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 $mimetype; // varchar(50)
32 public $size; // int(4)
33 public $title; // varchar(255)
34 public $date; // int(4)
35 public $protected; // int(4)
36 public $filename; // varchar(255)
37 public $width; // int(4)
38 public $height; // int(4)
39 public $modified; // timestamp() not_null default_CURRENT_TIMESTAMP
41 const URLHASH_ALG = 'sha256';
43 public static function schemaDef()
47 'id' => array('type' => 'serial', 'not null' => true),
48 'urlhash' => array('type' => 'varchar', 'length' => 64, 'description' => 'sha256 of destination URL (url field)'),
49 'url' => array('type' => 'text', 'description' => 'destination URL after following redirections'),
50 'mimetype' => array('type' => 'varchar', 'length' => 50, 'description' => 'mime type of resource'),
51 'size' => array('type' => 'int', 'description' => 'size of resource when available'),
52 'title' => array('type' => 'varchar', 'length' => 255, 'description' => 'title of resource when available'),
53 'date' => array('type' => 'int', 'description' => 'date of resource according to http query'),
54 'protected' => array('type' => 'int', 'description' => 'true when URL is private (needs login)'),
55 'filename' => array('type' => 'varchar', 'length' => 255, 'description' => 'if a local file, name of the file'),
56 'width' => array('type' => 'int', 'description' => 'width in pixels, if it can be described as such and data is available'),
57 'height' => array('type' => 'int', 'description' => 'height in pixels, if it can be described as such and data is available'),
59 'modified' => array('type' => 'timestamp', 'not null' => true, 'description' => 'date this record was modified'),
61 'primary key' => array('id'),
62 'unique keys' => array(
63 'file_urlhash_key' => array('urlhash'),
68 function isProtected($url) {
69 return 'http://www.facebook.com/login.php' === $url;
73 * Save a new file record.
75 * @param array $redir_data lookup data eg from File_redirection::where()
76 * @param string $given_url
79 public static function saveNew(array $redir_data, $given_url) {
81 // I don't know why we have to keep doing this but I'm adding this last check to avoid
84 $file = File::getKV('urlhash', self::hashurl($given_url));
86 if (!$file instanceof File) {
88 $file->urlhash = self::hashurl($given_url);
89 $file->url = $given_url;
90 if (!empty($redir_data['protected'])) $file->protected = $redir_data['protected'];
91 if (!empty($redir_data['title'])) $file->title = $redir_data['title'];
92 if (!empty($redir_data['type'])) $file->mimetype = $redir_data['type'];
93 if (!empty($redir_data['size'])) $file->size = intval($redir_data['size']);
94 if (isset($redir_data['time']) && $redir_data['time'] > 0) $file->date = intval($redir_data['time']);
95 $file_id = $file->insert();
98 Event::handle('EndFileSaveNew', array($file, $redir_data, $given_url));
99 assert ($file instanceof File);
104 * Go look at a URL and possibly save data about it if it's new:
105 * - follow redirect chains and store them in file_redirection
106 * - if a thumbnail is available, save it in file_thumbnail
107 * - save file record with basic info
108 * - optionally save a file_to_post record
109 * - return the File object with the full reference
111 * @fixme refactor this mess, it's gotten pretty scary.
112 * @param string $given_url the URL we're looking at
113 * @param int $notice_id (optional)
114 * @param bool $followRedirects defaults to true
116 * @return mixed File on success, -1 on some errors
118 * @throws ServerException on failure
120 public static function processNew($given_url, $notice_id=null, $followRedirects=true) {
121 if (empty($given_url)) {
122 throw new ServerException('No given URL to process');
125 $given_url = File_redirection::_canonUrl($given_url);
126 if (empty($given_url)) {
127 throw new ServerException('No canonical URL from given URL to process');
130 $file = File::getKV('url', $given_url);
131 if (!$file instanceof File) {
132 // First check if we have a lookup trace for this URL already
133 $file_redir = File_redirection::getKV('url', $given_url);
134 if ($file_redir instanceof File_redirection) {
135 $file = File::getKV('id', $file_redir->file_id);
136 if (!$file instanceof File) {
137 // File did not exist, let's clean up the File_redirection entry
138 $file_redir->delete();
142 // If we still don't have a File object, let's create one now!
143 if (!$file instanceof File) {
144 // @fixme for new URLs this also looks up non-redirect data
145 // such as target content type, size, etc, which we need
146 // for File::saveNew(); so we call it even if not following
148 $redir_data = File_redirection::where($given_url);
149 if (is_array($redir_data)) {
150 $redir_url = $redir_data['url'];
151 } elseif (is_string($redir_data)) {
152 $redir_url = $redir_data;
153 $redir_data = array();
155 // TRANS: Server exception thrown when a URL cannot be processed.
156 throw new ServerException(sprintf(_("Cannot process URL '%s'"), $given_url));
159 // TODO: max field length
160 if ($redir_url === $given_url || strlen($redir_url) > 255 || !$followRedirects) {
161 // Save the File object based on our lookup trace
162 $file = File::saveNew($redir_data, $given_url);
164 // This seems kind of messed up... for now skipping this part
165 // if we're already under a redirect, so we don't go into
166 // horrible infinite loops if we've been given an unstable
167 // redirect (where the final destination of the first request
168 // doesn't match what we get when we ask for it again).
170 // Seen in the wild with clojure.org, which redirects through
171 // wikispaces for auth and appends session data in the URL params.
172 $file = self::processNew($redir_url, $notice_id, /*followRedirects*/false);
173 File_redirection::saveNew($redir_data, $file->id, $given_url);
177 if (!$file instanceof File) {
178 // This should only happen if File::saveNew somehow did not return a File object,
179 // though we have an assert for that in case the event there might've gone wrong.
180 // If anything else goes wrong, there should've been an exception thrown.
181 throw new ServerException('URL processing failed without new File object');
185 if (!empty($notice_id)) {
186 File_to_post::processNew($file->id, $notice_id);
191 public static function respectsQuota(Profile $scoped, $fileSize) {
192 if ($fileSize > common_config('attachments', 'file_quota')) {
193 // TRANS: Message used to be inserted as %2$s in the text "No file may
194 // TRANS: be larger than %1$d byte and the file you sent was %2$s.".
195 // TRANS: %1$d is the number of bytes of an uploaded file.
196 $fileSizeText = sprintf(_m('%1$d byte','%1$d bytes',$fileSize),$fileSize);
198 $fileQuota = common_config('attachments', 'file_quota');
199 // TRANS: Message given if an upload is larger than the configured maximum.
200 // TRANS: %1$d (used for plural) is the byte limit for uploads,
201 // TRANS: %2$s is the proper form of "n bytes". This is the only ways to have
202 // TRANS: gettext support multiple plurals in the same message, unfortunately...
203 throw new ClientException(
204 sprintf(_m('No file may be larger than %1$d byte and the file you sent was %2$s. Try to upload a smaller version.',
205 'No file may be larger than %1$d bytes and the file you sent was %2$s. Try to upload a smaller version.',
207 $fileQuota, $fileSizeText));
212 $query = "select sum(size) as total from file join file_to_post on file_to_post.file_id = file.id join notice on file_to_post.post_id = notice.id where profile_id = {$scoped->id} and file.url like '%/notice/%/file'";
213 $file->query($query);
215 $total = $file->total + $fileSize;
216 if ($total > common_config('attachments', 'user_quota')) {
217 // TRANS: Message given if an upload would exceed user quota.
218 // TRANS: %d (number) is the user quota in bytes and is used for plural.
219 throw new ClientException(
220 sprintf(_m('A file this large would exceed your user quota of %d byte.',
221 'A file this large would exceed your user quota of %d bytes.',
222 common_config('attachments', 'user_quota')),
223 common_config('attachments', 'user_quota')));
225 $query .= ' AND EXTRACT(month FROM file.modified) = EXTRACT(month FROM now()) and EXTRACT(year FROM file.modified) = EXTRACT(year FROM now())';
226 $file->query($query);
228 $total = $file->total + $fileSize;
229 if ($total > common_config('attachments', 'monthly_quota')) {
230 // TRANS: Message given id an upload would exceed a user's monthly quota.
231 // TRANS: $d (number) is the monthly user quota in bytes and is used for plural.
232 throw new ClientException(
233 sprintf(_m('A file this large would exceed your monthly quota of %d byte.',
234 'A file this large would exceed your monthly quota of %d bytes.',
235 common_config('attachments', 'monthly_quota')),
236 common_config('attachments', 'monthly_quota')));
241 // where should the file go?
243 static function filename(Profile $profile, $origname, $mimetype)
246 $ext = common_supported_mime_to_ext($mimetype);
247 } catch (Exception $e) {
248 // We don't support this mimetype, but let's guess the extension
249 $ext = substr(strrchr($mimetype, '/'), 1);
252 // Normalize and make the original filename more URL friendly.
253 $origname = basename($origname, ".$ext");
254 if (class_exists('Normalizer')) {
255 // http://php.net/manual/en/class.normalizer.php
256 // http://www.unicode.org/reports/tr15/
257 $origname = Normalizer::normalize($origname, Normalizer::FORM_KC);
259 $origname = preg_replace('/[^A-Za-z0-9\.\_]/', '_', $origname);
261 $nickname = $profile->getNickname();
262 $datestamp = strftime('%Y%m%d', time());
264 // generate new random strings until we don't run into a filename collision.
265 $random = strtolower(common_confirmation_code(16));
266 $filename = "$nickname-$datestamp-$origname-$random.$ext";
267 } while (file_exists(self::path($filename)));
272 * Validation for as-saved base filenames
274 static function validFilename($filename)
276 return preg_match('/^[A-Za-z0-9._-]+$/', $filename);
280 * @throws ClientException on invalid filename
282 static function path($filename)
284 if (!self::validFilename($filename)) {
285 // TRANS: Client exception thrown if a file upload does not have a valid name.
286 throw new ClientException(_("Invalid filename."));
288 $dir = common_config('attachments', 'dir');
290 if ($dir[strlen($dir)-1] != '/') {
294 return $dir . $filename;
297 static function url($filename)
299 if (!self::validFilename($filename)) {
300 // TRANS: Client exception thrown if a file upload does not have a valid name.
301 throw new ClientException(_("Invalid filename."));
304 if (common_config('site','private')) {
306 return common_local_url('getfile',
307 array('filename' => $filename));
311 if (StatusNet::useHTTPS()) {
313 $sslserver = common_config('attachments', 'sslserver');
315 if (empty($sslserver)) {
316 // XXX: this assumes that background dir == site dir + /file/
317 // not true if there's another server
318 if (is_string(common_config('site', 'sslserver')) &&
319 mb_strlen(common_config('site', 'sslserver')) > 0) {
320 $server = common_config('site', 'sslserver');
321 } else if (common_config('site', 'server')) {
322 $server = common_config('site', 'server');
324 $path = common_config('site', 'path') . '/file/';
326 $server = $sslserver;
327 $path = common_config('attachments', 'sslpath');
329 $path = common_config('attachments', 'path');
335 $path = common_config('attachments', 'path');
336 $server = common_config('attachments', 'server');
338 if (empty($server)) {
339 $server = common_config('site', 'server');
342 $ssl = common_config('attachments', 'ssl');
344 $protocol = ($ssl) ? 'https' : 'http';
347 if ($path[strlen($path)-1] != '/') {
351 if ($path[0] != '/') {
355 return $protocol.'://'.$server.$path.$filename;
358 function getEnclosure(){
359 $enclosure = (object) array();
360 foreach (array('title', 'url', 'date', 'modified', 'size', 'mimetype') as $key) {
361 $enclosure->$key = $this->$key;
364 $needMoreMetadataMimetypes = array(null, 'application/xhtml+xml');
366 if (!isset($this->filename) && in_array(common_bare_mime($enclosure->mimetype), $needMoreMetadataMimetypes)) {
367 // This fetches enclosure metadata for non-local links with unset/HTML mimetypes,
368 // which may be enriched through oEmbed or similar (implemented as plugins)
369 Event::handle('FileEnclosureMetadata', array($this, &$enclosure));
371 if (empty($enclosure->mimetype) || in_array(common_bare_mime($enclosure->mimetype), $needMoreMetadataMimetypes)) {
372 // This means we either don't know what it is, so it can't
373 // be shown as an enclosure, or it is an HTML link which
374 // does not link to a resource with further metadata.
375 throw new ServerException('Unknown enclosure mimetype, not enough metadata');
381 * Get the attachment's thumbnail record, if any.
382 * Make sure you supply proper 'int' typed variables (or null).
384 * @param $width int Max width of thumbnail in pixels. (if null, use common_config values)
385 * @param $height int Max height of thumbnail in pixels. (if null, square-crop to $width)
386 * @param $crop bool Crop to the max-values' aspect ratio
388 * @return File_thumbnail
390 public function getThumbnail($width=null, $height=null, $crop=false, $force_still=true)
392 // Get some more information about this file through our ImageFile class
393 $image = ImageFile::fromFileObject($this);
394 if ($image->animated && !common_config('thumbnail', 'animated')) {
395 // null means "always use file as thumbnail"
396 // false means you get choice between frozen frame or original when calling getThumbnail
397 if (is_null(common_config('thumbnail', 'animated')) || !$force_still) {
398 throw new UseFileAsThumbnailException($this->id);
402 if ($width === null) {
403 $width = common_config('thumbnail', 'width');
404 $height = common_config('thumbnail', 'height');
405 $crop = common_config('thumbnail', 'crop');
408 if ($height === null) {
413 // Get proper aspect ratio width and height before lookup
414 // We have to do it through an ImageFile object because of orientation etc.
415 // Only other solution would've been to rotate + rewrite uploaded files.
416 list($width, $height, $x, $y, $w, $h) =
417 $image->scaleToFit($width, $height, $crop);
419 $params = array('file_id'=> $this->id,
421 'height' => $height);
422 $thumb = File_thumbnail::pkeyGet($params);
423 if ($thumb instanceof File_thumbnail) {
427 // throws exception on failure to generate thumbnail
428 $outname = "thumb-{$width}x{$height}-" . $image->filename;
429 $outpath = self::path($outname);
431 // The boundary box for our resizing
432 $box = array('width'=>$width, 'height'=>$height,
436 // Doublecheck that parameters are sane and integers.
437 if ($box['width'] < 1 || $box['width'] > common_config('thumbnail', 'maxsize')
438 || $box['height'] < 1 || $box['height'] > common_config('thumbnail', 'maxsize')
439 || $box['w'] < 1 || $box['x'] >= $image->width
440 || $box['h'] < 1 || $box['y'] >= $image->height) {
441 // Fail on bad width parameter. If this occurs, it's due to algorithm in ImageFile->scaleToFit
442 common_debug("Boundary box parameters for resize of {$image->filepath} : ".var_export($box,true));
443 throw new ServerException('Bad thumbnail size parameters.');
446 common_debug(sprintf('Generating a thumbnail of File id==%u of size %ux%u', $this->id, $width, $height));
447 // Perform resize and store into file
448 $image->resizeTo($outpath, $box);
450 // Avoid deleting the original
451 if ($image->getPath() != self::path($image->filename)) {
454 return File_thumbnail::saveThumbnail($this->id,
460 public function getPath()
462 return self::path($this->filename);
465 public function getUrl()
467 if (!empty($this->filename)) {
468 // A locally stored file, so let's generate a URL for our instance.
469 $url = self::url($this->filename);
470 if (self::urlhash($url) !== $this->urlhash) {
471 // For indexing purposes, in case we do a lookup on the 'url' field.
472 // also we're fixing possible changes from http to https, or paths
473 $this->updateUrl($url);
478 // No local filename available, return the URL we have stored
482 static public function getByUrl($url)
485 $file->urlhash = self::hashurl($url);
486 if (!$file->find(true)) {
487 throw new NoResultException($file);
492 public function updateUrl($url)
494 $file = File::getKV('urlhash', self::hashurl($url));
495 if ($file instanceof File) {
496 throw new ServerException('URL already exists in DB');
498 $sql = 'UPDATE %1$s SET urlhash=%2$s, url=%3$s WHERE urlhash=%4$s;';
499 $result = $this->query(sprintf($sql, $this->__table,
500 $this->_quote((string)self::hashurl($url)),
501 $this->_quote((string)$url),
502 $this->_quote((string)$this->urlhash)));
503 if ($result === false) {
504 common_log_db_error($this, 'UPDATE', __FILE__);
505 throw new ServerException("Could not UPDATE {$this->__table}.url");
512 * Blow the cache of notices that link to this URL
514 * @param boolean $last Whether to blow the "last" cache too
519 function blowCache($last=false)
521 self::blow('file:notice-ids:%s', $this->urlhash);
523 self::blow('file:notice-ids:%s;last', $this->urlhash);
525 self::blow('file:notice-count:%d', $this->id);
529 * Stream of notices linking to this URL
531 * @param integer $offset Offset to show; default is 0
532 * @param integer $limit Limit of notices to show
533 * @param integer $since_id Since this notice
534 * @param integer $max_id Before this notice
536 * @return array ids of notices that link to this file
539 function stream($offset=0, $limit=NOTICES_PER_PAGE, $since_id=0, $max_id=0)
541 $stream = new FileNoticeStream($this);
542 return $stream->getNotices($offset, $limit, $since_id, $max_id);
545 function noticeCount()
547 $cacheKey = sprintf('file:notice-count:%d', $this->id);
549 $count = self::cacheGet($cacheKey);
551 if ($count === false) {
553 $f2p = new File_to_post();
555 $f2p->file_id = $this->id;
557 $count = $f2p->count();
559 self::cacheSet($cacheKey, $count);
565 public function isLocal()
567 return !empty($this->filename);
570 public function delete($useWhere=false)
572 // Delete the file, if it exists locally
573 if (!empty($this->filename) && file_exists(self::path($this->filename))) {
574 $deleted = @unlink(self::path($this->filename));
576 common_log(LOG_ERR, sprintf('Could not unlink existing file: "%s"', self::path($this->filename)));
580 // Clear out related things in the database and filesystem, such as thumbnails
581 if (Event::handle('FileDeleteRelated', array($this))) {
582 $thumbs = new File_thumbnail();
583 $thumbs->file_id = $this->id;
584 if ($thumbs->find()) {
585 while ($thumbs->fetch()) {
591 // And finally remove the entry from the database
592 return parent::delete($useWhere);
595 public function getTitle()
597 $title = $this->title ?: $this->filename;
599 return $title ?: null;
602 static public function hashurl($url)
605 throw new Exception('No URL provided to hash algorithm.');
607 return hash(self::URLHASH_ALG, $url);
610 static public function beforeSchemaUpdate()
612 $table = strtolower(get_called_class());
613 $schema = Schema::get();
614 $schemadef = $schema->getTableDef($table);
616 // 2015-02-19 We have to upgrade our table definitions to have the urlhash field populated
617 if (isset($schemadef['fields']['urlhash']) && in_array('file_urlhash_key', $schemadef['unique keys'])) {
618 // We already have the urlhash field, so no need to migrate it.
621 echo "\nFound old $table table, upgrading it to contain 'urlhash' field...\n";
622 // We have to create a urlhash that is _not_ the primary key,
623 // transfer data and THEN run checkSchema
624 $schemadef['fields']['urlhash'] = array (
627 'description' => 'sha256 of destination URL after following redirections',
629 $schema->ensureTable($table, $schemadef);
632 $classname = ucfirst($table);
633 $tablefix = new $classname;
634 // urlhash is hash('sha256', $url) in the File table
635 echo "Updating urlhash fields in $table table...\n";
636 // Maybe very MySQL specific :(
637 $tablefix->query(sprintf('UPDATE %1$s SET %2$s=%3$s;',
638 $schema->quoteIdentifier($table),
640 // The line below is "result of sha256 on column `url`"
643 echo "Resuming core schema upgrade...";