LLM OCP API: Implement private backend code + add ILanguageModelTask

Signed-off-by: Marcel Klehr <mklehr@gmx.net>
(cherry picked from commit 34138736538f604af2c6c52aa43662d1d66087d0)

4 files changed, 326 insertions, 0 deletions

diff --git a/lib/private/LanguageModel/Db/Task.php b/lib/private/LanguageModel/Db/Task.php
new file mode 100644
index 00000000000..cee6c2fd8b9
--- /dev/null
+++ b/lib/private/LanguageModel/Db/Task.php
@@ -0,0 +1,58 @@
+<?php
+
+namespace OC\LanguageModel\Db;
+
+use OCP\AppFramework\Db\Entity;
+use OCP\LanguageModel\ILanguageModelTask;
+
+/**
+ * @method setType(string $type)
+ * @method string getType()
+ * @method setInput(string $type)
+ * @method string getInput()
+ * @method setStatus(int $type)
+ * @method int getStatus()
+ * @method setUserId(string $type)
+ * @method string getuserId()
+ * @method setAppId(string $type)
+ * @method string getAppId()
+ */
+class Task extends Entity {
+
+	protected $type;
+	protected $input;
+	protected $status;
+	protected $userId;
+	protected $appId;
+
+	/**
+	 * @var string[]
+	 */
+	public static array $columns = ['id', 'type', 'input', 'status', 'user_id', 'app_id'];
+
+	/**
+	 * @var string[]
+	 */
+	public static array $fields = ['id', 'type', 'input', 'status', 'userId', 'appId'];
+
+
+	public function __construct() {
+		// add types in constructor
+		$this->addType('id', 'integer');
+		$this->addType('type', 'string');
+		$this->addType('input', 'string');
+		$this->addType('status', 'integer');
+		$this->addType('userId', 'string');
+		$this->addType('appId', 'string');
+	}
+
+	public static function fromLanguageModelTask(ILanguageModelTask $task): Task {
+		return Task::fromParams([
+			'type' => $task->getType(),
+			'status' => ILanguageModelTask::STATUS_UNKNOWN,
+			'input' => $task->getInput(),
+			'userId' => $task->getUserId(),
+			'appId' => $task->getAppId(),
+		]);
+	}
+}
diff --git a/lib/private/LanguageModel/Db/TaskMapper.php b/lib/private/LanguageModel/Db/TaskMapper.php
new file mode 100644
index 00000000000..0b9004c4d96
--- /dev/null
+++ b/lib/private/LanguageModel/Db/TaskMapper.php
@@ -0,0 +1,34 @@
+<?php
+
+namespace OC\LanguageModel\Db;
+
+use OCP\AppFramework\Db\DoesNotExistException;
+use OCP\AppFramework\Db\MultipleObjectsReturnedException;
+use OCP\AppFramework\Db\QBMapper;
+use OCP\DB\Exception;
+use OCP\IDBConnection;
+
+/**
+ * @extends QBMapper<Task>
+ */
+class TaskMapper extends QBMapper {
+
+	public function __construct(IDBConnection $db) {
+		parent::__construct($db, 'oc_llm_tasks', Task::class);
+	}
+
+	/**
+	 * @param int $id
+	 * @return Task
+	 * @throws Exception
+	 * @throws DoesNotExistException
+	 * @throws MultipleObjectsReturnedException
+	 */
+	public function find(int $id): Task {
+		$qb = $this->db->getQueryBuilder();
+		$qb->select(Task::$columns)
+			->from($this->tableName)
+			->where($qb->expr()->eq('id', $qb->createPositionalParameter($id)));
+		return $this->findEntity($qb);
+	}
+}
diff --git a/lib/private/LanguageModel/LanguageModelManager.php b/lib/private/LanguageModel/LanguageModelManager.php
new file mode 100644
index 00000000000..f9f13b15d6e
--- /dev/null
+++ b/lib/private/LanguageModel/LanguageModelManager.php
@@ -0,0 +1,162 @@
+<?php
+
+namespace OC\LanguageModel;
+
+use OC\AppFramework\Bootstrap\Coordinator;
+use OC\LanguageModel\Db\Task;
+use OC\LanguageModel\Db\TaskMapper;
+use OCP\LanguageModel\AbstractLanguageModelTask;
+use OCP\LanguageModel\FreePromptTask;
+use OCP\LanguageModel\SummaryTask;
+use OCP\AppFramework\Db\DoesNotExistException;
+use OCP\AppFramework\Db\MultipleObjectsReturnedException;
+use OCP\BackgroundJob\IJobList;
+use OCP\DB\Exception;
+use OCP\IServerContainer;
+use OCP\LanguageModel\ILanguageModelManager;
+use OCP\LanguageModel\ILanguageModelProvider;
+use OCP\LanguageModel\ILanguageModelTask;
+use OCP\LanguageModel\ISummaryProvider;
+use OCP\PreConditionNotMetException;
+use Psr\Container\ContainerExceptionInterface;
+use Psr\Container\NotFoundExceptionInterface;
+use Psr\Log\LoggerInterface;
+use RuntimeException;
+use Throwable;
+
+class LanguageModelManager implements ILanguageModelManager {
+
+	/** @var ?ILanguageModelProvider[] */
+	private ?array $providers = null;
+
+	public function __construct(
+		private IServerContainer $serverContainer,
+		private Coordinator $coordinator,
+		private LoggerInterface $logger,
+		private IJobList $jobList,
+		private TaskMapper $taskMapper,
+	) {
+	}
+
+	public function getProviders(): array {
+		$context = $this->coordinator->getRegistrationContext();
+		if ($context === null) {
+			return [];
+		}
+
+		if ($this->providers !== null) {
+			return $this->providers;
+		}
+
+		$this->providers = [];
+
+		foreach ($context->getSpeechToTextProviders() as $providerServiceRegistration) {
+			$class = $providerServiceRegistration->getService();
+			try {
+				$this->providers[$class] = $this->serverContainer->get($class);
+			} catch (NotFoundExceptionInterface|ContainerExceptionInterface|Throwable $e) {
+				$this->logger->error('Failed to load LanguageModel provider ' . $class, [
+					'exception' => $e,
+				]);
+			}
+		}
+
+		return $this->providers;
+	}
+
+	public function hasProviders(): bool {
+		$context = $this->coordinator->getRegistrationContext();
+		if ($context === null) {
+			return false;
+		}
+		return !empty($context->getSpeechToTextProviders());
+	}
+
+	/**
+	 * @inheritDoc
+	 */
+	public function getAvailableTasks(): array {
+		$tasks = [];
+		foreach ($this->getProviders() as $provider) {
+			$tasks[FreePromptTask::class] = true;
+			if ($provider instanceof ISummaryProvider) {
+				$tasks[SummaryTask::class] = true;
+			}
+		}
+		return array_keys($tasks);
+	}
+
+	public function canHandleTask(ILanguageModelTask $task): bool {
+		return !empty(array_filter($this->getAvailableTasks(), fn ($class) => $task instanceof $class));
+	}
+
+	/**
+	 * @inheritDoc
+	 */
+	public function runTask(ILanguageModelTask $task): string {
+		if (!$this->canHandleTask($task)) {
+			throw new PreConditionNotMetException('No LanguageModel provider is installed that can handle this task');
+		}
+		foreach ($this->getProviders() as $provider) {
+			if (!$task->canUseProvider($provider)) {
+				continue;
+			}
+			try {
+				$task->setStatus(ILanguageModelTask::STATUS_RUNNING);
+				$this->taskMapper->update(Task::fromLanguageModelTask($task));
+				$output = $task->visitProvider($provider);
+				$task->setStatus(ILanguageModelTask::STATUS_SUCCESSFUL);
+				$this->taskMapper->update(Task::fromLanguageModelTask($task));
+				return $output;
+			} catch (\RuntimeException $e) {
+				$this->logger->info('LanguageModel call using provider ' . $provider->getName() . ' failed', ['exception' => $e]);
+				$task->setStatus(ILanguageModelTask::STATUS_FAILED);
+				$this->taskMapper->update(Task::fromLanguageModelTask($task));
+				throw $e;
+			} catch (\Throwable $e) {
+				$this->logger->info('LanguageModel call using provider ' . $provider->getName() . ' failed', ['exception' => $e]);
+				$task->setStatus(ILanguageModelTask::STATUS_FAILED);
+				$this->taskMapper->update(Task::fromLanguageModelTask($task));
+				throw new RuntimeException('LanguageModel call using provider ' . $provider->getName() . ' failed: ' . $e->getMessage());
+			}
+		}
+
+		throw new RuntimeException('Could not transcribe file');
+	}
+
+	/**
+	 * @inheritDoc
+	 * @throws Exception
+	 */
+	public function scheduleTask(ILanguageModelTask $task): void {
+		if (!$this->canHandleTask($task)) {
+			throw new PreConditionNotMetException('No LanguageModel provider is installed that can handle this task');
+		}
+		$taskEntity = Task::fromLanguageModelTask($task);
+		$this->taskMapper->insert($taskEntity);
+		$task->setId($taskEntity->getId());
+		$task->setStatus(ILanguageModelTask::STATUS_SCHEDULED);
+		$this->jobList->add(TaskBackgroundJob::class, [
+			'taskId' => $task->getId()
+		]);
+	}
+
+	/**
+	 * @param int $id The id of the task
+	 * @return ILanguageModelTask
+	 * @throws RuntimeException If the query failed
+	 * @throws \ValueError If the task could not be found
+	 */
+	public function getTask(int $id): ILanguageModelTask {
+		try {
+			$taskEntity = $this->taskMapper->find($id);
+			return AbstractLanguageModelTask::fromTaskEntity($taskEntity);
+		} catch (DoesNotExistException $e) {
+			throw new \ValueError('Could not find task with the provided id');
+		} catch (MultipleObjectsReturnedException $e) {
+			throw new RuntimeException('Could not uniquely identify task with given id');
+		} catch (Exception $e) {
+			throw new RuntimeException('Failure while trying to find task by id: '.$e->getMessage());
+		}
+	}
+}
diff --git a/lib/private/LanguageModel/TaskBackgroundJob.php b/lib/private/LanguageModel/TaskBackgroundJob.php
new file mode 100644
index 00000000000..55413ba3714
--- /dev/null
+++ b/lib/private/LanguageModel/TaskBackgroundJob.php
@@ -0,0 +1,72 @@
+<?php
+
+declare(strict_types=1);
+
+/**
+ * @copyright Copyright (c) 2023 Marcel Klehr <mklehr@gmx.net>
+ *
+ * @author Marcel Klehr <mklehr@gmx.net>
+ *
+ * @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\LanguageModel;
+
+use OC\User\NoUserException;
+use OCP\AppFramework\Utility\ITimeFactory;
+use OCP\BackgroundJob\QueuedJob;
+use OCP\EventDispatcher\IEventDispatcher;
+use OCP\Files\File;
+use OCP\Files\IRootFolder;
+use OCP\Files\NotFoundException;
+use OCP\Files\NotPermittedException;
+use OCP\LanguageModel\Events\TaskFailedEvent;
+use OCP\LanguageModel\Events\TaskSuccessfulEvent;
+use OCP\LanguageModel\ILanguageModelManager;
+use OCP\PreConditionNotMetException;
+use OCP\SpeechToText\Events\TranscriptionFailedEvent;
+use OCP\SpeechToText\Events\TranscriptionSuccessfulEvent;
+use OCP\SpeechToText\ISpeechToTextManager;
+use Psr\Log\LoggerInterface;
+
+class TaskBackgroundJob extends QueuedJob {
+	public function __construct(
+		ITimeFactory $timeFactory,
+		private ILanguageModelManager $languageModelManager,
+		private IEventDispatcher $eventDispatcher,
+	) {
+		parent::__construct($timeFactory);
+		$this->setAllowParallelRuns(false);
+	}
+
+	/**
+	 * @param array{taskId: int} $argument
+	 * @inheritDoc
+	 */
+	protected function run($argument) {
+		$taskId = $argument['taskId'];
+		$task = $this->languageModelManager->getTask($taskId);
+		try {
+			$output = $this->languageModelManager->runTask($task);
+			$event = new TaskSuccessfulEvent($task, $output);
+
+		} catch (\RuntimeException|PreConditionNotMetException $e) {
+			$event = new TaskFailedEvent($task, $e->getMessage());
+		}
+		$this->eventDispatcher->dispatchTyped($event);
+	}
+}