@gaddario98/react-queries
Una potente biblioteca basada en TanStack React Query y Jotai. Proporciona una API declarativa unificada para definir consultas, mutaciones y WebSockets.
Vedi su GitHub@gaddario98/react-queries
A powerful React Hook library built on top of TanStack React Query and Jotai. It provides a unified, declarative API to manage queries, mutations, and WebSockets with baked-in global state persistence, easy caching, and auto-mapping.
Features
- Unified API (
useApi): Manage multiple queries, mutations, and WebSockets through a single hook using a declarative configuration array. - Jotai Integration: Automatically syncs query and mutation states to Jotai atoms, allowing you to access cached states across different components easily.
- WebSocket Native Support: Built-in support for declaring WebSockets alongside your REST endpoints, cleanly unified into the same hook.
- Typescript First: Strongly typed query arrays mapped perfectly to your responses, inputs, and mutations (
allQuery.yourKey.data,allMutation.yourMutationKey.mutate). - Flexible Configuration: Support for custom request functions, global authentication control, local or global header merging, and built-in local persistence config.
- Backward Compatible: Ships with standalone
useQueryApianduseMutateApihooks for simpler usage.
Installation
yarn add @gaddario98/react-queries @tanstack/react-query jotai axios
(If you are within the monorepo, it's typically aliased to packages/react-base-core/queries or imported as @gaddario98/react-queries if already built.)
Global Setup
Wrap your application in the QueriesProvider. This component initializes the TanStack QueryClient and (optionally) the Persist provider.
import { QueriesProvider } from "@gaddario98/react-queries"; function App() { return ( <QueriesProvider> <YourAppComponents /> </QueriesProvider> ); }
Initializing the API Configuration
The package relies on an internal apiConfigAtom (built using @gaddario98/react-state) that holds the default endpoints, network request behaviour, and query client. Before fetching, ensure your layout/app initializes this if you need custom endpoints.
You can set up multiple endpoints, define default headers, add token validators, enable query persistence, configure WebSocket connections, and even set up auto-encryption/decryption for outgoing and incoming payloads.
import { useApiConfigState } from "@gaddario98/react-queries"; import { QueryClient } from "@tanstack/react-query"; // Setting up the base config somewhere early in your App tree const [config, setConfig] = useApiConfigState(); useEffect(() => { setConfig({ // 1. Define base endpoints here endpoints: { custom: "", api: "https://api.my-app.com", auth: "https://auth.my-app.com", }, // 2. Custom fetch function (defaults to internal axios wrapper) requestFn: customApiRequestFn, // 3. Optional: global auth validation validateAuthFn: () => !!localStorage.getItem("token"), // 4. Global headers appended to all requests defaultHeaders: { "Cache-Control": "no-cache", }, // 5. Tanstack global QueryClient instance queryClient: new QueryClient({ defaultOptions: { queries: { retry: 2 } }, }), // 6. Optional: Persister options for offline caching persistOptions: { persister: myAsyncStoragePersister, maxAge: 1000 * 60 * 60 * 24, // 24 hours }, // 7. Optional: Global WebSocket Configuration websocketConfig: { url: "wss://api.my-app.com/socket", autoConnect: true, onMessage: (msg) => console.log("Global socket msg", msg), }, // 8. Optional: Payload Encryption options encryption: { enabled: true, secretKey: "my-super-secret-key", encryptFn: customEncryptionFn, // Defaults to AES-GCM decryptFn: customDecryptionFn, }, // 9. Optional: Global notification handler interface showNotification: ({ message, type }) => toast(message, { type }), // 10. Persist internal Jotai atoms holding query states (default: true) persistQueries: true, }); }, []);
Basic Usage: Declarative Configurations (useApi)
The recommended pattern is to define an array of type QueriesArray consisting of your required queries, mutations, and websockets. Pass this definition to the useApi hook to get heavily typed state.
1. Define your Queries Array
import type { QueriesArray } from "@gaddario98/react-queries"; export const myFeatureQueries = [ { type: "query", key: "userProfile", queryConfig: { endpoint: ["api", "v1/users/me"], queryKey: ["user", "profile"], enabled: true, // map response from the network converter: (res: any) => ({ name: res.fullName }), }, }, { type: "mutation", key: "updateProfile", mutationConfig: { endpoint: ["api", "v1/users/me"], method: "PUT", queryKeyToInvalidate: ["user", "profile"], }, }, ] as const satisfies QueriesArray;
2. Consume in your Component
Pass your array to the useApi hook. Assign a scopeId to isolate Jotai atom keys so caching doesn't clash between different features using the same query names.
import { useApi } from "@gaddario98/react-queries"; const MyProfileComponent = () => { const { allQuery, allMutation, refreshQueries } = useApi(myFeatureQueries, { scopeId: "profile-page", }); const { data: profile, isLoading } = allQuery.userProfile; const updateProfileMutation = allMutation.updateProfile; if (isLoading) return <Spinner />; const handleUpdate = () => { updateProfileMutation.mutate({ body: { fullName: "New Name" } }); }; return ( <div> <h1>{profile?.name}</h1> <button onClick={handleUpdate}>Save</button> <button onClick={refreshQueries}>Reload Details</button> </div> ); };
Advanced Features
WebSockets
You can manage WebSockets inside the same array!
const queries = [ { type: "websocket", key: "notifications", endpoint: "wss://api.my-app.com/socket", autoConnect: true, onMessage: (msg) => console.log("New message:", msg), invalidateQueriesOnMessage: ["someQueryKey"], // auto-invalidates specific queries when messages arrive! }, ] as const satisfies QueriesArray; // Usage: const { allWebSocket } = useApi(queries, { scopeId: "notifications-feature" }); const { lastMessage, sendMessage, status } = allWebSocket.notifications;
Fine-Grained Subscriptions (useApiValues)
When working with large query objects or components that only need a tiny piece of the query state (e.g., just the isLoading flag or a nested field in data), useApiValues helps prevent unnecessary re-renders. It subscribes to Jotai atoms but only triggers a re-render when the specific path you request changes.
import { useApiValues } from "@gaddario98/react-queries"; import type { myFeatureQueries } from "./queries"; const LoadingIndicator = () => { const { get } = useApiValues<typeof myFeatureQueries>({ scopeId: "profile-page", }); // Only re-renders when `isLoading` changes, ignoring `data` or other states const isLoading = get("query", "userProfile.isLoading"); if (!isLoading) return null; return <Spinner />; };
Standalone Hooks
If you prefer dealing with queries one by one instead of a configuration array, you can use the standalone hooks:
import { useQueryApi, useMutateApi } from "@gaddario98/react-queries"; // Single Query const { data, isLoading, refetch } = useQueryApi( { endpoint: ["api", "v1/stats"], queryKey: ["stats"], enabled: true, }, "default-scope", ); // Single Mutation const mutation = useMutateApi( { endpoint: ["api", "v1/stats"], method: "POST", queryKeyToInvalidate: ["stats"], }, "default-scope", );
Jotai Interoperability
Every time any query executes via useApi or useQueryApi, its result is synchronously stored in global Jotai atoms. This prevents redundant loading flashes and provides deep cache accessibility across distinct unmounted pages.
To read raw values outside a component tree or react to cross-component data without refetching, you can fetch directly from Jotai atoms via queriesAtom and mutationsAtom or use helper utilities exported by the package.