REST API

REST API Standards

Standards for fetching REST API data using TanStack Query with OpenTelemetry tracing.

When to Use

Use REST API only when data is NOT available in GraphQL schema. Always verify GraphQL schema first.

Use these standards when:

• Creating new API endpoints not in GraphQL
• Fetching from legacy REST endpoints
• Integrating with external APIs
• Implementing infinite scroll pagination
• Setting up cache invalidation

Why: GraphQL provides better type safety and reduces over-fetching, but REST is needed for legacy endpoints and external APIs.

Core Principles

useFetch Hook with OpenTelemetry

Always use useFetch hook instead of native fetch(). It provides automatic OpenTelemetry tracing, request ID generation, and authentication middleware.

✅ Good:

import { useFetch } from '@libs/networking-client';

export const useGetProducts = (params: GetProductsRequest) => {
  const fetch = useFetch(); // Provides traced fetch with request IDs

  return useQuery({
    queryKey: PRODUCT_QUERY_KEYS.list(params),
    queryFn: () => getProducts(params, queryKey, fetch),
  });
};

❌ Bad:

// Never use native fetch directly
export const useGetProducts = (params: GetProductsRequest) => {
  return useQuery({
    queryKey: ['products', params],
    queryFn: async () => {
      const response = await fetch('/api/products'); // ❌ No tracing
      return response.json();
    },
  });
};

Why: useFetch automatically:

• Creates OpenTelemetry spans for distributed tracing
• Adds unique request IDs for debugging
• Handles authentication middleware
• Provides consistent error handling

Real-world usage: 200+ usages in shared-mobile-modules codebase.

Hierarchical Query Keys

Use pattern: [domain, operation, params]

✅ Good:

export const PRODUCT_QUERY_KEYS = {
  all: ['products'] as const,
  list: (params?) => ['products', 'list', params] as const,
  detail: (id: string) => ['products', 'detail', id] as const,
} as const;

// Usage
queryKey: PRODUCT_QUERY_KEYS.list({ category: 'electronics' })
// Result: ['products', 'list', { category: 'electronics' }]

❌ Bad:

// Flat keys - hard to invalidate
queryKey: ['products']
queryKey: ['productList']
queryKey: ['product', id, 'random'] // Inconsistent structure

Why: Hierarchical keys enable precise cache invalidation:

• queryClient.invalidateQueries({ queryKey: PRODUCT_QUERY_KEYS.all }) - Invalidates all product queries
• queryClient.invalidateQueries({ queryKey: PRODUCT_QUERY_KEYS.list() }) - Invalidates only list queries

File Organization

Organize by domain:

src/data-access/query/{domain}/
├── constants.ts      # Query keys & endpoints
├── service.ts        # Fetch functions
├── schema.ts         # TypeScript types
├── hooks.ts          # TanStack Query hooks
└── index.ts          # Exports

Why: Domain-based organization keeps API logic together and makes it easy to find related code.

Type Definitions

Define explicit interfaces for all requests and responses.

// schema.ts

/**
 * Product entity from API
 */
export interface Product {
  /** Unique identifier */
  id: string;
  /** Product name */
  name: string;
  /** Price in cents */
  price: number;
  /** Product category */
  category: 'electronics' | 'clothing' | 'books';
  /** Availability status */
  status: 'active' | 'inactive';
  /** Creation timestamp (ISO 8601) */
  createdAt: string;
}

/**
 * Request parameters for fetching products
 */
export interface GetProductsRequest {
  cursor?: string;
  limit?: number;
  category?: Product['category'];
  search?: string;
}

/**
 * Response for product list with pagination
 */
export interface GetProductsResponse {
  data: Product[];
  pagination: {
    nextCursor?: string;
    hasMore: boolean;
  };
}

Why: Explicit types provide autocomplete, catch errors at compile time, and serve as documentation.

Service Functions

Implement fetch functions using the traced fetch parameter from useFetch.

// service.ts
import type { DataAccessGet } from './schema';

export const getProducts: DataAccessGet<GetProductsResponse, GetProductsRequest> = async (
  params,
  queryKey,
  fetch
) => {
  // Build query params
  const searchParams = new URLSearchParams();
  if (params.cursor) searchParams.append('cursor', params.cursor);
  if (params.limit) searchParams.append('limit', params.limit.toString());
  if (params.category) searchParams.append('category', params.category);

  // fetch parameter comes from useFetch() - includes tracing
  const response = await fetch(`${PRODUCT_ENDPOINTS.LIST}?${searchParams}`, queryKey, {
    method: 'GET',
    headers: {
      'Content-Type': 'application/json',
    },
  });

  if (!response.ok) {
    throw new Error(`Failed to fetch products: ${response.status}`);
  }

  return await response.json();
};

Why: Centralized service functions make API calls consistent and easier to test.

Custom Hooks

Create domain-specific hooks using TanStack Query.

// hooks.ts
import { useQuery } from '@tanstack/react-query';
import { useFetch, useLocalizeParams } from '@/hooks/data-access';

/**
 * Hook for fetching products list
 */
export const useGetProducts = (params: GetProductsRequest, options?) => {
  const fetch = useFetch();
  const localizeParams = useLocalizeParams();
  const requestParams = { ...params, ...localizeParams };

  return useQuery({
    queryKey: PRODUCT_QUERY_KEYS.list(requestParams),
    queryFn: () => getProducts(requestParams, queryKey, fetch),
    staleTime: 5 * 60 * 1000, // 5 minutes
    ...options,
  });
};

/**
 * Hook for fetching single product
 */
export const useGetProduct = (productId: string, options?) => {
  const fetch = useFetch();
  const localizeParams = useLocalizeParams();

  return useQuery({
    queryKey: PRODUCT_QUERY_KEYS.detail(productId),
    queryFn: () => getProduct({ id: productId, ...localizeParams }, queryKey, fetch),
    enabled: !!productId,
    staleTime: 10 * 60 * 1000, // 10 minutes
    ...options,
  });
};

Why: Custom hooks abstract TanStack Query details, include localization automatically, and provide consistent patterns.

Infinite Queries for Pagination

Use useInfiniteQuery for paginated lists with infinite scroll.

// hooks.ts
import { useInfiniteQuery } from '@tanstack/react-query';

export const useGetProductsInfinite = (params: Omit<GetProductsRequest, 'cursor'>, options?) => {
  const fetch = useFetch();
  const localizeParams = useLocalizeParams();
  const requestParams = { ...params, ...localizeParams };

  return useInfiniteQuery({
    queryKey: [...PRODUCT_QUERY_KEYS.list(requestParams), 'infinite'],
    queryFn: ({ pageParam }) => {
      return getProducts({ ...requestParams, cursor: pageParam }, queryKey, fetch);
    },
    initialPageParam: undefined,
    getNextPageParam: (lastPage) => lastPage.pagination.nextCursor,
    staleTime: 5 * 60 * 1000,
    ...options,
  });
};

Why: useInfiniteQuery manages pagination state automatically and enables smooth infinite scroll UX.

Component Integration

Standard Query Usage

// ProductList.tsx
import { useGetProducts } from '@/data-access/query/products';

export const ProductList = () => {
  const { data, isLoading, error, refetch } = useGetProducts({
    limit: 20,
    sort: 'createdAt',
  });

  if (isLoading && !data) return <LoadingSpinner />;
  if (error && !data) return <ErrorMessage error={error} />;
  if (!data?.data.length) return <EmptyState />;

  return (
    <FlatList
      data={data.data}
      keyExtractor={(item) => item.id}
      renderItem={({ item }) => <ProductItem product={item} />}
    />
  );
};

Infinite Scroll Usage

// ProductInfiniteList.tsx
import { useGetProductsInfinite } from '@/data-access/query/products';

export const ProductInfiniteList = ({ category }) => {
  const {
    data: infiniteData,
    fetchNextPage,
    hasNextPage,
    isFetchingNextPage,
    isLoading,
    error,
  } = useGetProductsInfinite({ category, limit: 20 });

  // Flatten pages into single array
  const products = infiniteData?.pages.flatMap((page) => page.data) || [];

  if (isLoading && !infiniteData) return <LoadingSpinner />;
  if (error && !infiniteData) return <ErrorMessage error={error} />;
  if (!products.length) return <EmptyState />;

  return (
    <FlatList
      data={products}
      keyExtractor={(item) => item.id}
      renderItem={({ item }) => <ProductItem product={item} />}
      => {
        if (hasNextPage && !isFetchingNextPage) {
          fetchNextPage();
        }
      }}
     
      ListFooterComponent={isFetchingNextPage ? <LoadingSpinner /> : null}
    />
  );
};

Why: Flattening pages simplifies FlatList integration. onEndReached triggers pagination automatically.

Mutations with Cache Invalidation

// hooks.ts
export const useCreateProduct = (options?) => {
  const queryClient = useQueryClient();
  const fetch = useFetch();

  return useMutation({
    mutationFn: (params: CreateProductRequest) => {
      return createProduct(params, queryKey, fetch);
    },
    onSuccess: () => {
      // Invalidate list to refetch
      queryClient.invalidateQueries({ queryKey: PRODUCT_QUERY_KEYS.all });
    },
    ...options,
  });
};

Why: Proper cache invalidation keeps UI in sync with server state after mutations.

Performance Optimization

Configure appropriate stale times based on data volatility.

// List data changes frequently
useQuery({
  queryKey: PRODUCT_QUERY_KEYS.list(params),
  queryFn: () => getProducts(params, queryKey, fetch),
  staleTime: 5 * 60 * 1000, // 5 minutes
  gcTime: 10 * 60 * 1000, // 10 minutes
});

// Detail data is more stable
useQuery({
  queryKey: PRODUCT_QUERY_KEYS.detail(id),
  queryFn: () => getProduct({ id }, queryKey, fetch),
  staleTime: 10 * 60 * 1000, // 10 minutes
  gcTime: 15 * 60 * 1000, // 15 minutes
});

Why: Proper cache timing reduces unnecessary API calls while keeping data fresh.

Testing

Mock service functions and test hooks in isolation.

import { renderHook, waitFor } from '@testing-library/react-native';
import { QueryClient, QueryClientProvider } from '@tanstack/react-query';
import { useGetProducts } from './hooks';
import * as service from './service';

jest.mock('./service');
jest.mock('@/hooks/data-access', () => ({
  useFetch: () => jest.fn(),
  useLocalizeParams: () => ({ locale: 'en' }),
}));

const mockGetProducts = service.getProducts as jest.MockedFunction<typeof service.getProducts>;

test('returns products data', async () => {
  const mockResponse = {
    data: [{ id: '1', name: 'Test Product', price: 2999 }],
    pagination: { hasMore: false },
  };

  mockGetProducts.mockResolvedValue(mockResponse);

  const queryClient = new QueryClient({ defaultOptions: { queries: { retry: false } } });
  const wrapper = ({ children }) => (
    <QueryClientProvider client={queryClient}>{children}</QueryClientProvider>
  );

  const { result } = renderHook(() => useGetProducts({ limit: 10 }), { wrapper });

  await waitFor(() => expect(result.current.isSuccess).toBe(true));
  expect(result.current.data).toEqual(mockResponse);
});

Why: Mocking enables testing without actual API calls and validates query structure.

Common Mistakes

❌ Don't Use REST When GraphQL Exists

// GraphQL query exists - use that instead
const products = useQuery({
  queryKey: ['products'],
  queryFn: fetchProductsRest, // Wrong
});

❌ Don't Call Fetch Directly in Components

useEffect(() => {
  fetch('/api/products').then(setData); // Wrong - no caching, no tracing
}, []);

❌ Don't Use Inconsistent Query Keys

// Inconsistent structure
const queryKey = ['products', id, 'random'];

✅ Do Use Custom Hooks with Standard Patterns

const { data, isLoading, error } = useGetProducts(params);

Quick Reference

Query Keys:

const KEYS = {
  all: ['domain'] as const,
  list: (params?) => ['domain', 'list', params] as const,
  detail: (id) => ['domain', 'detail', id] as const,
} as const;

Standard Hook:

export const useGet = (params, options?) => {
  const fetch = useFetch();
  return useQuery({
    queryKey: KEYS.list(params),
    queryFn: () => service(params, queryKey, fetch),
    ...options,
  });
};

Infinite Query:

useInfiniteQuery({
  queryKey: [...KEYS.list(params), 'infinite'],
  queryFn: ({ pageParam }) => service({ ...params, cursor: pageParam }, queryKey, fetch),
  initialPageParam: undefined,
  getNextPageParam: (lastPage) => lastPage.pagination.nextCursor,
});

Additional Resources

For detailed examples and patterns, see:

• references/examples.md - Real API hook implementations
• references/patterns.md - REST API patterns and anti-patterns