索引

RefDocInfo dataclass

基类：DataClassJsonMixin

表示已摄取文档的 Dataclass。

参数

名称	类型	描述	默认值
node_ids	List	内置可变序列。 如果没有给定参数，构造函数将创建一个新的空列表。如果指定了参数，则必须是可迭代对象。	<动态>

源代码位于 llama-index-core/llama_index/core/storage/docstore/types.py

@dataclass
class RefDocInfo(DataClassJsonMixin):
    """Dataclass to represent ingested documents."""

    node_ids: List = field(default_factory=list)
    metadata: Dict[str, Any] = field(default_factory=dict)

BaseDocumentStore

基类：ABC

源代码位于 llama-index-core/llama_index/core/storage/docstore/types.py

class BaseDocumentStore(ABC):
    # ===== Save/load =====
    def persist(
        self,
        persist_path: str = DEFAULT_PERSIST_PATH,
        fs: Optional[fsspec.AbstractFileSystem] = None,
    ) -> None:
        """Persist the docstore to a file."""

    # ===== Main interface =====
    @property
    @abstractmethod
    def docs(self) -> Dict[str, BaseNode]:
        ...

    @abstractmethod
    def add_documents(
        self,
        docs: Sequence[BaseNode],
        allow_update: bool = True,
        batch_size: int = DEFAULT_BATCH_SIZE,
        store_text: bool = True,
    ) -> None:
        ...

    @abstractmethod
    async def async_add_documents(
        self,
        docs: Sequence[BaseNode],
        allow_update: bool = True,
        batch_size: int = DEFAULT_BATCH_SIZE,
        store_text: bool = True,
    ) -> None:
        ...

    @abstractmethod
    def get_document(self, doc_id: str, raise_error: bool = True) -> Optional[BaseNode]:
        ...

    @abstractmethod
    async def aget_document(
        self, doc_id: str, raise_error: bool = True
    ) -> Optional[BaseNode]:
        ...

    @abstractmethod
    def delete_document(self, doc_id: str, raise_error: bool = True) -> None:
        """Delete a document from the store."""
        ...

    @abstractmethod
    async def adelete_document(self, doc_id: str, raise_error: bool = True) -> None:
        """Delete a document from the store."""
        ...

    @abstractmethod
    def document_exists(self, doc_id: str) -> bool:
        ...

    @abstractmethod
    async def adocument_exists(self, doc_id: str) -> bool:
        ...

    # ===== Hash =====
    @abstractmethod
    def set_document_hash(self, doc_id: str, doc_hash: str) -> None:
        ...

    @abstractmethod
    async def aset_document_hash(self, doc_id: str, doc_hash: str) -> None:
        ...

    @abstractmethod
    def set_document_hashes(self, doc_hashes: Dict[str, str]) -> None:
        ...

    @abstractmethod
    async def aset_document_hashes(self, doc_hashes: Dict[str, str]) -> None:
        ...

    @abstractmethod
    def get_document_hash(self, doc_id: str) -> Optional[str]:
        ...

    @abstractmethod
    async def aget_document_hash(self, doc_id: str) -> Optional[str]:
        ...

    @abstractmethod
    def get_all_document_hashes(self) -> Dict[str, str]:
        ...

    @abstractmethod
    async def aget_all_document_hashes(self) -> Dict[str, str]:
        ...

    # ==== Ref Docs =====
    @abstractmethod
    def get_all_ref_doc_info(self) -> Optional[Dict[str, RefDocInfo]]:
        """Get a mapping of ref_doc_id -> RefDocInfo for all ingested documents."""

    @abstractmethod
    async def aget_all_ref_doc_info(self) -> Optional[Dict[str, RefDocInfo]]:
        """Get a mapping of ref_doc_id -> RefDocInfo for all ingested documents."""

    @abstractmethod
    def get_ref_doc_info(self, ref_doc_id: str) -> Optional[RefDocInfo]:
        """Get the RefDocInfo for a given ref_doc_id."""

    @abstractmethod
    async def aget_ref_doc_info(self, ref_doc_id: str) -> Optional[RefDocInfo]:
        """Get the RefDocInfo for a given ref_doc_id."""

    @abstractmethod
    def delete_ref_doc(self, ref_doc_id: str, raise_error: bool = True) -> None:
        """Delete a ref_doc and all it's associated nodes."""

    @abstractmethod
    async def adelete_ref_doc(self, ref_doc_id: str, raise_error: bool = True) -> None:
        """Delete a ref_doc and all it's associated nodes."""

    # ===== Nodes =====
    def get_nodes(
        self, node_ids: List[str], raise_error: bool = True
    ) -> List[BaseNode]:
        """
        Get nodes from docstore.

        Args:
            node_ids (List[str]): node ids
            raise_error (bool): raise error if node_id not found

        """
        return [self.get_node(node_id, raise_error=raise_error) for node_id in node_ids]

    async def aget_nodes(
        self, node_ids: List[str], raise_error: bool = True
    ) -> List[BaseNode]:
        """
        Get nodes from docstore.

        Args:
            node_ids (List[str]): node ids
            raise_error (bool): raise error if node_id not found

        """
        return [
            await self.aget_node(node_id, raise_error=raise_error)
            for node_id in node_ids
        ]

    def get_node(self, node_id: str, raise_error: bool = True) -> BaseNode:
        """
        Get node from docstore.

        Args:
            node_id (str): node id
            raise_error (bool): raise error if node_id not found

        """
        doc = self.get_document(node_id, raise_error=raise_error)
        if not isinstance(doc, BaseNode):
            raise ValueError(f"Document {node_id} is not a Node.")
        return doc

    async def aget_node(self, node_id: str, raise_error: bool = True) -> BaseNode:
        """
        Get node from docstore.

        Args:
            node_id (str): node id
            raise_error (bool): raise error if node_id not found

        """
        doc = await self.aget_document(node_id, raise_error=raise_error)
        if not isinstance(doc, BaseNode):
            raise ValueError(f"Document {node_id} is not a Node.")
        return doc

    def get_node_dict(self, node_id_dict: Dict[int, str]) -> Dict[int, BaseNode]:
        """
        Get node dict from docstore given a mapping of index to node ids.

        Args:
            node_id_dict (Dict[int, str]): mapping of index to node ids

        """
        return {
            index: self.get_node(node_id) for index, node_id in node_id_dict.items()
        }

    async def aget_node_dict(self, node_id_dict: Dict[int, str]) -> Dict[int, BaseNode]:
        """
        Get node dict from docstore given a mapping of index to node ids.

        Args:
            node_id_dict (Dict[int, str]): mapping of index to node ids

        """
        return {
            index: await self.aget_node(node_id)
            for index, node_id in node_id_dict.items()
        }

persist

persist(persist_path: str = DEFAULT_PERSIST_PATH, fs: Optional[AbstractFileSystem] = None) -> None

将文档存储持久化到文件。

源代码位于 llama-index-core/llama_index/core/storage/docstore/types.py

def persist(
    self,
    persist_path: str = DEFAULT_PERSIST_PATH,
    fs: Optional[fsspec.AbstractFileSystem] = None,
) -> None:
    """Persist the docstore to a file."""

delete_document abstractmethod

delete_document(doc_id: str, raise_error: bool = True) -> None

从存储中删除文档。

源代码位于 llama-index-core/llama_index/core/storage/docstore/types.py

@abstractmethod
def delete_document(self, doc_id: str, raise_error: bool = True) -> None:
    """Delete a document from the store."""
    ...

adelete_document abstractmethod async

adelete_document(doc_id: str, raise_error: bool = True) -> None

从存储中删除文档。

源代码位于 llama-index-core/llama_index/core/storage/docstore/types.py

@abstractmethod
async def adelete_document(self, doc_id: str, raise_error: bool = True) -> None:
    """Delete a document from the store."""
    ...

get_all_ref_doc_info abstractmethod

get_all_ref_doc_info() -> Optional[Dict[str, RefDocInfo]]

获取所有已摄取文档的 ref_doc_id -> RefDocInfo 映射。

源代码位于 llama-index-core/llama_index/core/storage/docstore/types.py

@abstractmethod
def get_all_ref_doc_info(self) -> Optional[Dict[str, RefDocInfo]]:
    """Get a mapping of ref_doc_id -> RefDocInfo for all ingested documents."""

aget_all_ref_doc_info abstractmethod async

aget_all_ref_doc_info() -> Optional[Dict[str, RefDocInfo]]

获取所有已摄取文档的 ref_doc_id -> RefDocInfo 映射。

源代码位于 llama-index-core/llama_index/core/storage/docstore/types.py

@abstractmethod
async def aget_all_ref_doc_info(self) -> Optional[Dict[str, RefDocInfo]]:
    """Get a mapping of ref_doc_id -> RefDocInfo for all ingested documents."""

get_ref_doc_info abstractmethod

get_ref_doc_info(ref_doc_id: str) -> Optional[RefDocInfo]

获取给定 ref_doc_id 的 RefDocInfo。

源代码位于 llama-index-core/llama_index/core/storage/docstore/types.py

@abstractmethod
def get_ref_doc_info(self, ref_doc_id: str) -> Optional[RefDocInfo]:
    """Get the RefDocInfo for a given ref_doc_id."""

aget_ref_doc_info abstractmethod async

aget_ref_doc_info(ref_doc_id: str) -> Optional[RefDocInfo]

获取给定 ref_doc_id 的 RefDocInfo。

源代码位于 llama-index-core/llama_index/core/storage/docstore/types.py

@abstractmethod
async def aget_ref_doc_info(self, ref_doc_id: str) -> Optional[RefDocInfo]:
    """Get the RefDocInfo for a given ref_doc_id."""

delete_ref_doc abstractmethod

delete_ref_doc(ref_doc_id: str, raise_error: bool = True) -> None

删除 ref_doc 及其所有关联节点。

源代码位于 llama-index-core/llama_index/core/storage/docstore/types.py

@abstractmethod
def delete_ref_doc(self, ref_doc_id: str, raise_error: bool = True) -> None:
    """Delete a ref_doc and all it's associated nodes."""

adelete_ref_doc abstractmethod async

adelete_ref_doc(ref_doc_id: str, raise_error: bool = True) -> None

删除 ref_doc 及其所有关联节点。

源代码位于 llama-index-core/llama_index/core/storage/docstore/types.py

@abstractmethod
async def adelete_ref_doc(self, ref_doc_id: str, raise_error: bool = True) -> None:
    """Delete a ref_doc and all it's associated nodes."""

get_nodes

get_nodes(node_ids: List[str], raise_error: bool = True) -> List[BaseNode]

从文档存储获取节点。

参数

名称	类型	描述	默认值
node_ids	List[str]	节点 ID	必需
raise_error	bool	如果未找到 node_id 则抛出错误	True

源代码位于 llama-index-core/llama_index/core/storage/docstore/types.py

def get_nodes(
    self, node_ids: List[str], raise_error: bool = True
) -> List[BaseNode]:
    """
    Get nodes from docstore.

    Args:
        node_ids (List[str]): node ids
        raise_error (bool): raise error if node_id not found

    """
    return [self.get_node(node_id, raise_error=raise_error) for node_id in node_ids]

aget_nodes async

aget_nodes(node_ids: List[str], raise_error: bool = True) -> List[BaseNode]

从文档存储获取节点。

参数

名称	类型	描述	默认值
node_ids	List[str]	节点 ID	必需
raise_error	bool	如果未找到 node_id 则抛出错误	True

源代码位于 llama-index-core/llama_index/core/storage/docstore/types.py

async def aget_nodes(
    self, node_ids: List[str], raise_error: bool = True
) -> List[BaseNode]:
    """
    Get nodes from docstore.

    Args:
        node_ids (List[str]): node ids
        raise_error (bool): raise error if node_id not found

    """
    return [
        await self.aget_node(node_id, raise_error=raise_error)
        for node_id in node_ids
    ]

get_node

get_node(node_id: str, raise_error: bool = True) -> BaseNode

从文档存储中获取节点。

参数

名称	类型	描述	默认值
node_id	str	节点ID	必需
raise_error	bool	如果未找到 node_id 则抛出错误	True

源代码位于 llama-index-core/llama_index/core/storage/docstore/types.py

def get_node(self, node_id: str, raise_error: bool = True) -> BaseNode:
    """
    Get node from docstore.

    Args:
        node_id (str): node id
        raise_error (bool): raise error if node_id not found

    """
    doc = self.get_document(node_id, raise_error=raise_error)
    if not isinstance(doc, BaseNode):
        raise ValueError(f"Document {node_id} is not a Node.")
    return doc

aget_node async

aget_node(node_id: str, raise_error: bool = True) -> BaseNode

从文档存储中获取节点。

参数

名称	类型	描述	默认值
node_id	str	节点ID	必需
raise_error	bool	如果未找到 node_id 则抛出错误	True

源代码位于 llama-index-core/llama_index/core/storage/docstore/types.py

async def aget_node(self, node_id: str, raise_error: bool = True) -> BaseNode:
    """
    Get node from docstore.

    Args:
        node_id (str): node id
        raise_error (bool): raise error if node_id not found

    """
    doc = await self.aget_document(node_id, raise_error=raise_error)
    if not isinstance(doc, BaseNode):
        raise ValueError(f"Document {node_id} is not a Node.")
    return doc

get_node_dict

get_node_dict(node_id_dict: Dict[int, str]) -> Dict[int, BaseNode]

从文档存储中获取节点字典，根据索引到节点ID的映射。

参数

名称	类型	描述	默认值
node_id_dict	Dict[int, str]	索引到节点ID的映射	必需

源代码位于 llama-index-core/llama_index/core/storage/docstore/types.py

def get_node_dict(self, node_id_dict: Dict[int, str]) -> Dict[int, BaseNode]:
    """
    Get node dict from docstore given a mapping of index to node ids.

    Args:
        node_id_dict (Dict[int, str]): mapping of index to node ids

    """
    return {
        index: self.get_node(node_id) for index, node_id in node_id_dict.items()
    }

aget_node_dict async

aget_node_dict(node_id_dict: Dict[int, str]) -> Dict[int, BaseNode]

从文档存储中获取节点字典，根据索引到节点ID的映射。

参数

名称	类型	描述	默认值
node_id_dict	Dict[int, str]	索引到节点ID的映射	必需

源代码位于 llama-index-core/llama_index/core/storage/docstore/types.py

async def aget_node_dict(self, node_id_dict: Dict[int, str]) -> Dict[int, BaseNode]:
    """
    Get node dict from docstore given a mapping of index to node ids.

    Args:
        node_id_dict (Dict[int, str]): mapping of index to node ids

    """
    return {
        index: await self.aget_node(node_id)
        for index, node_id in node_id_dict.items()
    }