Implement Markdown FormaAST Parser

Goal

Implement P0 Markdown reading and FormaAST enrichment for entries and views.

Sources

• architecture/forma-core-technical-direction
• decisions/forma-p0-core-architecture
• Historical parser spike report: knowledge/workspace/tiscs/research/forma-markdown-parser-spike-report.md

Context

Forma is parser-and-renderer-first, not Markdown-editor-first. P0 should split frontmatter and body, parse YAML frontmatter, parse Markdown with the selected AST parser, and enrich the result with Forma-specific references and directives.

In Scope

• Split frontmatter and Markdown body without rewriting source.
• Parse YAML frontmatter into generic values for entries and typed values for config where appropriate.
• Integrate the chosen Markdown AST parser.
• Detect ordinary wikilinks, Markdown links, Obsidian-style embeds, and Forma comment directives such as <!-- forma-view -->.
• Produce source locations when available and fall back to field/file locations when not.
• Add fixtures for valid Markdown, invalid frontmatter, wikilinks, embeds, and view mount comments.

Out Of Scope

• Markdown body rewriting.
• Rich Markdown editing.
• Expanding embedded content.
• Transclusion cycle handling.

Acceptance Criteria

• Parser fixture tests produce stable structured output.
• Obsidian-style embeds are represented as embed intent, not expanded content.
• View mount comments are detected without requiring non-Markdown syntax.
• Source location behavior is documented and tested for the parser capabilities available in P0.

Relationship Notes

Blocked by workspace scaffold. Can proceed in parallel with Schema DSL after the scaffold exists.

Open Questions

• None.