The first draft of the Preprocessing API.

llama_stack/providers/inline/preprocessing/docling/__init__.py

@@ -0,0 +1,18 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from .config import InlineDoclingConfig
+
+
+async def get_provider_impl(
+    config: InlineDoclingConfig,
+    _deps,
+):
+    from .docling import InclineDoclingPreprocessorImpl
+
+    impl = InclineDoclingPreprocessorImpl(config)
+    await impl.initialize()
+    return impl

llama_stack/providers/inline/preprocessing/docling/config.py

@@ -0,0 +1,9 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+from pydantic import BaseModel
+
+
+class InlineDoclingConfig(BaseModel): ...

llama_stack/providers/inline/preprocessing/docling/docling.py

@@ -0,0 +1,36 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+from typing import List
+
+from llama_stack.apis.preprocessing import (
+    Preprocessing,
+    PreprocessingInput,
+    PreprocessingResponse,
+    Preprocessor,
+    PreprocessorOptions,
+)
+from llama_stack.providers.datatypes import PreprocessorsProtocolPrivate
+from llama_stack.providers.inline.preprocessing.docling import InlineDoclingConfig
+
+
+class InclineDoclingPreprocessorImpl(Preprocessing, PreprocessorsProtocolPrivate):
+    def __init__(self, config: InlineDoclingConfig) -> None:
+        self.config = config
+
+    async def initialize(self) -> None: ...
+
+    async def shutdown(self) -> None: ...
+
+    async def register_preprocessor(self, preprocessor: Preprocessor) -> None: ...
+
+    async def unregister_preprocessor(self, preprocessor_id: str) -> None: ...
+
+    async def preprocess(
+        self,
+        preprocessor_id: str,
+        preprocessor_inputs: List[PreprocessingInput],
+        options: PreprocessorOptions,
+    ) -> PreprocessingResponse: ...