3 * StatusNet - the distributed open-source microblogging tool
4 * Copyright (C) 2008, 2009, StatusNet, Inc.
6 * This program is free software: you can redistribute it and/or modify
7 * it under the terms of the GNU Affero General Public License as published by
8 * the Free Software Foundation, either version 3 of the License, or
9 * (at your option) any later version.
11 * This program is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 * GNU Affero General Public License for more details.
16 * You should have received a copy of the GNU Affero General Public License
17 * along with this program. If not, see <http://www.gnu.org/licenses/>.
20 if (!defined('GNUSOCIAL')) { exit(1); }
23 * Table Definition for file
25 class File extends Managed_DataObject
27 public $__table = 'file'; // table name
28 public $id; // int(4) primary_key not_null
29 public $urlhash; // varchar(64) unique_key
31 public $filehash; // varchar(64) indexed
32 public $mimetype; // varchar(50)
33 public $size; // int(4)
34 public $title; // text()
35 public $date; // int(4)
36 public $protected; // int(4)
37 public $filename; // text()
38 public $width; // int(4)
39 public $height; // int(4)
40 public $modified; // timestamp() not_null default_CURRENT_TIMESTAMP
42 const URLHASH_ALG = 'sha256';
43 const FILEHASH_ALG = 'sha256';
45 public static function schemaDef()
49 'id' => array('type' => 'serial', 'not null' => true),
50 'urlhash' => array('type' => 'varchar', 'length' => 64, 'not null' => true, 'description' => 'sha256 of destination URL (url field)'),
51 'url' => array('type' => 'text', 'description' => 'destination URL after following possible redirections'),
52 'filehash' => array('type' => 'varchar', 'length' => 64, 'not null' => false, 'description' => 'sha256 of the file contents, only for locally stored files of course'),
53 'mimetype' => array('type' => 'varchar', 'length' => 50, 'description' => 'mime type of resource'),
54 'size' => array('type' => 'int', 'description' => 'size of resource when available'),
55 'title' => array('type' => 'text', 'description' => 'title of resource when available'),
56 'date' => array('type' => 'int', 'description' => 'date of resource according to http query'),
57 'protected' => array('type' => 'int', 'description' => 'true when URL is private (needs login)'),
58 'filename' => array('type' => 'text', 'description' => 'if file is stored locally (too) this is the filename'),
59 'width' => array('type' => 'int', 'description' => 'width in pixels, if it can be described as such and data is available'),
60 'height' => array('type' => 'int', 'description' => 'height in pixels, if it can be described as such and data is available'),
62 'modified' => array('type' => 'timestamp', 'not null' => true, 'description' => 'date this record was modified'),
64 'primary key' => array('id'),
65 'unique keys' => array(
66 'file_urlhash_key' => array('urlhash'),
69 'file_filehash_idx' => array('filehash'),
74 function isProtected($url) {
75 return 'http://www.facebook.com/login.php' === $url;
79 * Save a new file record.
81 * @param array $redir_data lookup data eg from File_redirection::where()
82 * @param string $given_url
85 public static function saveNew(array $redir_data, $given_url)
89 // I don't know why we have to keep doing this but we run a last check to avoid
91 $file = File::getByUrl($given_url);
93 } catch (NoResultException $e) {
94 // We don't have the file's URL since before, so let's continue.
98 $file->url = $given_url;
99 if (!empty($redir_data['protected'])) $file->protected = $redir_data['protected'];
100 if (!empty($redir_data['title'])) $file->title = $redir_data['title'];
101 if (!empty($redir_data['type'])) $file->mimetype = $redir_data['type'];
102 if (!empty($redir_data['size'])) $file->size = intval($redir_data['size']);
103 if (isset($redir_data['time']) && $redir_data['time'] > 0) $file->date = intval($redir_data['time']);
108 public function saveFile() {
109 $this->urlhash = self::hashurl($this->url);
111 if (!Event::handle('StartFileSaveNew', array(&$this))) {
112 throw new ServerException('File not saved due to an aborted StartFileSaveNew event.');
115 $this->id = $this->insert();
117 if ($this->id === false) {
118 throw new ServerException('File/URL metadata could not be saved to the database.');
121 Event::handle('EndFileSaveNew', array($this));
125 * Go look at a URL and possibly save data about it if it's new:
126 * - follow redirect chains and store them in file_redirection
127 * - if a thumbnail is available, save it in file_thumbnail
128 * - save file record with basic info
129 * - optionally save a file_to_post record
130 * - return the File object with the full reference
132 * @param string $given_url the URL we're looking at
133 * @param Notice $notice (optional)
134 * @param bool $followRedirects defaults to true
136 * @return mixed File on success, -1 on some errors
138 * @throws ServerException on failure
140 public static function processNew($given_url, Notice $notice=null, $followRedirects=true) {
141 if (empty($given_url)) {
142 throw new ServerException('No given URL to process');
145 $given_url = File_redirection::_canonUrl($given_url);
146 if (empty($given_url)) {
147 throw new ServerException('No canonical URL from given URL to process');
150 $redir = File_redirection::where($given_url);
151 $file = $redir->getFile();
153 if (!$file instanceof File || empty($file->id)) {
154 // This should not happen
155 throw new ServerException('URL processing failed without new File object');
158 if ($notice instanceof Notice) {
159 File_to_post::processNew($file, $notice);
165 public static function respectsQuota(Profile $scoped, $fileSize) {
166 if ($fileSize > common_config('attachments', 'file_quota')) {
167 // TRANS: Message used to be inserted as %2$s in the text "No file may
168 // TRANS: be larger than %1$d byte and the file you sent was %2$s.".
169 // TRANS: %1$d is the number of bytes of an uploaded file.
170 $fileSizeText = sprintf(_m('%1$d byte','%1$d bytes',$fileSize),$fileSize);
172 $fileQuota = common_config('attachments', 'file_quota');
173 // TRANS: Message given if an upload is larger than the configured maximum.
174 // TRANS: %1$d (used for plural) is the byte limit for uploads,
175 // TRANS: %2$s is the proper form of "n bytes". This is the only ways to have
176 // TRANS: gettext support multiple plurals in the same message, unfortunately...
177 throw new ClientException(
178 sprintf(_m('No file may be larger than %1$d byte and the file you sent was %2$s. Try to upload a smaller version.',
179 'No file may be larger than %1$d bytes and the file you sent was %2$s. Try to upload a smaller version.',
181 $fileQuota, $fileSizeText));
186 $query = "select sum(size) as total from file join file_to_post on file_to_post.file_id = file.id join notice on file_to_post.post_id = notice.id where profile_id = {$scoped->id} and file.url like '%/notice/%/file'";
187 $file->query($query);
189 $total = $file->total + $fileSize;
190 if ($total > common_config('attachments', 'user_quota')) {
191 // TRANS: Message given if an upload would exceed user quota.
192 // TRANS: %d (number) is the user quota in bytes and is used for plural.
193 throw new ClientException(
194 sprintf(_m('A file this large would exceed your user quota of %d byte.',
195 'A file this large would exceed your user quota of %d bytes.',
196 common_config('attachments', 'user_quota')),
197 common_config('attachments', 'user_quota')));
199 $query .= ' AND EXTRACT(month FROM file.modified) = EXTRACT(month FROM now()) and EXTRACT(year FROM file.modified) = EXTRACT(year FROM now())';
200 $file->query($query);
202 $total = $file->total + $fileSize;
203 if ($total > common_config('attachments', 'monthly_quota')) {
204 // TRANS: Message given id an upload would exceed a user's monthly quota.
205 // TRANS: $d (number) is the monthly user quota in bytes and is used for plural.
206 throw new ClientException(
207 sprintf(_m('A file this large would exceed your monthly quota of %d byte.',
208 'A file this large would exceed your monthly quota of %d bytes.',
209 common_config('attachments', 'monthly_quota')),
210 common_config('attachments', 'monthly_quota')));
215 public function getFilename()
217 if (!self::validFilename($this->filename)) {
218 // TRANS: Client exception thrown if a file upload does not have a valid name.
219 throw new ClientException(_("Invalid filename."));
221 return $this->filename;
224 // where should the file go?
226 static function filename(Profile $profile, $origname, $mimetype)
228 $ext = self::guessMimeExtension($mimetype);
230 // Normalize and make the original filename more URL friendly.
231 $origname = basename($origname, ".$ext");
232 if (class_exists('Normalizer')) {
233 // http://php.net/manual/en/class.normalizer.php
234 // http://www.unicode.org/reports/tr15/
235 $origname = Normalizer::normalize($origname, Normalizer::FORM_KC);
237 $origname = preg_replace('/[^A-Za-z0-9\.\_]/', '_', $origname);
239 $nickname = $profile->getNickname();
240 $datestamp = strftime('%Y%m%d', time());
242 // generate new random strings until we don't run into a filename collision.
243 $random = strtolower(common_confirmation_code(16));
244 $filename = "$nickname-$datestamp-$origname-$random.$ext";
245 } while (file_exists(self::path($filename)));
249 static function guessMimeExtension($mimetype)
252 $ext = common_supported_mime_to_ext($mimetype);
253 } catch (Exception $e) {
254 // We don't support this mimetype, but let's guess the extension
256 if (!preg_match('/\/([a-z0-9]+)/', mb_strtolower($mimetype), $matches)) {
257 throw new Exception('Malformed mimetype: '.$mimetype);
261 return mb_strtolower($ext);
265 * Validation for as-saved base filenames
267 static function validFilename($filename)
269 return preg_match('/^[A-Za-z0-9._-]+$/', $filename);
273 * @throws ClientException on invalid filename
275 static function path($filename)
277 if (!self::validFilename($filename)) {
278 // TRANS: Client exception thrown if a file upload does not have a valid name.
279 throw new ClientException(_("Invalid filename."));
281 $dir = common_config('attachments', 'dir');
283 if ($dir[strlen($dir)-1] != '/') {
287 return $dir . $filename;
290 static function url($filename)
292 if (!self::validFilename($filename)) {
293 // TRANS: Client exception thrown if a file upload does not have a valid name.
294 throw new ClientException(_("Invalid filename."));
297 if (common_config('site','private')) {
299 return common_local_url('getfile',
300 array('filename' => $filename));
304 if (GNUsocial::useHTTPS()) {
306 $sslserver = common_config('attachments', 'sslserver');
308 if (empty($sslserver)) {
309 // XXX: this assumes that background dir == site dir + /file/
310 // not true if there's another server
311 if (is_string(common_config('site', 'sslserver')) &&
312 mb_strlen(common_config('site', 'sslserver')) > 0) {
313 $server = common_config('site', 'sslserver');
314 } else if (common_config('site', 'server')) {
315 $server = common_config('site', 'server');
317 $path = common_config('site', 'path') . '/file/';
319 $server = $sslserver;
320 $path = common_config('attachments', 'sslpath');
322 $path = common_config('attachments', 'path');
328 $path = common_config('attachments', 'path');
329 $server = common_config('attachments', 'server');
331 if (empty($server)) {
332 $server = common_config('site', 'server');
335 $ssl = common_config('attachments', 'ssl');
337 $protocol = ($ssl) ? 'https' : 'http';
340 if ($path[strlen($path)-1] != '/') {
344 if ($path[0] != '/') {
348 return $protocol.'://'.$server.$path.$filename;
351 static $_enclosures = array();
353 function getEnclosure(){
354 if (isset(self::$_enclosures[$this->getID()])) {
355 return self::$_enclosures[$this->getID()];
358 $enclosure = (object) array();
359 foreach (array('title', 'url', 'date', 'modified', 'size', 'mimetype', 'width', 'height') as $key) {
360 if ($this->$key !== '') {
361 $enclosure->$key = $this->$key;
365 $needMoreMetadataMimetypes = array(null, 'application/xhtml+xml', 'text/html');
367 if (!isset($this->filename) && in_array(common_bare_mime($enclosure->mimetype), $needMoreMetadataMimetypes)) {
368 // This fetches enclosure metadata for non-local links with unset/HTML mimetypes,
369 // which may be enriched through oEmbed or similar (implemented as plugins)
370 Event::handle('FileEnclosureMetadata', array($this, &$enclosure));
372 if (empty($enclosure->mimetype)) {
373 // This means we either don't know what it is, so it can't
374 // be shown as an enclosure, or it is an HTML link which
375 // does not link to a resource with further metadata.
376 throw new ServerException('Unknown enclosure mimetype, not enough metadata');
379 self::$_enclosures[$this->getID()] = $enclosure;
383 public function hasThumbnail()
386 $this->getThumbnail();
387 } catch (Exception $e) {
394 * Get the attachment's thumbnail record, if any.
395 * Make sure you supply proper 'int' typed variables (or null).
397 * @param $width int Max width of thumbnail in pixels. (if null, use common_config values)
398 * @param $height int Max height of thumbnail in pixels. (if null, square-crop to $width)
399 * @param $crop bool Crop to the max-values' aspect ratio
401 * @return File_thumbnail
403 * @throws UseFileAsThumbnailException if the file is considered an image itself and should be itself as thumbnail
404 * @throws UnsupportedMediaException if, despite trying, we can't understand how to make a thumbnail for this format
405 * @throws ServerException on various other errors
407 public function getThumbnail($width=null, $height=null, $crop=false, $force_still=true)
409 // Get some more information about this file through our ImageFile class
410 $image = ImageFile::fromFileObject($this);
411 if ($image->animated && !common_config('thumbnail', 'animated')) {
412 // null means "always use file as thumbnail"
413 // false means you get choice between frozen frame or original when calling getThumbnail
414 if (is_null(common_config('thumbnail', 'animated')) || !$force_still) {
415 throw new UseFileAsThumbnailException($this->id);
419 return $image->getFileThumbnail($width, $height, $crop);
422 public function getPath()
424 $filepath = self::path($this->filename);
425 if (!file_exists($filepath)) {
426 throw new FileNotFoundException($filepath);
431 public function getUrl($prefer_local=true)
433 if ($prefer_local && !empty($this->filename)) {
434 // A locally stored file, so let's generate a URL for our instance.
435 return self::url($this->filename);
438 // No local filename available, return the URL we have stored
442 static public function getByUrl($url)
445 $file->urlhash = self::hashurl($url);
446 if (!$file->find(true)) {
447 throw new NoResultException($file);
453 * @param string $hashstr String of (preferrably lower case) hexadecimal characters, same as result of 'hash_file(...)'
455 static public function getByHash($hashstr)
458 $file->filehash = strtolower($hashstr);
459 if (!$file->find(true)) {
460 throw new NoResultException($file);
465 public function updateUrl($url)
467 $file = File::getKV('urlhash', self::hashurl($url));
468 if ($file instanceof File) {
469 throw new ServerException('URL already exists in DB');
471 $sql = 'UPDATE %1$s SET urlhash=%2$s, url=%3$s WHERE urlhash=%4$s;';
472 $result = $this->query(sprintf($sql, $this->tableName(),
473 $this->_quote((string)self::hashurl($url)),
474 $this->_quote((string)$url),
475 $this->_quote((string)$this->urlhash)));
476 if ($result === false) {
477 common_log_db_error($this, 'UPDATE', __FILE__);
478 throw new ServerException("Could not UPDATE {$this->tableName()}.url");
485 * Blow the cache of notices that link to this URL
487 * @param boolean $last Whether to blow the "last" cache too
492 function blowCache($last=false)
494 self::blow('file:notice-ids:%s', $this->id);
496 self::blow('file:notice-ids:%s;last', $this->id);
498 self::blow('file:notice-count:%d', $this->id);
502 * Stream of notices linking to this URL
504 * @param integer $offset Offset to show; default is 0
505 * @param integer $limit Limit of notices to show
506 * @param integer $since_id Since this notice
507 * @param integer $max_id Before this notice
509 * @return array ids of notices that link to this file
512 function stream($offset=0, $limit=NOTICES_PER_PAGE, $since_id=0, $max_id=0)
514 $stream = new FileNoticeStream($this);
515 return $stream->getNotices($offset, $limit, $since_id, $max_id);
518 function noticeCount()
520 $cacheKey = sprintf('file:notice-count:%d', $this->id);
522 $count = self::cacheGet($cacheKey);
524 if ($count === false) {
526 $f2p = new File_to_post();
528 $f2p->file_id = $this->id;
530 $count = $f2p->count();
532 self::cacheSet($cacheKey, $count);
538 public function isLocal()
540 return !empty($this->filename);
543 public function delete($useWhere=false)
545 // Delete the file, if it exists locally
546 if (!empty($this->filename) && file_exists(self::path($this->filename))) {
547 $deleted = @unlink(self::path($this->filename));
549 common_log(LOG_ERR, sprintf('Could not unlink existing file: "%s"', self::path($this->filename)));
553 // Clear out related things in the database and filesystem, such as thumbnails
554 if (Event::handle('FileDeleteRelated', array($this))) {
555 $thumbs = new File_thumbnail();
556 $thumbs->file_id = $this->id;
557 if ($thumbs->find()) {
558 while ($thumbs->fetch()) {
563 $f2p = new File_to_post();
564 $f2p->file_id = $this->id;
566 while ($f2p->fetch()) {
572 // And finally remove the entry from the database
573 return parent::delete($useWhere);
576 public function getTitle()
578 $title = $this->title ?: $this->filename;
580 return $title ?: null;
583 static public function hashurl($url)
586 throw new Exception('No URL provided to hash algorithm.');
588 return hash(self::URLHASH_ALG, $url);
591 static public function beforeSchemaUpdate()
593 $table = strtolower(get_called_class());
594 $schema = Schema::get();
595 $schemadef = $schema->getTableDef($table);
597 // 2015-02-19 We have to upgrade our table definitions to have the urlhash field populated
598 if (isset($schemadef['fields']['urlhash']) && isset($schemadef['unique keys']['file_urlhash_key'])) {
599 // We already have the urlhash field, so no need to migrate it.
602 echo "\nFound old $table table, upgrading it to contain 'urlhash' field...";
605 $file->query(sprintf('SELECT id, LEFT(url, 191) AS shortenedurl, COUNT(*) AS c FROM %1$s WHERE LENGTH(url)>191 GROUP BY shortenedurl HAVING c > 1', $schema->quoteIdentifier($table)));
606 print "\nFound {$file->N} URLs with too long entries in file table\n";
607 while ($file->fetch()) {
608 // We've got a URL that is too long for our future file table
609 // so we'll cut it. We could save the original URL, but there is
610 // no guarantee it is complete anyway since the previous max was 255 chars.
611 $dupfile = new File();
612 // First we find file entries that would be duplicates of this when shortened
613 // ... and we'll just throw the dupes out the window for now! It's already so borken.
614 $dupfile->query(sprintf('SELECT * FROM file WHERE LEFT(url, 191) = "%1$s"', $file->shortenedurl));
615 // Leave one of the URLs in the database by using ->find(true) (fetches first entry)
616 if ($dupfile->find(true)) {
617 print "\nShortening url entry for $table id: {$file->id} [";
618 $orig = clone($dupfile);
619 $dupfile->url = $file->shortenedurl; // make sure it's only 191 chars from now on
620 $dupfile->update($orig);
621 print "\nDeleting duplicate entries of too long URL on $table id: {$file->id} [";
622 // only start deleting with this fetch.
623 while($dupfile->fetch()) {
629 print "\nWarning! URL suddenly disappeared from database: {$file->url}\n";
632 echo "...and now all the non-duplicates which are longer than 191 characters...\n";
633 $file->query('UPDATE file SET url=LEFT(url, 191) WHERE LENGTH(url)>191');
635 echo "\n...now running hacky pre-schemaupdate change for $table:";
636 // We have to create a urlhash that is _not_ the primary key,
637 // transfer data and THEN run checkSchema
638 $schemadef['fields']['urlhash'] = array (
641 'not null' => false, // this is because when adding column, all entries will _be_ NULL!
642 'description' => 'sha256 of destination URL (url field)',
644 $schemadef['fields']['url'] = array (
646 'description' => 'destination URL after following possible redirections',
648 unset($schemadef['unique keys']);
649 $schema->ensureTable($table, $schemadef);
652 $classname = ucfirst($table);
653 $tablefix = new $classname;
654 // urlhash is hash('sha256', $url) in the File table
655 echo "Updating urlhash fields in $table table...";
656 // Maybe very MySQL specific :(
657 $tablefix->query(sprintf('UPDATE %1$s SET %2$s=%3$s;',
658 $schema->quoteIdentifier($table),
660 // The line below is "result of sha256 on column `url`"
663 echo "Resuming core schema upgrade...";