50 lines
1.2 KiB
Python
50 lines
1.2 KiB
Python
"""
|
|
天目引擎协议（Protocol）。
|
|
|
|
定义统一的引擎接口，两个引擎实现（RapidOCR / PaddleOCR-VL）均遵循此协议。
|
|
调用方只需依赖协议，不依赖具体实现。
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
from typing import Protocol, runtime_checkable
|
|
|
|
|
|
@runtime_checkable
|
|
class OCREngine(Protocol):
|
|
"""OCR 引擎统一协议。"""
|
|
|
|
loaded: bool
|
|
|
|
async def load(self) -> None:
|
|
"""初始化并加载模型。"""
|
|
...
|
|
|
|
async def close(self) -> None:
|
|
"""释放模型资源。"""
|
|
...
|
|
|
|
async def health(self) -> dict:
|
|
"""返回引擎健康状态。"""
|
|
...
|
|
|
|
|
|
class RapidOCRProtocol(OCREngine, Protocol):
|
|
"""RapidOCR 引擎协议（含 recognize 方法）。"""
|
|
|
|
async def recognize(self, image_data: bytes | str) -> object:
|
|
"""执行 OCR 识别，返回结构化响应。"""
|
|
...
|
|
|
|
|
|
class PaddleOCRVlProtocol(OCREngine, Protocol):
|
|
"""PaddleOCR-VL 引擎协议（含 parse_document 方法）。"""
|
|
|
|
async def parse_document(
|
|
self,
|
|
image_data: bytes | str,
|
|
output_format: str = "markdown",
|
|
) -> object:
|
|
"""解析文档图片，返回结构化文本。"""
|
|
...
|