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 $modified; // timestamp() not_null default_CURRENT_TIMESTAMP
42 const URLHASH_ALG = 'sha256';
43 const FILEHASH_ALG = 'sha256';
45 public static function schemaDef()
49 'id' => array('type' => 'serial', 'not null' => true),
50 'urlhash' => array('type' => 'varchar', 'length' => 64, 'not null' => true, 'description' => 'sha256 of destination URL (url field)'),
51 'url' => array('type' => 'text', 'description' => 'destination URL after following possible redirections'),
52 'filehash' => array('type' => 'varchar', 'length' => 64, 'not null' => false, 'description' => 'sha256 of the file contents, only for locally stored files of course'),
53 'mimetype' => array('type' => 'varchar', 'length' => 50, 'description' => 'mime type of resource'),
54 'size' => array('type' => 'int', 'description' => 'size of resource when available'),
55 'title' => array('type' => 'text', 'description' => 'title of resource when available'),
56 'date' => array('type' => 'int', 'description' => 'date of resource according to http query'),
57 'protected' => array('type' => 'int', 'description' => 'true when URL is private (needs login)'),
58 'filename' => array('type' => 'text', 'description' => 'if file is stored locally (too) this is the filename'),
59 'width' => array('type' => 'int', 'description' => 'width in pixels, if it can be described as such and data is available'),
60 'height' => array('type' => 'int', 'description' => 'height in pixels, if it can be described as such and data is available'),
62 'modified' => array('type' => 'timestamp', 'not null' => true, 'description' => 'date this record was modified'),
64 'primary key' => array('id'),
65 'unique keys' => array(
66 'file_urlhash_key' => array('urlhash'),
69 'file_filehash_idx' => array('filehash'),
74 public static function isProtected($url) {
76 $protected_urls_exps = array(
77 'https://www.facebook.com/login.php',
78 common_path('main/login')
81 foreach ($protected_urls_exps as $protected_url_exp) {
82 if (preg_match('!^'.preg_quote($protected_url_exp).'(.*)$!i', $url) === 1) {
91 * Save a new file record.
93 * @param array $redir_data lookup data eg from File_redirection::where()
94 * @param string $given_url
97 public static function saveNew(array $redir_data, $given_url)
101 // I don't know why we have to keep doing this but we run a last check to avoid
103 $file = File::getByUrl($given_url);
105 } catch (NoResultException $e) {
106 // We don't have the file's URL since before, so let's continue.
109 // if the given url is an local attachment url and the id already exists, don't
110 // save a new file record. This should never happen, but let's make it foolproof
111 $attachment_path = common_path('attachment/');
112 if(strpos($given_url,$attachment_path) == 0) {
113 $possible_file_id = substr($given_url,strlen($attachment_path));
114 if(is_numeric($possible_file_id)) {
115 $file = File::getKV('id',$possible_file_id);
116 if($file instanceof File) {
123 $file->url = $given_url;
124 if (!empty($redir_data['protected'])) $file->protected = $redir_data['protected'];
125 if (!empty($redir_data['title'])) $file->title = $redir_data['title'];
126 if (!empty($redir_data['type'])) $file->mimetype = $redir_data['type'];
127 if (!empty($redir_data['size'])) $file->size = intval($redir_data['size']);
128 if (isset($redir_data['time']) && $redir_data['time'] > 0) $file->date = intval($redir_data['time']);
133 public function saveFile() {
134 $this->urlhash = self::hashurl($this->url);
136 if (!Event::handle('StartFileSaveNew', array(&$this))) {
137 throw new ServerException('File not saved due to an aborted StartFileSaveNew event.');
140 $this->id = $this->insert();
142 if ($this->id === false) {
143 throw new ServerException('File/URL metadata could not be saved to the database.');
146 Event::handle('EndFileSaveNew', array($this));
150 * Go look at a URL and possibly save data about it if it's new:
151 * - follow redirect chains and store them in file_redirection
152 * - if a thumbnail is available, save it in file_thumbnail
153 * - save file record with basic info
154 * - optionally save a file_to_post record
155 * - return the File object with the full reference
157 * @param string $given_url the URL we're looking at
158 * @param Notice $notice (optional)
159 * @param bool $followRedirects defaults to true
161 * @return mixed File on success, -1 on some errors
163 * @throws ServerException on failure
165 public static function processNew($given_url, Notice $notice=null, $followRedirects=true) {
166 if (empty($given_url)) {
167 throw new ServerException('No given URL to process');
170 $given_url = File_redirection::_canonUrl($given_url);
171 if (empty($given_url)) {
172 throw new ServerException('No canonical URL from given URL to process');
175 $redir = File_redirection::where($given_url);
176 $file = $redir->getFile();
178 if (!$file instanceof File || empty($file->id)) {
179 // This should not happen
180 throw new ServerException('URL processing failed without new File object');
183 if ($notice instanceof Notice) {
184 File_to_post::processNew($file, $notice);
190 public static function respectsQuota(Profile $scoped, $fileSize) {
191 if ($fileSize > common_config('attachments', 'file_quota')) {
192 // TRANS: Message used to be inserted as %2$s in the text "No file may
193 // TRANS: be larger than %1$d byte and the file you sent was %2$s.".
194 // TRANS: %1$d is the number of bytes of an uploaded file.
195 $fileSizeText = sprintf(_m('%1$d byte','%1$d bytes',$fileSize),$fileSize);
197 $fileQuota = common_config('attachments', 'file_quota');
198 // TRANS: Message given if an upload is larger than the configured maximum.
199 // TRANS: %1$d (used for plural) is the byte limit for uploads,
200 // TRANS: %2$s is the proper form of "n bytes". This is the only ways to have
201 // TRANS: gettext support multiple plurals in the same message, unfortunately...
202 throw new ClientException(
203 sprintf(_m('No file may be larger than %1$d byte and the file you sent was %2$s. Try to upload a smaller version.',
204 'No file may be larger than %1$d bytes and the file you sent was %2$s. Try to upload a smaller version.',
206 $fileQuota, $fileSizeText));
211 $query = "select sum(size) as total from file join file_to_post on file_to_post.file_id = file.id join notice on file_to_post.post_id = notice.id where profile_id = {$scoped->id} and file.url like '%/notice/%/file'";
212 $file->query($query);
214 $total = $file->total + $fileSize;
215 if ($total > common_config('attachments', 'user_quota')) {
216 // TRANS: Message given if an upload would exceed user quota.
217 // TRANS: %d (number) is the user quota in bytes and is used for plural.
218 throw new ClientException(
219 sprintf(_m('A file this large would exceed your user quota of %d byte.',
220 'A file this large would exceed your user quota of %d bytes.',
221 common_config('attachments', 'user_quota')),
222 common_config('attachments', 'user_quota')));
224 $query .= ' AND EXTRACT(month FROM file.modified) = EXTRACT(month FROM now()) and EXTRACT(year FROM file.modified) = EXTRACT(year FROM now())';
225 $file->query($query);
227 $total = $file->total + $fileSize;
228 if ($total > common_config('attachments', 'monthly_quota')) {
229 // TRANS: Message given id an upload would exceed a user's monthly quota.
230 // TRANS: $d (number) is the monthly user quota in bytes and is used for plural.
231 throw new ClientException(
232 sprintf(_m('A file this large would exceed your monthly quota of %d byte.',
233 'A file this large would exceed your monthly quota of %d bytes.',
234 common_config('attachments', 'monthly_quota')),
235 common_config('attachments', 'monthly_quota')));
240 public function getFilename()
242 if (!self::validFilename($this->filename)) {
243 // TRANS: Client exception thrown if a file upload does not have a valid name.
244 throw new ClientException(_("Invalid filename."));
246 return $this->filename;
249 // where should the file go?
251 static function filename(Profile $profile, $origname, $mimetype)
253 $ext = self::guessMimeExtension($mimetype, $origname);
255 // Normalize and make the original filename more URL friendly.
256 $origname = basename($origname, ".$ext");
257 if (class_exists('Normalizer')) {
258 // http://php.net/manual/en/class.normalizer.php
259 // http://www.unicode.org/reports/tr15/
260 $origname = Normalizer::normalize($origname, Normalizer::FORM_KC);
262 $origname = preg_replace('/[^A-Za-z0-9\.\_]/', '_', $origname);
264 $nickname = $profile->getNickname();
265 $datestamp = strftime('%Y%m%d', time());
267 // generate new random strings until we don't run into a filename collision.
268 $random = strtolower(common_confirmation_code(16));
269 $filename = "$nickname-$datestamp-$origname-$random.$ext";
270 } while (file_exists(self::path($filename)));
275 * @param $mimetype The mimetype we've discovered for this file.
276 * @param $filename An optional filename which we can use on failure.
278 static function guessMimeExtension($mimetype, $filename=null)
281 // first see if we know the extension for our mimetype
282 $ext = common_supported_mime_to_ext($mimetype);
285 } catch (Exception $e) { // FIXME: Make this exception more specific to "unknown mime=>ext relation"
286 // We don't know the extension for this mimetype, but let's guess.
288 // If we are very liberal with uploads ($config['attachments']['supported'] === true)
289 // then we try to do some guessing based on the filename, if it was supplied.
290 if (!is_null($filename) && common_config('attachments', 'supported')===true
291 && preg_match('/^.+\.([A-Za-z0-9]+)$/', $filename, $matches)) {
292 // we matched on a file extension, so let's see if it means something.
293 $ext = mb_strtolower($matches[1]);
295 $blacklist = common_config('attachments', 'extblacklist');
296 // If we got an extension from $filename we want to check if it's in a blacklist
297 // so we avoid people uploading .php files etc.
298 if (array_key_exists($ext, $blacklist)) {
299 if (!is_string($blacklist[$ext])) {
300 // we don't have a safe replacement extension
301 throw new ClientException(_('Blacklisted file extension.'));
303 common_debug('Found replaced extension for filename '._ve($filename).': '._ve($ext));
305 // return a safe replacement extension ('php' => 'phps' for example)
306 return $blacklist[$ext];
308 // the attachment extension based on its filename was not blacklisted so it's ok to use it
313 // If nothing else has given us a result, try to extract it from
314 // the mimetype value (this turns .jpg to .jpeg for example...)
316 // FIXME: try to build a regexp that will get jpeg from image/jpeg as well as json from application/jrd+json
317 if (!preg_match('/\/([a-z0-9]+)/', mb_strtolower($mimetype), $matches)) {
318 throw new Exception('Malformed mimetype: '.$mimetype);
320 return mb_strtolower($matches[1]);
324 * Validation for as-saved base filenames
326 static function validFilename($filename)
328 return preg_match('/^[A-Za-z0-9._-]+$/', $filename);
332 * @throws ClientException on invalid filename
334 static function path($filename)
336 if (!self::validFilename($filename)) {
337 // TRANS: Client exception thrown if a file upload does not have a valid name.
338 throw new ClientException(_("Invalid filename."));
340 $dir = common_config('attachments', 'dir');
342 if ($dir[strlen($dir)-1] != '/') {
346 return $dir . $filename;
349 static function url($filename)
351 if (!self::validFilename($filename)) {
352 // TRANS: Client exception thrown if a file upload does not have a valid name.
353 throw new ClientException(_("Invalid filename."));
356 if (common_config('site','private')) {
358 return common_local_url('getfile',
359 array('filename' => $filename));
363 if (GNUsocial::useHTTPS()) {
365 $sslserver = common_config('attachments', 'sslserver');
367 if (empty($sslserver)) {
368 // XXX: this assumes that background dir == site dir + /file/
369 // not true if there's another server
370 if (is_string(common_config('site', 'sslserver')) &&
371 mb_strlen(common_config('site', 'sslserver')) > 0) {
372 $server = common_config('site', 'sslserver');
373 } else if (common_config('site', 'server')) {
374 $server = common_config('site', 'server');
376 $path = common_config('site', 'path') . '/file/';
378 $server = $sslserver;
379 $path = common_config('attachments', 'sslpath');
381 $path = common_config('attachments', 'path');
387 $path = common_config('attachments', 'path');
388 $server = common_config('attachments', 'server');
390 if (empty($server)) {
391 $server = common_config('site', 'server');
394 $ssl = common_config('attachments', 'ssl');
396 $protocol = ($ssl) ? 'https' : 'http';
399 if ($path[strlen($path)-1] != '/') {
403 if ($path[0] != '/') {
407 return $protocol.'://'.$server.$path.$filename;
410 static $_enclosures = array();
412 function getEnclosure(){
413 if (isset(self::$_enclosures[$this->getID()])) {
414 return self::$_enclosures[$this->getID()];
417 $enclosure = (object) array();
418 foreach (array('title', 'url', 'date', 'modified', 'size', 'mimetype', 'width', 'height') as $key) {
419 if ($this->$key !== '') {
420 $enclosure->$key = $this->$key;
424 $needMoreMetadataMimetypes = array(null, 'application/xhtml+xml', 'text/html');
426 if (!isset($this->filename) && in_array(common_bare_mime($enclosure->mimetype), $needMoreMetadataMimetypes)) {
427 // This fetches enclosure metadata for non-local links with unset/HTML mimetypes,
428 // which may be enriched through oEmbed or similar (implemented as plugins)
429 Event::handle('FileEnclosureMetadata', array($this, &$enclosure));
431 if (empty($enclosure->mimetype)) {
432 // This means we either don't know what it is, so it can't
433 // be shown as an enclosure, or it is an HTML link which
434 // does not link to a resource with further metadata.
435 throw new ServerException('Unknown enclosure mimetype, not enough metadata');
438 self::$_enclosures[$this->getID()] = $enclosure;
442 public function hasThumbnail()
445 $this->getThumbnail();
446 } catch (Exception $e) {
453 * Get the attachment's thumbnail record, if any.
454 * Make sure you supply proper 'int' typed variables (or null).
456 * @param $width int Max width of thumbnail in pixels. (if null, use common_config values)
457 * @param $height int Max height of thumbnail in pixels. (if null, square-crop to $width)
458 * @param $crop bool Crop to the max-values' aspect ratio
460 * @return File_thumbnail
462 * @throws UseFileAsThumbnailException if the file is considered an image itself and should be itself as thumbnail
463 * @throws UnsupportedMediaException if, despite trying, we can't understand how to make a thumbnail for this format
464 * @throws ServerException on various other errors
466 public function getThumbnail($width=null, $height=null, $crop=false, $force_still=true, $upscale=null)
468 // Get some more information about this file through our ImageFile class
469 $image = ImageFile::fromFileObject($this);
470 if ($image->animated && !common_config('thumbnail', 'animated')) {
471 // null means "always use file as thumbnail"
472 // false means you get choice between frozen frame or original when calling getThumbnail
473 if (is_null(common_config('thumbnail', 'animated')) || !$force_still) {
474 throw new UseFileAsThumbnailException($this->id);
478 return $image->getFileThumbnail($width, $height, $crop,
479 !is_null($upscale) ? $upscale : common_config('thumbnail', 'upscale'));
482 public function getPath()
484 $filepath = self::path($this->filename);
485 if (!file_exists($filepath)) {
486 throw new FileNotFoundException($filepath);
491 public function getUrl($prefer_local=true)
493 if ($prefer_local && !empty($this->filename)) {
494 // A locally stored file, so let's generate a URL for our instance.
495 return self::url($this->filename);
498 // No local filename available, return the URL we have stored
502 static public function getByUrl($url)
505 $file->urlhash = self::hashurl($url);
506 if (!$file->find(true)) {
507 throw new NoResultException($file);
513 * @param string $hashstr String of (preferrably lower case) hexadecimal characters, same as result of 'hash_file(...)'
515 static public function getByHash($hashstr)
518 $file->filehash = strtolower($hashstr);
519 if (!$file->find(true)) {
520 throw new NoResultException($file);
525 public function updateUrl($url)
527 $file = File::getKV('urlhash', self::hashurl($url));
528 if ($file instanceof File) {
529 throw new ServerException('URL already exists in DB');
531 $sql = 'UPDATE %1$s SET urlhash=%2$s, url=%3$s WHERE urlhash=%4$s;';
532 $result = $this->query(sprintf($sql, $this->tableName(),
533 $this->_quote((string)self::hashurl($url)),
534 $this->_quote((string)$url),
535 $this->_quote((string)$this->urlhash)));
536 if ($result === false) {
537 common_log_db_error($this, 'UPDATE', __FILE__);
538 throw new ServerException("Could not UPDATE {$this->tableName()}.url");
545 * Blow the cache of notices that link to this URL
547 * @param boolean $last Whether to blow the "last" cache too
552 function blowCache($last=false)
554 self::blow('file:notice-ids:%s', $this->id);
556 self::blow('file:notice-ids:%s;last', $this->id);
558 self::blow('file:notice-count:%d', $this->id);
562 * Stream of notices linking to this URL
564 * @param integer $offset Offset to show; default is 0
565 * @param integer $limit Limit of notices to show
566 * @param integer $since_id Since this notice
567 * @param integer $max_id Before this notice
569 * @return array ids of notices that link to this file
572 function stream($offset=0, $limit=NOTICES_PER_PAGE, $since_id=0, $max_id=0)
574 $stream = new FileNoticeStream($this);
575 return $stream->getNotices($offset, $limit, $since_id, $max_id);
578 function noticeCount()
580 $cacheKey = sprintf('file:notice-count:%d', $this->id);
582 $count = self::cacheGet($cacheKey);
584 if ($count === false) {
586 $f2p = new File_to_post();
588 $f2p->file_id = $this->id;
590 $count = $f2p->count();
592 self::cacheSet($cacheKey, $count);
598 public function isLocal()
600 return !empty($this->filename);
603 public function delete($useWhere=false)
605 // Delete the file, if it exists locally
606 if (!empty($this->filename) && file_exists(self::path($this->filename))) {
607 $deleted = @unlink(self::path($this->filename));
609 common_log(LOG_ERR, sprintf('Could not unlink existing file: "%s"', self::path($this->filename)));
613 // Clear out related things in the database and filesystem, such as thumbnails
614 if (Event::handle('FileDeleteRelated', array($this))) {
615 $thumbs = new File_thumbnail();
616 $thumbs->file_id = $this->id;
617 if ($thumbs->find()) {
618 while ($thumbs->fetch()) {
623 $f2p = new File_to_post();
624 $f2p->file_id = $this->id;
626 while ($f2p->fetch()) {
632 // And finally remove the entry from the database
633 return parent::delete($useWhere);
636 public function getTitle()
638 $title = $this->title ?: $this->filename;
640 return $title ?: null;
643 static public function hashurl($url)
646 throw new Exception('No URL provided to hash algorithm.');
648 return hash(self::URLHASH_ALG, $url);
651 static public function beforeSchemaUpdate()
653 $table = strtolower(get_called_class());
654 $schema = Schema::get();
655 $schemadef = $schema->getTableDef($table);
657 // 2015-02-19 We have to upgrade our table definitions to have the urlhash field populated
658 if (isset($schemadef['fields']['urlhash']) && isset($schemadef['unique keys']['file_urlhash_key'])) {
659 // We already have the urlhash field, so no need to migrate it.
662 echo "\nFound old $table table, upgrading it to contain 'urlhash' field...";
665 $file->query(sprintf('SELECT id, LEFT(url, 191) AS shortenedurl, COUNT(*) AS c FROM %1$s WHERE LENGTH(url)>191 GROUP BY shortenedurl HAVING c > 1', $schema->quoteIdentifier($table)));
666 print "\nFound {$file->N} URLs with too long entries in file table\n";
667 while ($file->fetch()) {
668 // We've got a URL that is too long for our future file table
669 // so we'll cut it. We could save the original URL, but there is
670 // no guarantee it is complete anyway since the previous max was 255 chars.
671 $dupfile = new File();
672 // First we find file entries that would be duplicates of this when shortened
673 // ... and we'll just throw the dupes out the window for now! It's already so borken.
674 $dupfile->query(sprintf('SELECT * FROM file WHERE LEFT(url, 191) = "%1$s"', $file->shortenedurl));
675 // Leave one of the URLs in the database by using ->find(true) (fetches first entry)
676 if ($dupfile->find(true)) {
677 print "\nShortening url entry for $table id: {$file->id} [";
678 $orig = clone($dupfile);
679 $dupfile->url = $file->shortenedurl; // make sure it's only 191 chars from now on
680 $dupfile->update($orig);
681 print "\nDeleting duplicate entries of too long URL on $table id: {$file->id} [";
682 // only start deleting with this fetch.
683 while($dupfile->fetch()) {
689 print "\nWarning! URL suddenly disappeared from database: {$file->url}\n";
692 echo "...and now all the non-duplicates which are longer than 191 characters...\n";
693 $file->query('UPDATE file SET url=LEFT(url, 191) WHERE LENGTH(url)>191');
695 echo "\n...now running hacky pre-schemaupdate change for $table:";
696 // We have to create a urlhash that is _not_ the primary key,
697 // transfer data and THEN run checkSchema
698 $schemadef['fields']['urlhash'] = array (
701 'not null' => false, // this is because when adding column, all entries will _be_ NULL!
702 'description' => 'sha256 of destination URL (url field)',
704 $schemadef['fields']['url'] = array (
706 'description' => 'destination URL after following possible redirections',
708 unset($schemadef['unique keys']);
709 $schema->ensureTable($table, $schemadef);
712 $classname = ucfirst($table);
713 $tablefix = new $classname;
714 // urlhash is hash('sha256', $url) in the File table
715 echo "Updating urlhash fields in $table table...";
716 // Maybe very MySQL specific :(
717 $tablefix->query(sprintf('UPDATE %1$s SET %2$s=%3$s;',
718 $schema->quoteIdentifier($table),
720 // The line below is "result of sha256 on column `url`"
723 echo "Resuming core schema upgrade...";