在日常開發中，GraphQL Schema 設計最佳實踐的使用頻率越來越高。本文系統地講解其用法、原理和最佳化策略。

快速上手

實際專案中的用法會更復雜一些：

import { ApolloClient, InMemoryCache, createHttpLink } from '@apollo/client'
import { setContext } from '@apollo/client/link/context'

const httpLink = createHttpLink({ uri: '/graphql' })
const authLink = setContext((_, { headers }) => {
  const token = localStorage.getItem('token')
  return { headers: { ...headers, authorization: `Bearer ${token}` } }
})

const client = new ApolloClient({
  link: authLink.concat(httpLink),
  cache: new InMemoryCache({
    typePolicies: {
      Query: {
        fields: {
          users: { keyArgs: ['filter'], merge: (e, i) => ({ ...i, edges: [...(e?.edges||[]), ...i.edges] }) }
        }
      }
    }
  })
})

通過這種方式，程式碼的可測試性和可擴充套件性都得到了提升。

內部原理

以下是一個完整的示例：

import { ApolloClient, InMemoryCache, createHttpLink } from '@apollo/client'
import { setContext } from '@apollo/client/link/context'

const httpLink = createHttpLink({ uri: '/graphql' })
const authLink = setContext((_, { headers }) => {
  const token = localStorage.getItem('token')
  return { headers: { ...headers, authorization: `Bearer ${token}` } }
})

const client = new ApolloClient({
  link: authLink.concat(httpLink),
  cache: new InMemoryCache({
    typePolicies: {
      Query: {
        fields: {
          users: { keyArgs: ['filter'], merge: (e, i) => ({ ...i, edges: [...(e?.edges||[]), ...i.edges] }) }
        }
      }
    }
  })
})

注意邊界條件處理，這在生產環境中至關重要。

業務實戰

關鍵在於理解核心邏輯：

import { ApolloClient, InMemoryCache, createHttpLink } from '@apollo/client'
import { setContext } from '@apollo/client/link/context'

const httpLink = createHttpLink({ uri: '/graphql' })
const authLink = setContext((_, { headers }) => {
  const token = localStorage.getItem('token')
  return { headers: { ...headers, authorization: `Bearer ${token}` } }
})

const client = new ApolloClient({
  link: authLink.concat(httpLink),
  cache: new InMemoryCache({
    typePolicies: {
      Query: {
        fields: {
          users: { keyArgs: ['filter'], merge: (e, i) => ({ ...i, edges: [...(e?.edges||[]), ...i.edges] }) }
        }
      }
    }
  })
})

效能最佳化需要結合具體場景，不是所有情況都需要過度最佳化。

效能對比

我們可以通過以下方式來改進：

import { ApolloClient, InMemoryCache, createHttpLink } from '@apollo/client'
import { setContext } from '@apollo/client/link/context'

const httpLink = createHttpLink({ uri: '/graphql' })
const authLink = setContext((_, { headers }) => {
  const token = localStorage.getItem('token')
  return { headers: { ...headers, authorization: `Bearer ${token}` } }
})

const client = new ApolloClient({
  link: authLink.concat(httpLink),
  cache: new InMemoryCache({
    typePolicies: {
      Query: {
        fields: {
          users: { keyArgs: ['filter'], merge: (e, i) => ({ ...i, edges: [...(e?.edges||[]), ...i.edges] }) }
        }
      }
    }
  })
})

這套方案已經在線上穩定運行了半年以上，經過了實際驗證。

小結

• 程式碼示例僅供參考，需根據業務場景調整
• GraphQL Schema 設計最佳實踐不是銀彈，需要根據專案規模和技術棧選擇
• 理解底層原理比記住 API 更重要
• 生產環境使用前務必做好相容性驗證
• 團隊協作中約定和文件比技術本身更重要