Move to subfolders for preview files

Else the number of files can grow very large very quickly in the preview
folder. Esp on large systems.

This generates the md5 of the fileid. And then creates folders of the
first 7 charts. In that folder is then a folder with the fileid. And
inside there are the previews.

Signed-off-by: Roeland Jago Douma <roeland@famdouma.nl>

3 files changed, 151 insertions, 18 deletions

diff --git a/lib/private/Preview/BackgroundCleanupJob.php b/lib/private/Preview/BackgroundCleanupJob.php
index 15bb6bd7188..e0546c4a11c 100644
--- a/lib/private/Preview/BackgroundCleanupJob.php
+++ b/lib/private/Preview/BackgroundCleanupJob.php
@@ -27,8 +27,9 @@ declare(strict_types=1);
 namespace OC\Preview;
 
 use OC\BackgroundJob\TimedJob;
-use OC\Files\AppData\Factory;
+use OC\Preview\Storage\Root;
 use OCP\DB\QueryBuilder\IQueryBuilder;
+use OCP\Files\IMimeTypeLoader;
 use OCP\Files\NotFoundException;
 use OCP\Files\NotPermittedException;
 use OCP\IDBConnection;
@@ -38,28 +39,47 @@ class BackgroundCleanupJob extends TimedJob {
 	/** @var IDBConnection */
 	private $connection;
 
-	/** @var Factory */
-	private $appDataFactory;
+	/** @var Root */
+	private $previewFolder;
 
 	/** @var bool */
 	private $isCLI;
 
+	/** @var IMimeTypeLoader */
+	private $mimeTypeLoader;
+
 	public function __construct(IDBConnection $connection,
-								Factory $appDataFactory,
+								Root $previewFolder,
+								IMimeTypeLoader $mimeTypeLoader,
 								bool $isCLI) {
 		// Run at most once an hour
 		$this->setInterval(3600);
 
 		$this->connection = $connection;
-		$this->appDataFactory = $appDataFactory;
+		$this->previewFolder = $previewFolder;
 		$this->isCLI = $isCLI;
+		$this->mimeTypeLoader = $mimeTypeLoader;
 	}
 
 	public function run($argument) {
-		$previews = $this->appDataFactory->get('preview');
+		foreach ($this->getDeletedFiles() as $fileId) {
+			try {
+				$preview = $this->previewFolder->getFolder((string)$fileId);
+				$preview->delete();
+			} catch (NotFoundException $e) {
+				// continue
+			} catch (NotPermittedException $e) {
+				// continue
+			}
+		}
+	}
 
-		$previewFodlerId = $previews->getId();
+	private function getDeletedFiles(): \Iterator {
+		yield from $this->getOldPreviewLocations();
+		yield from $this->getNewPreviewLocations();
+	}
 
+	private function getOldPreviewLocations(): \Iterator {
 		$qb = $this->connection->getQueryBuilder();
 		$qb->select('a.name')
 			->from('filecache', 'a')
@@ -69,7 +89,9 @@ class BackgroundCleanupJob extends TimedJob {
 			->where(
 				$qb->expr()->isNull('b.fileid')
 			)->andWhere(
-				$qb->expr()->eq('a.parent', $qb->createNamedParameter($previewFodlerId))
+				$qb->expr()->eq('a.parent', $qb->createNamedParameter($this->previewFolder->getId()))
+			)->andWhere(
+				$qb->expr()->like('a.name', $qb->createNamedParameter('__%'))
 			);
 
 		if (!$this->isCLI) {
@@ -79,14 +101,54 @@ class BackgroundCleanupJob extends TimedJob {
 		$cursor = $qb->execute();
 
 		while ($row = $cursor->fetch()) {
-			try {
-				$preview = $previews->getFolder($row['name']);
-				$preview->delete();
-			} catch (NotFoundException $e) {
-				// continue
-			} catch (NotPermittedException $e) {
-				// continue
-			}
+			yield $row['name'];
+		}
+
+		$cursor->closeCursor();
+	}
+
+	private function getNewPreviewLocations(): \Iterator {
+		$qb = $this->connection->getQueryBuilder();
+		$qb->select('path', 'mimetype')
+			->from('filecache')
+			->where($qb->expr()->eq('fileid', $qb->createNamedParameter($this->previewFolder->getId())));
+		$cursor = $qb->execute();
+		$data = $cursor->fetch();
+		$cursor->closeCursor();
+
+		if ($data === null) {
+			return [];
+		}
+
+		/*
+		 * This lovely like is the result of the way the new previews are stored
+		 * We take the md5 of the name (fileid) and split the first 7 chars. That way
+		 * there are not a gazillion files in the root of the preview appdata.
+		 */
+		$like = $this->connection->escapeLikeParameter($data['path']) . '/_/_/_/_/_/_/_/%';
+
+		$qb = $this->connection->getQueryBuilder();
+		$qb->select('a.name')
+			->from('filecache', 'a')
+			->leftJoin('a', 'filecache', 'b', $qb->expr()->eq(
+				$qb->expr()->castColumn('a.name', IQueryBuilder::PARAM_INT), 'b.fileid'
+			))
+			->where(
+				$qb->expr()->andX(
+					$qb->expr()->isNull('b.fileid'),
+					$qb->expr()->like('a.path', $qb->createNamedParameter($like)),
+					$qb->expr()->eq('a.mimetype', $qb->createNamedParameter($this->mimeTypeLoader->getId('httpd/unix-directory')))
+				)
+			);
+
+		if (!$this->isCLI) {
+			$qb->setMaxResults(10);
+		}
+
+		$cursor = $qb->execute();
+
+		while ($row = $cursor->fetch()) {
+			yield $row['name'];
 		}
 
 		$cursor->closeCursor();
diff --git a/lib/private/Preview/Storage/Root.php b/lib/private/Preview/Storage/Root.php
new file mode 100644
index 00000000000..0e9d92a44c2
--- /dev/null
+++ b/lib/private/Preview/Storage/Root.php
@@ -0,0 +1,71 @@
+<?php
+
+declare(strict_types=1);
+/**
+ * @copyright Copyright (c) 2020, Roeland Jago Douma <roeland@famdouma.nl>
+ *
+ * @author Roeland Jago Douma <roeland@famdouma.nl>
+ *
+ * @license GNU AGPL version 3 or any later version
+ *
+ * This program is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU Affero General Public License as
+ * published by the Free Software Foundation, either version 3 of the
+ * License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU Affero General Public License for more details.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ *
+ */
+
+namespace OC\Preview\Storage;
+
+use OC\Files\AppData\AppData;
+use OC\SystemConfig;
+use OCP\Files\IRootFolder;
+use OCP\Files\NotFoundException;
+use OCP\Files\SimpleFS\ISimpleFolder;
+
+class Root extends AppData {
+	public function __construct(IRootFolder $rootFolder, SystemConfig $systemConfig) {
+		parent::__construct($rootFolder, $systemConfig, 'preview');
+	}
+
+
+	public function getFolder(string $name): ISimpleFolder {
+		$internalFolder = $this->getInternalFolder($name);
+
+		try {
+			return parent::getFolder($internalFolder);
+		} catch (NotFoundException $e) {
+			/*
+			 * The new folder structure is not found.
+			 * Lets try the old one
+			 */
+		}
+
+		return parent::getFolder($name);
+	}
+
+	public function newFolder(string $name): ISimpleFolder {
+		$internalFolder = $this->getInternalFolder($name);
+		return parent::newFolder($internalFolder);
+	}
+
+	/*
+	 * Do not allow directory listing on this special root
+	 * since it gets to big and time consuming
+	 */
+	public function getDirectoryListing(): array {
+		return [];
+	}
+
+	private function getInternalFolder(string $name): string {
+		return implode('/', str_split(substr(md5($name), 0, 7))) . '/' . $name;
+	}
+}
diff --git a/lib/private/Server.php b/lib/private/Server.php
index cf4f8cc097b..4fe89e3098e 100644
--- a/lib/private/Server.php
+++ b/lib/private/Server.php
@@ -281,7 +281,7 @@ class Server extends ServerContainer implements IServerContainer {
 			return new PreviewManager(
 				$c->getConfig(),
 				$c->getRootFolder(),
-				$c->getAppDataDir('preview'),
+				new \OC\Preview\Storage\Root($c->getRootFolder(), $c->getSystemConfig(), 'preview'),
 				$c->getEventDispatcher(),
 				$c->getGeneratorHelper(),
 				$c->getSession()->get('user_id')
@@ -291,7 +291,7 @@ class Server extends ServerContainer implements IServerContainer {
 
 		$this->registerService(\OC\Preview\Watcher::class, function (Server $c) {
 			return new \OC\Preview\Watcher(
-				$c->getAppDataDir('preview')
+				new \OC\Preview\Storage\Root($c->getRootFolder(), $c->getSystemConfig(), 'preview')
 			);
 		});