generate-ts-docs

Utilities to parse type information and JSDoc annotations from TypeScript source files, and render Markdown documentation

Usage no npm install needed!

<script type="module">
  import generateTsDocs from 'https://cdn.skypack.dev/generate-ts-docs';
</script>

README

generate-ts-docs npm Version build

Utilities to parse type information and JSDoc annotations from TypeScript source files, and render Markdown documentation

Usage

First:

$ npm install --save-dev generate-ts-docs

Suppose that we have the following (highly contrived) toy example.ts source file:

/**
 * Adds two numbers.
 *
 * @param x  First number to add.
 * @param y  Second number to add.
 * @return The sum of `x` and `y`.
 */
export function add(x: number, y: number): number {
  return x + y
}

…and the following generate-ts-docs.ts script:

import {
  parseExportedFunctionsAsync,
  renderFunctionDataToMarkdown
} from 'generate-ts-docs'

async function main(): Promise<void> {
  const functionsData = await parseExportedFunctionsAsync(['./example.ts'])
  for (const functionData of functionsData) {
    console.log(renderFunctionDataToMarkdown(functionData))
  }
}
main()

parseExportedFunctionsAsync receives an array of globs of TypeScript source files, and parses the functions in these files that have the export keyword. It returns an array of FunctionData objects with the following shape:

[
  {
    description: 'Adds two numbers.',
    jsDocTags: null,
    name: 'add',
    parameters: [
      {
        description: 'First number to add.',
        name: 'x',
        optional: false,
        type: 'number'
      },
      {
        description: 'Second number to add.',
        name: 'y',
        optional: false,
        type: 'number'
      }
    ],
    returnType: { description: 'The sum of `x` and `y`.', type: 'number' },
    typeParameters: []
  }
]

renderFunctionDataToMarkdown renders the given array of FunctionData objects to a Markdown string.

Now, let’s run the generate-ts-docs.ts script, piping its output to a file:

$ npm install --dev ts-node
$ node --loader ts-node/esm generate-ts-docs.ts > README.md

The output README.md will be as follows:

# add(x, y)

Adds two numbers.

***Parameters***

- **`x`** (`number`) – First number to add.
- **`y`** (`number`) – Second number to add.

***Return type***

The sum of `x` and `y`.

```
number
```

API

type FunctionData = {
  description: null | string
  name: string
  parameters: Array<ParameterData>
  returnType: null | ReturnTypeData
  jsDocTags: null | JsDocTagsData
  typeParameters: Array<TypeParameterData>
}

type JsDocTagsData = Record<string, null | string>

type ObjectData = {
  keys: Array<ParameterData>
  type: 'object'
}

type ParameterData = {
  description: null | string
  name: string
  optional: boolean
  type: string | ObjectData
}

type ReturnTypeData = {
  description: null | string
  type: string
}

type TypeParameterData = {
  name: string
  defaultType: null | string
  type: null | string
}

Functions data

parseExportedFunctionsAsync(globs [, options])

Parses the exported functions defined in the given globs of TypeScript files.

  • Functions with the @ignore JSDoc tag will be skipped.
  • Functions will be sorted in ascending order of their @weight JSDoc tag. A function with the @weight tag will be ranked before a function without the @weight tag.

Parameters

  • globs (Array<string>) – One or more globs of TypeScript files.
  • options (object) – Optional.
    • tsconfigFilePath (string) – Path to a TypeScript configuration file. Defaults to ./tsconfig.json.

Return type

Promise<Array<FunctionData>>

createCategories(functionsData)

Groups each object in functionsData by the value of each function’s tags.category key.

Parameters

  • functionsData (Array<FunctionData>)

Return type

Array<{
  name: string;
  functionsData: Array<FunctionData>;
}>

Markdown

renderCategoryToMarkdown(category [, options])

Parameters

  • category (object)
    • name (string)
    • functionsData (Array<FunctionData>)
  • options (object) – Optional.
    • headerLevel (number) – Header level to be used for rendering the category name. Defaults to 1 (ie. #).

Return type

string

renderFunctionDataToMarkdown(functionData [, options])

Parameters

  • functionData (FunctionData)
  • options (object) – Optional.
    • headerLevel (number) – Header level to be used for rendering the function name. Defaults to 1 (ie. #).

Return type

string

Markdown table of contents

renderCategoriesToMarkdownToc(categories)

Generate a Markdown table of contents for the given categories.

Parameters

  • categories (Array<{ name: string; functionsData: Array<FunctionData>; }>)

Return type

string

renderFunctionsDataToMarkdownToc(functionsData)

Generate a Markdown table of contents for the given functionsData.

Parameters

  • functionsData (Array<FunctionData>)

Return type

string

Installation

$ npm install --save-dev generate-ts-docs

Implementation details

The parseExportedFunctionsAsync function works via the following two-step process:

  1. Generate type declarations for the given TypeScript source files.
  2. Traverse and extract relevant information from the AST of the generated type declarations.

(The TypeScript AST Viewer tool is useful for figuring out how nodes should be parsed.)

See also

License

MIT