captpar

captpar is a hypothetical data model used in digital publishing to encode a single paragraph of text together with its associated caption and metadata. The term is not tied to a formal standard and is used in discussions of document structure, accessibility, and content reuse to illustrate how paragraph-level information can be packaged for machine readability.

it presents a paragraph as a unit with fields such as id, text, language, style, and version, plus optional caption fields that attach to the paragraph or to a linked figure or table. a caption field might include caption_text, caption_source, and caption_role. additional metadata may include provenance, edit history, formatting hints such as alignment, indentation, emphasis, and references to related content blocks.

serialization and interoperability: a paragraph unit can be serialized in formats such as JSON, JSON-LD, or XML. implementations typically expose an API for creating, updating, and rendering captpar units, and systems can index and translate paragraphs using the metadata.

uses and benefits: it is used in content management systems, editorial workflows, e-learning platforms, and accessibility tooling to improve captioning, translation, archiving, and content reuse. by packaging paragraph text with captions and metadata, captpar supports consistent rendering across devices and helps meet accessibility requirements such as screen readers.

status and adoption: as of now captpar is not an official standard; it remains a concept explored in research and prototype tooling.