Adding New Components to Next SEO

July 25, 2025 · View on GitHub

This guide walks through the process of adding new JSON-LD structured data components to next-seo. We'll use the ArticleJsonLd component as a reference implementation.

Research Phase
Type Definitions
Component Implementation
Export Configuration
Unit Tests
Documentation
Example Pages
E2E Tests
Final Verification

1. Research Phase

Before implementing, thoroughly research the structured data specification:

Visit Google's Documentation
- Go to Google's Structured Data Gallery
- Find the specific type you're implementing (e.g., Article, Product, Recipe)
- Note all required and recommended properties
Analyze Schema Types
- Identify all subtypes (e.g., Article has NewsArticle, BlogPosting, Blog)
- Note property variations between types
- Check for special formatting requirements (dates, images, etc.)
Review Existing Implementation
- If updating from an older version, fetch the previous implementation
- Identify any missing features or properties
- Ensure backward compatibility where possible

2. Type Definitions

Create comprehensive TypeScript types in src/types/[component].types.ts:

// src/types/article.types.ts
import type { ImageObject, Person, Organization, Author } from "./common.types";

// Note: Common types like ImageObject, Person, Organization, and Author
// are now defined in common.types.ts to avoid duplication

// Base interface with common properties
export interface ArticleBase {
  headline: string;
  url?: string;
  author?: Author | Author[];
  datePublished?: string;
  dateModified?: string;
  image?: string | ImageObject | (string | ImageObject)[];
  publisher?: Organization;
  description?: string;
  isAccessibleForFree?: boolean;
  mainEntityOfPage?:
    | string
    | {
        "@type": "WebPage";
        "@id": string;
      };
}

// Specific schema type interfaces
export interface Article extends ArticleBase {
  "@type": "Article";
}

export interface NewsArticle extends ArticleBase {
  "@type": "NewsArticle";
}

export interface BlogPosting extends ArticleBase {
  "@type": "BlogPosting";
}

// Component props type
export type ArticleJsonLdProps = (
  | Omit<Article, "@type">
  | Omit<NewsArticle, "@type">
  | Omit<BlogPosting, "@type">
) & {
  type?: "Article" | "NewsArticle" | "BlogPosting";
  scriptId?: string;
  scriptKey?: string;
};

Key Patterns:

Use union types for flexible inputs (e.g., string | Person | Organization)
Support both single items and arrays where appropriate
Extend common interfaces to reduce duplication
Make all properties optional except truly required ones
Include component-specific props like scriptId and scriptKey
Important: Reuse types from common.types.ts for shared definitions like ImageObject, Person, Organization, and Author

The @type Optional Pattern

A core design principle of next-seo is that developers should not need to specify @type properties manually. This provides better developer experience while maintaining full Schema.org compliance.

How It Works:

Type Definitions: Use Omit<Type, "@type"> to create props that don't require @type:

export type ArticleJsonLdProps = (
  | Omit<Article, "@type">
  | Omit<NewsArticle, "@type">
  | Omit<BlogPosting, "@type">
) & {
  type?: "Article" | "NewsArticle" | "BlogPosting";
  // ... other props
};

Process Functions: Automatically add the correct @type based on input:

// Developers can pass a simple string
author="John Doe"

// Process function converts it to a proper Person object
processAuthor("John Doe") // → { "@type": "Person", name: "John Doe" }

// Or pass an object without @type
author={{ name: "John Doe", url: "https://example.com" }}

// Process function adds @type intelligently
processAuthor({...}) // → { "@type": "Person", name: "John Doe", url: "..." }

Intelligent Type Detection: Process functions use property analysis to determine types:
- Objects with logo, address, or contactPoint → Organization
- Objects with familyName or givenName → Person
- Default fallbacks ensure valid Schema.org output

Benefits:

Less Boilerplate: Developers don't need to remember Schema.org type names
Flexible Input: Accept strings, objects with/without @type
Type Safety: Full TypeScript support throughout
Forward Compatible: Can still accept objects with @type if provided

3. Component Implementation

Create the component in src/components/[Component]JsonLd.tsx:

// src/components/ArticleJsonLd.tsx
import { JsonLdScript } from "~/core/JsonLdScript";
import type { ArticleJsonLdProps } from "~/types/article.types";
import { processAuthor, processImage } from "~/utils/processors";

// Note: Common processing functions like processAuthor and processImage
// are now available in ~/utils/processors.ts to avoid duplication

export default function ArticleJsonLd({
  type = "Article",
  scriptId,
  scriptKey,
  headline,
  url,
  author,
  datePublished,
  dateModified,
  image,
  publisher,
  description,
  isAccessibleForFree,
  mainEntityOfPage,
}: ArticleJsonLdProps) {
  const data = {
    "@context": "https://schema.org",
    "@type": type,
    headline,
    ...(url && { url }),
    ...(author && {
      author: Array.isArray(author)
        ? author.map(processAuthor)
        : processAuthor(author),
    }),
    ...(datePublished && { datePublished }),
    ...(dateModified && { dateModified }),
    // Apply defaults where appropriate
    ...(!dateModified && datePublished && { dateModified: datePublished }),
    ...(image && {
      image: Array.isArray(image) ? image.map(processImage) : processImage(image),
    }),
    ...(publisher && { publisher }),
    ...(description && { description }),
    ...(isAccessibleForFree !== undefined && { isAccessibleForFree }),
    ...(mainEntityOfPage && { mainEntityOfPage }),
  };

  return (
    <JsonLdScript
      data={data}
      id={scriptId}
      scriptKey={scriptKey || `article-jsonld-${type}`}
    />
  );
}

export type { ArticleJsonLdProps };

Implementation Guidelines:

Use the existing JsonLdScript component for rendering (now with TypeScript generics support)
Process flexible inputs to proper schema format using shared utilities from ~/utils/processors
Use object spread with conditional inclusion for optional properties
Handle arrays appropriately with .map()
Apply sensible defaults (e.g., dateModified defaults to datePublished)
Ensure boolean values are explicitly checked with !== undefined
Always use process functions for properties that accept flexible types (strings, objects with/without @type)
Never require developers to specify @type - the component should set the main @type from the type prop, and process functions should handle nested objects

4. Export Configuration

Update src/index.ts to export your component:

export { JsonLdScript } from "./core/JsonLdScript";
export {
  default as ArticleJsonLd,
  type ArticleJsonLdProps,
} from "./components/ArticleJsonLd";
// Add your new component here
export const version = "7.0.0-alpha.0";

5. Unit Tests

Create comprehensive tests in src/components/[Component]JsonLd.test.tsx:

import { render } from "@testing-library/react";
import { describe, it, expect } from "vitest";
import ArticleJsonLd from "./ArticleJsonLd";

describe("ArticleJsonLd", () => {
  it("renders basic Article with minimal props", () => {
    const { container } = render(
      <ArticleJsonLd
        headline="Test Article"
        datePublished="2024-01-01T00:00:00.000Z"
      />
    );

    const script = container.querySelector('script[type="application/ld+json"]');
    expect(script).toBeTruthy();

    const jsonData = JSON.parse(script!.textContent!);
    expect(jsonData).toEqual({
      "@context": "https://schema.org",
      "@type": "Article",
      headline: "Test Article",
      datePublished: "2024-01-01T00:00:00.000Z",
      dateModified: "2024-01-01T00:00:00.000Z", // defaults to datePublished
    });
  });

  // Test each schema type
  it("renders NewsArticle type when specified", () => {
    // ... test implementation
  });

  // Test flexible inputs
  it("handles string author", () => {
    // ... converts string to Person object
  });

  it("handles multiple authors", () => {
    // ... test array handling
  });

  // Test all properties
  it("handles all optional properties", () => {
    // ... comprehensive test with all props
  });

  // Test edge cases
  it("handles isAccessibleForFree as false", () => {
    // ... ensure boolean false is included
  });
});

Testing Checklist:

✅ Basic rendering with minimal props
✅ All schema type variations
✅ String to object conversions
✅ Array handling for authors and images
✅ All optional properties
✅ Default value application
✅ Boolean value handling
✅ Custom scriptId and scriptKey

6. Documentation

Add comprehensive documentation to README.md:

### ArticleJsonLd

The `ArticleJsonLd` component helps you add structured data for articles, blog posts, and news articles to improve their appearance in search results.

#### Basic Usage

```tsx
import { ArticleJsonLd } from "next-seo";

<ArticleJsonLd
  headline="My Amazing Article"
  datePublished="2024-01-01T08:00:00+08:00"
  author="John Doe"
  image="https://example.com/article-image.jpg"
  description="This article explains amazing things"
/>;
```

Props

Property	Type	Description
`type`	`"Article" \| "NewsArticle" \| "BlogPosting"`	The type of article. Defaults to "Article"
`headline`	`string`	Required. The headline of the article
...	...	...

Best Practices

Always include images: Google recommends multiple aspect ratios
Use ISO 8601 dates: Include timezone information
Multiple authors: List all authors when applicable


## 7. Example Pages

Create example pages in `examples/app-router-showcase/app/[component]/page.tsx`:

```tsx
import { ArticleJsonLd } from "next-seo";

export default function ArticlePage() {
  return (
    <div className="container mx-auto p-8">
      <ArticleJsonLd
        headline="Understanding Next.js App Router"
        url="https://example.com/articles/nextjs-app-router"
        datePublished="2024-01-01T08:00:00+00:00"
        author="Sarah Johnson"
        image="https://example.com/images/nextjs-article.jpg"
        description="A comprehensive guide to Next.js App Router"
      />

      <article className="prose lg:prose-xl">
        <h1>Understanding Next.js App Router</h1>
        {/* Article content */}
      </article>
    </div>
  );
}

Create examples for:

Basic usage (minimal props)
Advanced usage (all features)
Each schema type variation

8. E2E Tests

Create Playwright tests in tests/e2e/[component]JsonLd.e2e.spec.ts:

Important E2E Testing Guidelines

ALL E2E tests must use real example pages! E2E tests should test the actual component behavior through real pages in the example app. Never mock or inject content in E2E tests.

❌ DO NOT use page.route() to inject mock HTML:

// BAD - This is not a real E2E test!
await page.route("/test-page", async (route) => {
  await route.fulfill({
    body: `<html>...</html>`,
  });
});

✅ DO create real example pages and test them:

// GOOD - Test real pages with actual components
await page.goto("/article");

Creating E2E Tests

For every E2E test scenario, you must:

Create a real example page in examples/app-router-showcase/app/
Write the E2E test to navigate to that page
Test the actual rendered output

import { test, expect } from "@playwright/test";

test.describe("ArticleJsonLd", () => {
  test("renders basic Article structured data", async ({ page }) => {
    // Navigate to the real example page
    await page.goto("/article");

    const jsonLdScript = await page
      .locator('script[type="application/ld+json"]')
      .textContent();
    expect(jsonLdScript).toBeTruthy();

    const jsonData = JSON.parse(jsonLdScript!);

    // Verify all properties
    expect(jsonData["@context"]).toBe("https://schema.org");
    expect(jsonData["@type"]).toBe("Article");
    expect(jsonData.headline).toBe("Understanding Next.js App Router");
    // ... test all properties
  });

  test("properly escapes HTML entities in content", async ({ page }) => {
    // Navigate to a real example page with special characters
    await page.goto("/article-special-chars");

    const jsonLdScript = await page
      .locator('script[type="application/ld+json"]')
      .textContent();

    // Verify JSON is valid and content is properly escaped
    const jsonData = JSON.parse(jsonLdScript!);
    expect(jsonData.headline).toContain("Special & Characters");

    // Check that dangerous content is escaped in the raw JSON
    expect(jsonLdScript).toContain("\\u003C/script>");
  });
});

When to Create Additional Example Pages

Create new example pages for:

Basic usage with minimal props
Advanced usage with all features
Each schema type variation (e.g., Article, NewsArticle, BlogPosting)
Special characters and HTML entities
Edge cases with unusual data
Different data combinations

Example structure:

examples/app-router-showcase/app/
├── article/                    # Basic article example
├── article-advanced/           # All features
├── news-article/              # NewsArticle type
├── blog-posting/              # BlogPosting type
└── article-special-chars/     # Special characters test

You should also add a valid JSON test in tests/e2e/jsonValidation.e2e.spec.ts

Security and Escaping Tests

DO NOT add escape/security tests to individual component E2E tests!

Security testing for escaping dangerous sequences (like </script>, HTML comments, etc.) is handled centrally in tests/e2e/security.e2e.spec.ts. This test file comprehensively covers:

Script tag injection prevention
HTML comment escaping
Edge cases with mixed dangerous patterns
Safe rendering in Next.js-like environments

Individual component E2E tests should focus on:

Component-specific functionality
Correct data structure output
Schema type variations
Required and optional properties

The escaping functionality is a core library feature handled by the stringify utility, not something each component needs to test individually.

9. Final Verification

Before completing, run all quality checks:

# 1. Run unit tests
pnpm test:unit

# 2. Type checking
pnpm typecheck

# 3. Linting
pnpm lint

# 4. Build the package
pnpm build

Developer will run e2e manually as they can take a long time.

Common Patterns and Best Practices

Shared Utilities

The library now provides shared utilities to avoid code duplication:

Common Types (~/types/common.types.ts):
- ImageObject, Person, Organization, Author
- Base interfaces like Thing
Processing Functions (~/utils/processors.ts):
- processAuthor(author: Author): Person | Organization
- processImage(image: string | ImageObject): string | ImageObject

Flexible Input Processing

Use the shared processing functions from ~/utils/processors:

import { processAuthor, processImage } from "~/utils/processors";

// These functions handle string-to-object conversions automatically
// and add the appropriate @type without developers needing to specify it

Important: Always create or use existing process functions for properties that can accept multiple formats. This maintains the pattern of not requiring developers to specify @type and ensures consistent behavior across all components.

Conditional Property Inclusion

Use object spread with conditional checks:

const data = {
  "@context": "https://schema.org",
  "@type": type,
  headline,
  ...(url && { url }), // Only include if truthy
  ...(isAccessibleForFree !== undefined && { isAccessibleForFree }), // Include false values
};

Default Values

Apply sensible defaults where appropriate:

// If dateModified is not provided but datePublished is, use datePublished
...(!dateModified && datePublished && { dateModified: datePublished }),

Array Handling

Support both single items and arrays:

...(author && {
  author: Array.isArray(author)
    ? author.map(processAuthor)
    : processAuthor(author),
}),

Troubleshooting

Common Issues

ESLint errors about unused React import
- Remove import React from 'react' - it's not needed with modern JSX transform
Test failures with dateModified
- Remember that dateModified defaults to datePublished when not provided
Boolean properties not appearing
- Use !== undefined check instead of truthy check for booleans
Type errors with union types
- Ensure proper type guards in processing functions