Adopt consistent, explicit identifier naming across diagrams and TypeScript APIs.

Rules 1) Diagram ids/syntax keywords

• Use lower-case diagram identifiers.
• For new/unstable syntax, append -beta to the syntax/keyword the parser accepts.
• Keep semantic parts of names uniform (don’t introduce ad-hoc prefixes like Mini or extra suffixes like Language unless the codebase has a single established pattern).

2) TypeScript naming style

• Types: use UpperCamelCase (e.g., DiagramStylesProvider).
• Interfaces: prefer interface for object shapes when possible.

3) Type-only files

• Avoid using .d.ts for internal project types.
• Use a dedicated convention like .type.ts (e.g., gantt.type.ts).

4) Exported fields/props

• Prefer explicit names for public types (e.g., width/height over w/h) to improve readability for downstream users.

5) Stability/compatibility for naming formats

• If an identifier format is consumed externally (e.g., edge id strings), treat it as a contract: don’t rename/separator-change without an explicit compatibility plan.

Example

// diagram detector keyword
const detector: DiagramDetector = (txt) => /^\s*contextMap-beta/.test(txt);

// stable diagram config key naming: one established pattern
export interface MermaidConfig {
  contextMap?: MiniContextMapLanguageDiagramConfig; // avoid “Mini…Language…” unless consistent everywhere
}

// type-only file convention
// packages/mermaid/src/diagrams/gantt/gantt.type.ts
export type DiagramStylesProvider = (options?: Record<string, unknown>) => string;

// explicit fields
export interface NodeMetaData {
  width?: string;
  height?: string;
}