84 lines
1.9 KiB
Python
84 lines
1.9 KiB
Python
"""
|
||
Offline RAG Indexer module.
|
||
|
||
提供完整的离线索引构建功能，包括：
|
||
- 文档加载（PDF、Word、TXT 等）
|
||
- 文本切分（递归、语义、父子块）
|
||
- 向量嵌入（支持 llama.cpp）
|
||
- 向量存储（Qdrant）
|
||
- 父文档存储（PostgreSQL）
|
||
|
||
示例用法：
|
||
>>> from rag_indexer import IndexBuilder, IndexBuilderConfig, SplitterType
|
||
>>>
|
||
>>> config = IndexBuilderConfig(
|
||
... collection_name="my_docs",
|
||
... splitter_type=SplitterType.PARENT_CHILD,
|
||
... )
|
||
>>> builder = IndexBuilder(config)
|
||
>>>
|
||
>>> # 或直接传参（向后兼容）
|
||
>>> builder = IndexBuilder(collection_name="my_docs")
|
||
>>>
|
||
>>> await builder.build_from_file("document.pdf")
|
||
"""
|
||
|
||
from .index_builder import IndexBuilder, IndexBuilderConfig, DocstoreConfig
|
||
from .loaders import DocumentLoader
|
||
from .splitters import SplitterType, get_splitter
|
||
from .config import (
|
||
QDRANT_URL,
|
||
QDRANT_API_KEY,
|
||
LLAMACPP_EMBEDDING_URL,
|
||
LLAMACPP_API_KEY,
|
||
DB_URI,
|
||
DOCSTORE_URI,
|
||
RAG_OCR_LANGUAGES,
|
||
RAG_DOC_LANGUAGES,
|
||
)
|
||
|
||
# 从 rag_core 重新导出常用组件
|
||
import sys
|
||
from pathlib import Path
|
||
sys.path.insert(0, str(Path(__file__).parent.parent / "backend"))
|
||
|
||
from backend.rag_core import (
|
||
LlamaCppEmbedder,
|
||
QdrantVectorStore,
|
||
PostgresDocStore,
|
||
create_docstore,
|
||
)
|
||
|
||
__version__ = "2.0.0"
|
||
|
||
__all__ = [
|
||
# 核心构建器与配置
|
||
"IndexBuilder",
|
||
"IndexBuilderConfig",
|
||
"DocstoreConfig",
|
||
|
||
# 加载器
|
||
"DocumentLoader",
|
||
|
||
# 切分相关
|
||
"SplitterType",
|
||
"get_splitter",
|
||
|
||
# 配置
|
||
"QDRANT_URL",
|
||
"QDRANT_API_KEY",
|
||
"LLAMACPP_EMBEDDING_URL",
|
||
"LLAMACPP_API_KEY",
|
||
"DB_URI",
|
||
"DOCSTORE_URI",
|
||
"RAG_OCR_LANGUAGES",
|
||
"RAG_DOC_LANGUAGES",
|
||
|
||
# 嵌入与向量存储
|
||
"LlamaCppEmbedder",
|
||
"QdrantVectorStore",
|
||
|
||
# 文档存储
|
||
"PostgresDocStore",
|
||
"create_docstore",
|
||
] |