified on remote server
106 common_debug(sprintf('%s: Ignoring remote media because we did not get a content length for file id==%u', __CLASS__, $file->getID()));
108 } elseif ($filesize > $this->max_image_bytes) {
109 //FIXME: When we perhaps start fetching videos etc. we'll need to differentiate max_image_bytes from that...
110 // file too big according to plugin configuration
111 common_debug(sprintf('%s: Skipping remote media because content length (%u) is larger than plugin configured max_image_bytes (%u) for file id==%u', __CLASS__, intval($filesize), $this->max_image_bytes, $file->getID()));
113 } elseif ($filesize > common_config('attachments', 'file_quota')) {
114 // file too big according to site configuration
115 common_debug(sprintf('%s: Skipping remote media because content length (%u) is larger than file_quota (%u) for file id==%u', __CLASS__, intval($filesize), common_config('attachments', 'file_quota'), $file->getID()));
119 // Then we download the file to memory and test whether it's actually an image file
120 common_debug(sprintf('Downloading remote file id==%u (should be size %u) with effective URL: %s', $file->getID(), $filesize, _ve($remoteUrl)));
121 $imgData = HTTPClient::quickGet($remoteUrl);
122 } catch (HTTP_Request2_ConnectionException $e) {
123 common_log(LOG_ERR, __CLASS__.': '._ve(get_class($e)).' on URL: '._ve($file->getUrl()).' threw exception: '.$e->getMessage());
126 $info = @getimagesizefromstring($imgData);
127 if ($info === false) {
128 throw new UnsupportedMediaException(_('Remote file format was not identified as an image.'), $remoteUrl);
129 } elseif (!$info[0] || !$info[1]) {
130 throw new UnsupportedMediaException(_('Image file had impossible geometry (0 width or height)'));
133 $filehash = hash(File::FILEHASH_ALG, $imgData);
135 // Exception will be thrown before $file is set to anything, so old $file value will be kept
136 $file = File::getByHash($filehash);
138 //FIXME: Add some code so we don't have to store duplicate File rows for same hash files.
139 } catch (NoResultException $e) {
140 $filename = $filehash . '.' . common_supported_mime_to_ext($info['mime']);
141 $fullpath = File::path($filename);
143 // Write the file to disk if it doesn't exist yet. Throw Exception on failure.
144 if (!file_exists($fullpath) && file_put_contents($fullpath, $imgData) === false) {
145 throw new ServerException(_('Could not write downloaded file to disk.'));
148 // Updated our database for the file record
149 $orig = clone($file);
150 $file->filehash = $filehash;
151 $file->filename = $filename;
152 $file->width = $info[0]; // array indexes documented on php.net:
153 $file->height = $info[1]; // https://php.net/manual/en/function.getimagesize.php
154 // Throws exception on failure.
155 $file->updateWithKeys($orig);
157 // Get rid of the file from memory
160 $imgPath = $file->getPath();
166 * @return boolean true if given url passes blacklist check
168 protected function checkBlackList($url)
170 if (!$this->check_blacklist) {
173 $host = parse_url($url, PHP_URL_HOST);
174 foreach ($this->domain_blacklist as $regex => $provider) {
175 if (preg_match("/$regex/", $host)) {
184 * @return boolean true if given url passes whitelist check
186 protected function checkWhiteList($url)
188 if (!$this->check_whitelist) {
191 $host = parse_url($url, PHP_URL_HOST);
192 foreach ($this->domain_whitelist as $regex => $provider) {
193 if (preg_match("/$regex/", $host)) {
201 public function onPluginVersion(array &$versions)
203 $versions[] = array('name' => 'StoreRemoteMedia',
204 'version' => GNUSOCIAL_VERSION,
205 'author' => 'Mikael Nordfeldth',
206 'homepage' => 'https://gnu.io/',
208 // TRANS: Plugin description.
209 _m('Plugin for downloading remotely attached files to local server.'));