Kreuzberg.Element (kreuzberg v4.4.2)

Semantic element extracted from a document.

Represents a logical unit of content with semantic classification, unique identifier, and metadata for tracking origin and position. Compatible with Unstructured.io element format.

Element Types

Elements are classified into semantic categories:

:title - Document title
:narrative_text - Main narrative text body
:heading - Section heading
:list_item - List item (bullet, numbered, etc.)
:table - Table element
:image - Image element
:page_break - Page break marker
:code_block - Code block
:block_quote - Block quote
:footer - Footer text
:header - Header text

Fields

:element_id - Unique deterministic identifier for the element
:element_type - Semantic type classification (atom)
:text - Text content of the element
:metadata - ElementMetadata struct with position and source info

Examples

iex> element = %Kreuzberg.Element{
...>   element_id: "elem-12345",
...>   element_type: :narrative_text,
...>   text: "This is a paragraph of text.",
...>   metadata: %Kreuzberg.ElementMetadata{
...>     page_number: 1,
...>     filename: "document.pdf"
...>   }
...> }
iex> element.element_type
:narrative_text
iex> element.text
"This is a paragraph of text."

Summary

Types

element_type()

Semantic element type classification

t()

Functions

from_map(data)

Creates an Element struct from a map.

to_map(element)

Converts an Element struct to a map.

type_description(arg1)

Returns a human-readable description of the element type.

Types

element_type()

@type element_type() ::
  :title
  | :narrative_text
  | :heading
  | :list_item
  | :table
  | :image
  | :page_break
  | :code_block
  | :block_quote
  | :footer
  | :header

Semantic element type classification

t()

@type t() :: %Kreuzberg.Element{
  element_id: String.t(),
  element_type: element_type(),
  metadata: Kreuzberg.ElementMetadata.t(),
  text: String.t()
}

Functions

from_map(data)

@spec from_map(map()) :: t()

Creates an Element struct from a map.

Converts a plain map (typically from NIF/Rust) into a proper struct, handling nested metadata and type conversion.

Parameters

data - A map containing element fields

Returns

An Element struct with properly typed fields.

Examples

iex> element_map = %{
...>   "element_id" => "elem-123",
...>   "element_type" => "narrative_text",
...>   "text" => "Paragraph content",
...>   "metadata" => %{"page_number" => 1}
...> }
iex> element = Kreuzberg.Element.from_map(element_map)
iex> element.element_type
:narrative_text

to_map(element)

@spec to_map(t()) :: map()

Converts an Element struct to a map.

Useful for serialization and passing to external systems.

Parameters

element - An Element struct

Returns

A map with string keys representing all fields. Element type is converted to string.

Examples

iex> element = %Kreuzberg.Element{
...>   element_id: "elem-123",
...>   element_type: :narrative_text,
...>   text: "Content",
...>   metadata: %Kreuzberg.ElementMetadata{}
...> }
iex> map = Kreuzberg.Element.to_map(element)
iex> map["element_type"]
"narrative_text"

type_description(arg1)

@spec type_description(element_type()) :: String.t()

Returns a human-readable description of the element type.

Parameters

element_type - An element type atom

Returns

A string description of the element type.

Examples

iex> Kreuzberg.Element.type_description(:narrative_text)
"Narrative text"

iex> Kreuzberg.Element.type_description(:code_block)
"Code block"