Elements

ragbits.document_search.documents.element.Element

Bases: BaseModel, ABC

An object representing an element in a document.

element_type instance-attribute

element_type: str

document_meta instance-attribute

document_meta: DocumentMeta

location class-attribute instance-attribute

location: ElementLocation | None = None

id property

id: str

Retrieve the ID of the element, primarily used to represent the element's data.

RETURNS	DESCRIPTION
str	string representing element TYPE: str

key property

key: str | None

Get the representation of the element for embedding.

RETURNS	DESCRIPTION
str | None	The representation for embedding.

text_representation abstractmethod property

text_representation: str | None

Get the text representation of the element.

RETURNS	DESCRIPTION
str | None	The text representation.

image_representation property

image_representation: bytes | None

Get the image representation of the element.

RETURNS	DESCRIPTION
bytes | None	The image representation.

get_id_components

get_id_components() -> dict[str, str]

Creates a dictionary of key value pairs of id components

RETURNS	DESCRIPTION
dict	a dictionary TYPE: dict[str, str]

Source code in packages/ragbits-document-search/src/ragbits/document_search/documents/element.py

def get_id_components(self) -> dict[str, str]:
    """
    Creates a dictionary of key value pairs of id components

    Returns:
        dict: a dictionary
    """
    id_components = {
        "meta": self.document_meta.id,
        "type": self.element_type,
        "key": str(self.key),
        "text": str(self.text_representation),
        "location": str(self.location),
    }
    return id_components

from_vector_db_entry classmethod

from_vector_db_entry(db_entry: VectorStoreEntry) -> Element

Create an element from a vector database entry.

PARAMETER	DESCRIPTION
db_entry	The vector database entry. TYPE: VectorStoreEntry

RETURNS	DESCRIPTION
Element	The element.

Source code in packages/ragbits-document-search/src/ragbits/document_search/documents/element.py

@classmethod
def from_vector_db_entry(cls, db_entry: VectorStoreEntry) -> "Element":
    """
    Create an element from a vector database entry.

    Args:
        db_entry: The vector database entry.

    Returns:
        The element.
    """
    element_type = db_entry.metadata["element_type"]
    element_cls = Element._elements_registry[element_type]
    if "embedding_type" in db_entry.metadata:
        del db_entry.metadata["embedding_type"]
    return element_cls(**db_entry.metadata)

to_vector_db_entry

to_vector_db_entry() -> VectorStoreEntry

Create a vector database entry from the element.

RETURNS	DESCRIPTION
VectorStoreEntry	The vector database entry

Source code in packages/ragbits-document-search/src/ragbits/document_search/documents/element.py

def to_vector_db_entry(self) -> VectorStoreEntry:
    """
    Create a vector database entry from the element.

    Returns:
        The vector database entry
    """
    id_components = [
        self.id,
    ]
    vector_store_entry_id = uuid.uuid5(uuid.NAMESPACE_OID, ";".join(id_components))
    metadata = self.model_dump(exclude={"id", "key"})
    metadata["document_meta"]["source"]["id"] = self.document_meta.source.id

    return VectorStoreEntry(
        id=vector_store_entry_id, text=self.key, image_bytes=self.image_representation, metadata=metadata
    )

ragbits.document_search.documents.element.TextElement

Bases: Element

An object representing a text element in a document.

document_meta instance-attribute

document_meta: DocumentMeta

location class-attribute instance-attribute

location: ElementLocation | None = None

id property

id: str

Retrieve the ID of the element, primarily used to represent the element's data.

RETURNS	DESCRIPTION
str	string representing element TYPE: str

key property

key: str | None

Get the representation of the element for embedding.

RETURNS	DESCRIPTION
str | None	The representation for embedding.

image_representation property

image_representation: bytes | None

Get the image representation of the element.

RETURNS	DESCRIPTION
bytes | None	The image representation.

element_type class-attribute instance-attribute

element_type: str = 'text'

content instance-attribute

content: str

text_representation property

text_representation: str

Get the text representation of the element.

RETURNS	DESCRIPTION
str	The text representation.

get_id_components

get_id_components() -> dict[str, str]

Creates a dictionary of key value pairs of id components

RETURNS	DESCRIPTION
dict	a dictionary TYPE: dict[str, str]

Source code in packages/ragbits-document-search/src/ragbits/document_search/documents/element.py

def get_id_components(self) -> dict[str, str]:
    """
    Creates a dictionary of key value pairs of id components

    Returns:
        dict: a dictionary
    """
    id_components = {
        "meta": self.document_meta.id,
        "type": self.element_type,
        "key": str(self.key),
        "text": str(self.text_representation),
        "location": str(self.location),
    }
    return id_components

from_vector_db_entry classmethod

from_vector_db_entry(db_entry: VectorStoreEntry) -> Element

Create an element from a vector database entry.

PARAMETER	DESCRIPTION
db_entry	The vector database entry. TYPE: VectorStoreEntry

RETURNS	DESCRIPTION
Element	The element.

Source code in packages/ragbits-document-search/src/ragbits/document_search/documents/element.py

@classmethod
def from_vector_db_entry(cls, db_entry: VectorStoreEntry) -> "Element":
    """
    Create an element from a vector database entry.

    Args:
        db_entry: The vector database entry.

    Returns:
        The element.
    """
    element_type = db_entry.metadata["element_type"]
    element_cls = Element._elements_registry[element_type]
    if "embedding_type" in db_entry.metadata:
        del db_entry.metadata["embedding_type"]
    return element_cls(**db_entry.metadata)

to_vector_db_entry

to_vector_db_entry() -> VectorStoreEntry

Create a vector database entry from the element.

RETURNS	DESCRIPTION
VectorStoreEntry	The vector database entry

Source code in packages/ragbits-document-search/src/ragbits/document_search/documents/element.py

def to_vector_db_entry(self) -> VectorStoreEntry:
    """
    Create a vector database entry from the element.

    Returns:
        The vector database entry
    """
    id_components = [
        self.id,
    ]
    vector_store_entry_id = uuid.uuid5(uuid.NAMESPACE_OID, ";".join(id_components))
    metadata = self.model_dump(exclude={"id", "key"})
    metadata["document_meta"]["source"]["id"] = self.document_meta.source.id

    return VectorStoreEntry(
        id=vector_store_entry_id, text=self.key, image_bytes=self.image_representation, metadata=metadata
    )

ragbits.document_search.documents.element.ImageElement

Bases: Element

An object representing an image element in a document.

document_meta instance-attribute

document_meta: DocumentMeta

location class-attribute instance-attribute

location: ElementLocation | None = None

id property

id: str

Retrieve the ID of the element, primarily used to represent the element's data.

RETURNS	DESCRIPTION
str	string representing element TYPE: str

key property

key: str | None

Get the representation of the element for embedding.

RETURNS	DESCRIPTION
str | None	The representation for embedding.

element_type class-attribute instance-attribute

element_type: str = 'image'

image_bytes instance-attribute

image_bytes: SerializableBytes

description class-attribute instance-attribute

description: str | None = None

ocr_extracted_text class-attribute instance-attribute

ocr_extracted_text: str | None = None

text_representation property

text_representation: str | None

Get the text representation of the element.

RETURNS	DESCRIPTION
str | None	The text representation.

image_representation property

image_representation: bytes

Get the image representation of the element.

RETURNS	DESCRIPTION
bytes	The image representation.

from_vector_db_entry classmethod

from_vector_db_entry(db_entry: VectorStoreEntry) -> Element

Create an element from a vector database entry.

PARAMETER	DESCRIPTION
db_entry	The vector database entry. TYPE: VectorStoreEntry

RETURNS	DESCRIPTION
Element	The element.

Source code in packages/ragbits-document-search/src/ragbits/document_search/documents/element.py

@classmethod
def from_vector_db_entry(cls, db_entry: VectorStoreEntry) -> "Element":
    """
    Create an element from a vector database entry.

    Args:
        db_entry: The vector database entry.

    Returns:
        The element.
    """
    element_type = db_entry.metadata["element_type"]
    element_cls = Element._elements_registry[element_type]
    if "embedding_type" in db_entry.metadata:
        del db_entry.metadata["embedding_type"]
    return element_cls(**db_entry.metadata)

to_vector_db_entry

to_vector_db_entry() -> VectorStoreEntry

Create a vector database entry from the element.

RETURNS	DESCRIPTION
VectorStoreEntry	The vector database entry

Source code in packages/ragbits-document-search/src/ragbits/document_search/documents/element.py

def to_vector_db_entry(self) -> VectorStoreEntry:
    """
    Create a vector database entry from the element.

    Returns:
        The vector database entry
    """
    id_components = [
        self.id,
    ]
    vector_store_entry_id = uuid.uuid5(uuid.NAMESPACE_OID, ";".join(id_components))
    metadata = self.model_dump(exclude={"id", "key"})
    metadata["document_meta"]["source"]["id"] = self.document_meta.source.id

    return VectorStoreEntry(
        id=vector_store_entry_id, text=self.key, image_bytes=self.image_representation, metadata=metadata
    )

get_id_components

get_id_components() -> dict[str, str]

Creates a dictionary of key value pairs of id components

RETURNS	DESCRIPTION
dict	a dictionary TYPE: dict[str, str]

Source code in packages/ragbits-document-search/src/ragbits/document_search/documents/element.py

def get_id_components(self) -> dict[str, str]:
    """
    Creates a dictionary of key value pairs of id components

    Returns:
        dict: a dictionary
    """
    id_components = super().get_id_components()
    id_components["image_hash"] = hashlib.sha256(self.image_bytes).hexdigest()
    return id_components