Expand description
Structural extraction for Markdown, MDX, reStructuredText and AsciiDoc.
Prose has no token structure worth the name: a " is a quotation mark, not
a literal, and // is part of a URL. So this reads lines directly rather
than through the tokenizer, which is the honest model for the format even
though it is the opposite of what every other extractor here does.
Two facts are worth having. A heading is a named anchor other documents link to, and nesting one heading under another is the document’s own table of contents. A link to a path in the repository is a dependency exactly as an import is - which is what turns a documentation tree into part of the graph rather than a pile of files beside it.
Functions§
- extract
- Extracts structural facts from one document.