Neenja Documentation / Internal Runtime Functions

Internal

Internal Runtime Functions

Internal runtime helpers resolve document sources, parse recognized files, and build reader models.

parserruntimereaderinternals

Related: Parser Pipeline Reader Navigation Internals Documentation Model Types

These functions are implementation details of the bundled reader, not part of a stable public API.

function

`readDocumentCollection`

readDocumentCollection(): Promise<DocumentCollection>

Load all recognized Neenja documents and return the full reader collection.

Behavior: no visible concepts.
Resolves the documents directory.
Reads recognized document files.
Parses documentation, project plan, and task tree documents.
Builds documents, documentsBySlug, and defaultDocument.
Uses the project plan as defaultDocument when documentation exists but has

function

`readDocumentationDocument`

readDocumentationDocument(): Promise<DocumentationDocument>

Load and parse the documentation document only.

Behavior: Supports single-document reader callers.
Reads NEENJA_DOCUMENTATION_PATH or the default documentation path.
Parses concepts, filters by visibility, and builds a type index.

function

`parseConcept`

parseConcept(block: string): Concept

Parse one ## Concept: block into the in-memory documentation concept model.

Behavior: Reads concept metadata such as ID, Privacy, Type, and Category.
Routes the body through the Markdown parser or structured entry parser based on ConceptKind.

function

`parsePlanSection`

parsePlanSection(block: string): PlanSection

Parse one ## Plan: block into a technical project-plan section.

Behavior: Reads ID, Area, and optional Summary.
Keeps non-metadata section content as Markdown.
Parses body content before the first ### heading as intro Markdown.
Parses each ### heading as a PlanDetailBlock.

function

`parsePlanDetailBlock`

parsePlanDetailBlock(block: string, usedIds: Set<string>): PlanDetailBlock

Parse one ### technical detail block inside a project-plan section.

Behavior: Derives a unique block ID from the heading.
Keeps the block body as Markdown content.

function

`parsePlanMetadataFields`

parsePlanMetadataFields(lines: string[]): { metadata: Record<string, string>; bodyStartIndex: number }

Parse the small metadata header at the start of a project-plan section.

Behavior: Accepts only ID, Area, and Summary as project-plan metadata.
Stops before the first non-metadata line so it can render as Markdown body content.
Returns the index where body parsing should continue.

function

`parseYamlTaskNode`

parseYamlTaskNode(rawTask: RawTaskNode, parentId: string | undefined, usedIds: Set<string>): { task: TaskNode; children: RawTaskNode[] }

Parse one YAML task object into a task node before derived graph relationships are applied.

Behavior: children.
Reads id, title, status, area, dependsOn, details, and nested
Normalizes status aliases such as doing to in-progress and completed to done.
Parses details as Markdown content.

function

`resolveDocumentDirectoryPath`

resolveDocumentDirectoryPath(): Promise<string>

Resolve the directory that should be scanned for recognized Neenja documents.

Behavior: Uses NEENJA_DOCUMENTS_DIR or NEENJA_DOCUMENTS_PATH when present.
Falls back to ${NEENJA_PROJECT_ROOT}/.neenja or ${process.cwd()}/.neenja.

function

`buildTypeIndex`

buildTypeIndex(concepts: Concept[]): Record<string, TypeReferenceTarget>

Collect the first visible occurrence of each documented type name so inline type mentions can navigate to the matching type entry.