phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-08-07 19:12:09 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Switched to using the existing Chunk type.

Browse source
This commit is contained in:
ilya-kolchinsky 2025-03-03 20:31:06 +01:00
parent c3515530bb
commit 224d2d2891
2 changed files with 5 additions and 10 deletions
Download patch file Download diff file Expand all files Collapse all files

11
llama_stack/apis/preprocessing/preprocessing.py
View file

@ -8,8 +8,9 @@ from typing import Any, Dict, List, Optional, Protocol, runtime_checkable
from pydantic import BaseModel
from llama_stack.apis.common.content_types import URL
from llama_stack.apis.common.content_types import URL, InterleavedContent
from llama_stack.apis.preprocessing.preprocessors import Preprocessor
from llama_stack.apis.vector_io import Chunk
from llama_stack.schema_utils import json_schema_type, webmethod
@ -28,16 +29,10 @@ class PreprocessingInput(BaseModel):
PreprocessorOptions = Dict[str, Any]
@json_schema_type
class PreprocessingResult(BaseModel):
metadata: dict[str, Any]
data: str
@json_schema_type
class PreprocessingResponse(BaseModel):
status: bool
results: Optional[List[str | PreprocessingResult]]
results: Optional[List[str | InterleavedContent | Chunk]]
class PreprocessorStore(Protocol):

4
llama_stack/providers/inline/preprocessing/docling/docling.py
View file

@ -13,10 +13,10 @@ from llama_stack.apis.preprocessing import (
Preprocessing,
PreprocessingInput,
PreprocessingResponse,
PreprocessingResult,
Preprocessor,
PreprocessorOptions,
)
from llama_stack.apis.vector_io import Chunk
from llama_stack.providers.datatypes import PreprocessorsProtocolPrivate
from llama_stack.providers.inline.preprocessing.docling import InlineDoclingConfig
@ -56,7 +56,7 @@ class InclineDoclingPreprocessorImpl(Preprocessing, PreprocessorsProtocolPrivate
converted_document = self.converter.convert(url).document
if self.config.chunk:
result = self.chunker.chunk(converted_document)
results.extend([PreprocessingResult(data=chunk.text, metadata=chunk.meta) for chunk in result])
results.extend([Chunk(content=chunk.text, metadata=chunk.meta) for chunk in result])
else:
result = converted_document.export_to_markdown()
results.append(result)