Logo
  1. Docs
  2. Guides

Section Data Fetching

Published on Nov 10, 2023, updated 4 days ago

Section Data Fetching

Table of Contents

Introduction

Weaverse's section components can fetch their own data, which provides several key advantages:

  • Modularity: Each component manages its own data dependencies
  • Performance: Only fetch what's needed when it's needed
  • Maintainability: Data fetching logic lives with the component that uses it
  • Reusability: Components can be used in multiple contexts with different data

This guide covers everything you need to know about fetching data in Weaverse section components.

Core Concepts

The Component Loader Pattern

Weaverse section components use the loader pattern for server-side data fetching:

// app/sections/featured-collection/index.tsximport type { ComponentLoaderArgs } from '@weaverse/hydrogen'
// Define the expected input data typetype FeaturedCollectionData = {  collection: { handle: string }}
// Component implementation...
// Define the loader function to fetch dataexport let loader = async (args: ComponentLoaderArgs<FeaturedCollectionData>) => {  const { weaverse, data } = args  const { storefront } = weaverse    // Access component settings through the data parameter  const { collection } = data    // Fetch and return the data  return await storefront.query(COLLECTION_QUERY, {    variables: {      handle: collection.handle    }  })}

The loader function:

  1. Receives arguments via ComponentLoaderArgs
  2. Accesses the Weaverse client and component data
  3. Fetches necessary data from APIs
  4. Returns data that will be automatically passed to the component as props.loaderData

Type Safety with TypeScript

Using TypeScript with your component loaders provides several benefits:

import type { ComponentLoaderArgs } from '@weaverse/hydrogen'import type { CollectionQuery } from 'storefrontapi.generated'
// Define input data shapetype FeaturedCollectionData = {  collection: { handle: string }  productsToShow: number}
// Define what the loader returns (will be available as props.loaderData)type LoaderReturnType = {  collection: CollectionQuery['collection']}
export let loader = async (  args: ComponentLoaderArgs<FeaturedCollectionData>): Promise<LoaderReturnType | null> => {  // Implementation...}

By specifying the input and output types, you get:

  • Auto-completion in your IDE
  • Type checking during development
  • Better documentation for component usage
  • Clearer contract between components and their data

Data Sources

Shopify Storefront API

The most common data source for Weaverse components is Shopify's Storefront API:

export let loader = async ({ weaverse, data }: ComponentLoaderArgs<ProductData>) => {  const { storefront } = weaverse  const { product } = data    if (!product?.handle) return null    return await storefront.query(PRODUCT_QUERY, {    variables: {      handle: product.handle,      language: storefront.i18n.language,      country: storefront.i18n.country    }  })}
// GraphQL query defined elsewhereconst PRODUCT_QUERY = `#graphql  query Product($handle: String!, $language: LanguageCode, $country: CountryCode)  @inContext(language: $language, country: $country) {    product(handle: $handle) {      id      title      description      # Other fields...    }  }` as const

External APIs

Weaverse components can also fetch data from any external API using the fetchWithCache utility:

export let loader = async ({ weaverse, data }: ComponentLoaderArgs<WeatherWidgetData>) => {  const { fetchWithCache, env } = weaverse  const { location = 'New York' } = data    try {    return await fetchWithCache(`https://api.weather.example/forecast`, {      method: 'POST',      headers: {        'Content-Type': 'application/json',        'API-Key': env.WEATHER_API_KEY      },      body: JSON.stringify({ location }),      // Use built-in caching      strategy: weaverse.storefront.CacheShort()    })  } catch (error) {    console.error('Weather API error:', error)    return { error: true, forecast: [] }  }}

The fetchWithCache function:

  • Works like the standard fetch API
  • Adds Hydrogen's caching capabilities
  • Makes external API calls more efficient
  • Provides a consistent interface for all data fetching

Component Data and Settings

The data argument in ComponentLoaderArgs contains all the component's settings configured by merchants in the Weaverse editor:

type MapComponentData = {  latitude: number  longitude: number  zoom: number  showTraffic: boolean  mapStyle: 'standard' | 'satellite' | 'terrain'}
export let loader = async ({ weaverse, data }: ComponentLoaderArgs<MapComponentData>) => {  const { fetchWithCache, env } = weaverse  const { latitude, longitude, zoom, showTraffic, mapStyle } = data    // Use component settings to customize the API request  return await fetchWithCache(    `https://api.maps.example/staticmap?lat=${latitude}&lng=${longitude}&zoom=${zoom}&traffic=${showTraffic}&style=${mapStyle}&key=${env.MAPS_API_KEY}`  )}

Implementation Patterns

Basic Data Fetching

The simplest pattern is direct data fetching based on component settings:

export let loader = async ({ weaverse, data }: ComponentLoaderArgs<ProductData>) => {  const { storefront } = weaverse  const { productHandle } = data    return await storefront.query(PRODUCT_QUERY, {    variables: { handle: productHandle }  })}

Conditional Fetching

Often you'll need to conditionally fetch data based on component settings:

export let loader = async ({ weaverse, data }: ComponentLoaderArgs<TestimonialData>) => {  const { storefront, fetchWithCache, env } = weaverse  const { source, productHandle, collectionHandle } = data    // Choose data source based on component settings  switch (source) {    case 'product-reviews':      return await fetchWithCache(`https://api.reviews.example/product/${productHandle}`, {        headers: { 'API-Key': env.REVIEWS_API_KEY }      })        case 'collection-reviews':      return await fetchWithCache(`https://api.reviews.example/collection/${collectionHandle}`, {        headers: { 'API-Key': env.REVIEWS_API_KEY }      })        case 'store-testimonials':      return await storefront.query(TESTIMONIALS_QUERY)          default:      return { testimonials: [] }  }}

Parallel Data Fetching

For optimal performance, fetch multiple data sources in parallel:

export let loader = async ({ weaverse, data }: ComponentLoaderArgs<ProductDetailData>) => {  const { storefront, fetchWithCache, env } = weaverse  const { productHandle } = data    // Fetch from multiple sources simultaneously  const [productData, reviewsData, inventoryData] = await Promise.all([    // Product data from Shopify    storefront.query(PRODUCT_QUERY, {      variables: { handle: productHandle }    }),        // Reviews from third-party API    fetchWithCache(`https://api.reviews.example/products/${productHandle}`, {      headers: { 'Authorization': `Bearer ${env.REVIEWS_API_KEY}` }    }),        // Inventory data from ERP system    fetchWithCache(`https://api.inventory.example/stock-levels`, {      method: 'POST',      headers: { 'API-Key': env.INVENTORY_API_KEY },      body: JSON.stringify({ sku: productHandle })    })  ])    // Combine the results  return {    product: productData.product,    reviews: reviewsData.reviews || [],    inventory: inventoryData.stockLevels || {}  }}

Error Handling

Robust error handling ensures your components degrade gracefully when APIs fail:

export let loader = async ({ weaverse, data }: ComponentLoaderArgs<NewsData>) => {  const { fetchWithCache } = weaverse  const { source, category, count = 3 } = data    try {    const response = await fetchWithCache(      `https://api.news.example/${source}?category=${category}&count=${count}`    )        // Validate the response    if (!response || !Array.isArray(response.articles)) {      console.warn('News API returned invalid data format')      return { articles: [], error: 'invalid_format' }    }        return {       articles: response.articles,      error: null    }  } catch (error) {    console.error('News API error:', error)        // Return structured error data for the component to handle    return {       articles: [],      error: error instanceof Error ? error.message : 'unknown_error'    }  }}

Dependent Queries

Sometimes you need to fetch data sequentially, where one request depends on the results of another:

export let loader = async ({ weaverse, data }: ComponentLoaderArgs<RelatedProductsData>) => {  const { storefront } = weaverse  const { productHandle } = data    // Step 1: Get the main product to find its type  const { product } = await storefront.query(PRODUCT_BASIC_QUERY, {    variables: { handle: productHandle }  })    if (!product) return { relatedProducts: [] }    // Step 2: Use the product type to find related products  const { products } = await storefront.query(RELATED_PRODUCTS_QUERY, {    variables: {       productType: product.productType,      excludeId: product.id,      first: 4    }  })    return { relatedProducts: products.nodes }}

Data Revalidation

Weaverse provides a powerful mechanism to automatically refresh component data when specific settings change. This ensures that the displayed content always reflects the current configuration.

Understanding shouldRevalidate

The shouldRevalidate property, when added to an input in your component schema, tells Weaverse to reload the component's data from its loader function whenever that specific input changes.

// In your component schemainspector: [  {    group: 'Settings',    inputs: [      {        type: 'select',        name: 'sortOrder',        label: 'Sort Products By',        defaultValue: 'best-selling',        shouldRevalidate: true, // This triggers data reload when changed        configs: {          options: [            { value: 'best-selling', label: 'Best Selling' },            { value: 'newest', label: 'Newest' },            { value: 'price-low-high', label: 'Price: Low to High' },            { value: 'price-high-low', label: 'Price: High to Low' }          ]        }      }    ]  }]

When a merchant changes the sortOrder in the editor, Weaverse will:

  1. Update the component's data with the new value
  2. Re-run the component's loader function
  3. Refresh the component with the updated data

This creates a seamless experience where changes in the editor immediately update the displayed content.

Auto-Revalidating Inputs

Some input types automatically trigger revalidation without needing the shouldRevalidate property:

  • product - When selecting a different product
  • collection - When selecting a different collection
  • blog - When selecting a different blog
  • product-list - When selecting different products
  • collection-list - When selecting different collections

These inputs deal with Shopify resources that typically require fresh data when changed.

Custom Revalidation Rules

You can create powerful data-driven components by combining shouldRevalidate with component loaders:

// app/sections/dynamic-product-grid/schema.tsexport const schema = {  // ... other schema properties  inspector: [    {      group: 'Content',      inputs: [        {          type: 'collection',          name: 'collection',          label: 'Collection',          // No need for shouldRevalidate as collection inputs auto-revalidate        },        {          type: 'select',          name: 'sortBy',          label: 'Sort By',          defaultValue: 'BEST_SELLING',          shouldRevalidate: true, // Will reload data when changed          configs: {            options: [              { value: 'BEST_SELLING', label: 'Best Selling' },              { value: 'CREATED_AT', label: 'Newest' },              { value: 'PRICE', label: 'Price: Low to High' },              { value: 'PRICE_DESC', label: 'Price: High to Low' }            ]          }        },        {          type: 'range',          name: 'productsToShow',          label: 'Number of Products',          defaultValue: 4,          shouldRevalidate: true, // Will reload data when changed          configs: {            min: 2,            max: 12,            step: 1          }        },        {          type: 'select',          name: 'viewStyle',          label: 'View Style',          defaultValue: 'grid',          shouldRevalidate: false, // UI change only, no data reload needed          configs: {            options: [              { value: 'grid', label: 'Grid' },              { value: 'slider', label: 'Slider' }            ]          }        }      ]    }  ]}
// app/sections/dynamic-product-grid/index.tsimport type { ComponentLoaderArgs } from '@weaverse/hydrogen'
type ProductGridData = {  collection: { handle: string }  sortBy: 'BEST_SELLING' | 'CREATED_AT' | 'PRICE' | 'PRICE_DESC'  productsToShow: number  viewStyle: string}
export let loader = async ({ weaverse, data }: ComponentLoaderArgs<ProductGridData>) => {  const { storefront } = weaverse  const { collection, sortBy, productsToShow } = data    if (!collection?.handle) return { products: [] }    // This query will re-run whenever collection, sortBy, or productsToShow changes  // because they have shouldRevalidate: true in the schema  return await storefront.query(COLLECTION_PRODUCTS_QUERY, {    variables: {      handle: collection.handle,      sortKey: sortBy,      first: productsToShow,      language: storefront.i18n.language,      country: storefront.i18n.country    }  })}
// Component implementation...

In this example:

  • Changing the collection, sort order, or number of products triggers a data reload
  • Changing the view style doesn't require new data, so shouldRevalidate is set to false

This optimizes performance by only reloading data when necessary, while ensuring that content stays fresh and relevant as merchants configure their components.

Caching Strategies

Weaverse components inherit Hydrogen's powerful caching system, allowing you to optimize performance based on how frequently your data changes.

Available Caching Options

StrategyCache Control HeaderBest Used For
CacheShort()public, max-age=1, stale-while-revalidate=9Frequently changing data (price, inventory)
CacheLong()public, max-age=3600, stale-while-revalidate=82800Rarely changing data (product details, images)
CacheNone()no-storePersonalized or uncacheable data
CacheCustom()Custom definedSpecial caching requirements

Custom Caching

For fine-tuned control over caching behavior:

export let loader = async ({ weaverse }: ComponentLoaderArgs<StockTickerData>) => {  const { fetchWithCache, storefront } = weaverse    // Fast-changing financial data needs custom caching  return await fetchWithCache('https://api.stocks.example/ticker', {    strategy: storefront.CacheCustom({      // Cache for 30 seconds      maxAge: 30,      // Allow stale content for up to 2 minutes while revalidating      staleWhileRevalidate: 120,      // If revalidation fails, serve stale content for up to 5 minutes      staleIfError: 300,      // Cache control mode      mode: 'public'    })  })}

Caching Best Practices

  1. Match cache duration to data volatility:

    • Product descriptions: CacheLong()
    • Prices and inventory: CacheShort()
    • User-specific content: CacheNone()
  2. Use stale-while-revalidate for a balance of freshness and performance

  3. Consider cache hierarchies when combining multiple data sources

  4. Be mindful of API rate limits when setting short cache durations

  5. Use cache debugging headers during development to verify caching behavior

Real-World Examples

E-commerce Examples

Product Recommendations Component

// app/sections/product-recommendations/index.tsximport type { ComponentLoaderArgs } from '@weaverse/hydrogen'import { RECOMMENDATIONS_QUERY } from '~/graphql/queries'
type RecommendationsData = {  product: { id: string; handle: string }  recommendationType: 'related' | 'complementary' | 'bestsellers'  maxProducts: number}
export let loader = async ({   weaverse,   data }: ComponentLoaderArgs<RecommendationsData>) => {  const { storefront } = weaverse  const { product, recommendationType, maxProducts = 4 } = data    if (!product?.id) return { recommendations: [] }    switch (recommendationType) {    case 'related':    case 'complementary':      return await storefront.query(RECOMMENDATIONS_QUERY, {        variables: {          productId: product.id,          intent: recommendationType.toUpperCase(),          count: maxProducts        }      })        case 'bestsellers':      return await storefront.query(BESTSELLERS_QUERY, {        variables: {          count: maxProducts        }      })  }}

Content Integration

Blog Feed with Categories

// app/sections/blog-feed/index.tsximport type { ComponentLoaderArgs } from '@weaverse/hydrogen'import { BLOG_ARTICLES_QUERY } from '~/graphql/queries'
type BlogFeedData = {  blog: { handle: string }  category?: string  tagsToInclude: string[]  tagsToExclude: string[]  postsCount: number  sortBy: 'newest' | 'oldest' | 'title'}
export let loader = async ({   weaverse,   data }: ComponentLoaderArgs<BlogFeedData>) => {  const { storefront } = weaverse  const {     blog,     category,     tagsToInclude = [],     tagsToExclude = [],    postsCount = 3,    sortBy = 'newest'  } = data    if (!blog?.handle) return { articles: [] }    // Get all articles for the blog  const response = await storefront.query(BLOG_ARTICLES_QUERY, {    variables: {      blogHandle: blog.handle,      first: 250 // Maximum to retrieve    }  })    // Client-side filtering and sorting (could be moved to a server query with a custom app extension)  let articles = response.blog.articles.nodes    // Filter by category if specified  if (category) {    articles = articles.filter(article =>       article.tags.includes(category)    )  }    // Filter by tags  if (tagsToInclude.length > 0) {    articles = articles.filter(article =>       tagsToInclude.some(tag => article.tags.includes(tag))    )  }    if (tagsToExclude.length > 0) {    articles = articles.filter(article =>       !tagsToExclude.some(tag => article.tags.includes(tag))    )  }    // Sort articles  switch (sortBy) {    case 'newest':      articles.sort((a, b) => new Date(b.publishedAt).getTime() - new Date(a.publishedAt).getTime())      break    case 'oldest':      articles.sort((a, b) => new Date(a.publishedAt).getTime() - new Date(b.publishedAt).getTime())      break    case 'title':      articles.sort((a, b) => a.title.localeCompare(b.title))      break  }    // Limit to requested count  articles = articles.slice(0, postsCount)    return { articles }}

Third-Party Services

Currency Converter Widget

// app/sections/currency-converter/index.tsximport type { ComponentLoaderArgs } from '@weaverse/hydrogen'
type CurrencyConverterData = {  baseCurrency: string  targetCurrencies: string[]  showChart: boolean}
type ExchangeRateResponse = {  base: string  rates: Record<string, number>  timestamp: number}
export let loader = async ({   weaverse,   data }: ComponentLoaderArgs<CurrencyConverterData>) => {  const { fetchWithCache, env, storefront } = weaverse  const {     baseCurrency = 'USD',     targetCurrencies = ['EUR', 'GBP', 'JPY', 'CAD'],    showChart = false   } = data    try {    // Fetch current exchange rates    const exchangeRates = await fetchWithCache<ExchangeRateResponse>(      `https://api.exchangerate.host/latest?base=${baseCurrency}&symbols=${targetCurrencies.join(',')}`,      {        // Exchange rates change throughout the day, but we don't need to fetch every request        strategy: storefront.CacheCustom({          maxAge: 900, // 15 minutes          staleWhileRevalidate: 3600 // 1 hour        })      }    )        // Fetch historical data for chart if needed    let historicalData = null    if (showChart) {      const today = new Date()      const lastMonth = new Date(today.setMonth(today.getMonth() - 1))      const formattedDate = lastMonth.toISOString().split('T')[0]            historicalData = await fetchWithCache(        `https://api.exchangerate.host/timeseries?start_date=${formattedDate}&end_date=${new Date().toISOString().split('T')[0]}&base=${baseCurrency}&symbols=${targetCurrencies.join(',')}`,        {          // Historical data can be cached longer          strategy: storefront.CacheLong()        }      )    }        return {      base: baseCurrency,      rates: exchangeRates.rates,      timestamp: exchangeRates.timestamp,      historicalData: historicalData?.rates || null    }  } catch (error) {    console.error('Currency API error:', error)    return {      base: baseCurrency,      rates: {},      error: 'Failed to fetch exchange rates'    }  }}

Performance Optimization

To ensure your section components load quickly:

  1. Use parallel fetching with Promise.all() for independent data sources

  2. Implement appropriate caching strategies based on data freshness requirements

  3. Filter data server-side whenever possible to reduce payload size

  4. Consider data dependencies to avoid waterfall requests

  5. Return only what you need to minimize response size

  6. Handle errors gracefully with fallback content

  7. Monitor API response times and optimize slow requests

Troubleshooting

Common issues and their solutions:

1. Missing or undefined data

// Problem: Data is sometimes undefinedexport let loader = async ({ data }: ComponentLoaderArgs<ProductData>) => {  // ❌ This might cause an error if data.product is undefined  return await storefront.query(QUERY, {     variables: { handle: data.product.handle }  })}
// Solution: Add proper validationexport let loader = async ({ data }: ComponentLoaderArgs<ProductData>) => {  // ✅ Check for existence before accessing properties  if (!data?.product?.handle) return null    return await storefront.query(QUERY, {     variables: { handle: data.product.handle }  })}

2. Type errors in loader data

Use TypeScript to catch issues early:

// Define explicit types for your component datatype ProductCarouselData = {  products: Array<{ handle: string }>  autoplay: boolean  autoplaySpeed: number}
// Use the type in your loaderexport let loader = async ({   data }: ComponentLoaderArgs<ProductCarouselData>) => {  // TypeScript will now validate that data matches the expected structure}

3. API rate limiting issues

Implement proper caching and error handling:

export let loader = async ({ weaverse }: ComponentLoaderArgs<ApiData>) => {  const { fetchWithCache, storefront } = weaverse    try {    return await fetchWithCache('https://rate-limited-api.example/data', {      // Use longer cache times for rate-limited APIs      strategy: storefront.CacheCustom({        maxAge: 600, // 10 minutes        staleWhileRevalidate: 3600, // 1 hour        staleIfError: 86400 // 1 day - use stale data on errors for a while      })    })  } catch (error) {    // Check for rate limit errors    if (error instanceof Error && error.message.includes('rate limit')) {      console.warn('API rate limit reached, using fallback data')      return getFallbackData()    }    throw error  }}

To further enhance your understanding of Weaverse's data fetching capabilities and component development, explore these related guides:

Was this article helpful?