Namespace Mythosia.Documents.Elements
Classes
- CodeItem
A code block item. Follows the docling CodeItem convention.
- DocItem
Base type for any element that carries content (can be a leaf node). Follows the docling DocItem convention.
- DocListItem
A list item within a list group. Follows the docling ListItem convention. Named DocListItem to avoid conflict with System.Collections types.
- FormulaItem
A mathematical formula item. Follows the docling FormulaItem convention.
- GridTableSerializer
Renders tables as standard Markdown pipe tables. Rows whose origin cells carry ColumnHeader = true are treated as header rows; if none exist, the first row is used as the header.
- GroupItem
A container node that groups other nodes (e.g. list container, chapter, section, slide). Follows the docling GroupItem convention. Cannot be a leaf node.
- MarkdownSerializer
Serializes a DoclingDocument to Markdown format. Walks the body tree in reading order and emits Markdown for each content item.
- NodeItem
Base class for all tree nodes in the document structure. Follows the docling NodeItem convention with self_ref, parent, and children pointers.
- PictureItem
A picture/image content item. Follows the docling PictureItem convention.
- RefItem
A JSON-pointer style reference to another item in the document. Follows the docling RefItem convention (e.g. "#/texts/0", "#/tables/1").
- SectionHeaderItem
A section heading item with a heading level. Follows the docling SectionHeaderItem convention.
- SemanticTableSerializer
Renders tables using BuildSemanticGroups(). Form-style tables (col-0 row labels, no explicit column headers) are rendered as bold-label + value lines. Grid-style tables are rendered as pipe tables.
- TableCell
Represents a single cell in a table. Follows the docling TableCell convention with span and offset information.
- TableData
Holds all table cell data and dimensions. Follows the docling TableData convention.
- TableItem
A table content item. Follows the docling TableItem convention.
- TableSemanticGroup
A set of consecutive data rows that share the same col-0 row label.
- TableSemanticView
Semantic view of a table, built from BuildSemanticGroups(). Groups rows by their col-0 row label and detects whether the table is better represented as a form/key-value structure than as a grid.
- TextItem
A text content item. Follows the docling TextItem convention. Carries both original and sanitized text representations.
- TitleItem
The document title item. Follows the docling TitleItem convention.
Interfaces
- IDocumentParser
Parses a document into a structured DoclingDocument representation.
- ITableSerializer
Strategy interface for rendering a TableItem to Markdown. Swap implementations on TableSerializer to change how tables are serialized without modifying any other code.
Enums
- ContentLayer
Content layer classification, following the docling ContentLayer convention.
- DocItemLabel
Labels for document content items, following the docling DocItemLabel convention.
- GroupLabel
Labels for group (container) items, following the docling GroupLabel convention.