README.md

Data Explorer

Overview

Data Explorer is an integral enhancement of the OpenSearch Dashboards that seeks to consolidate various data exploration facets into a unified platform. Built to provide an efficient data exploration experience, Data Explorer merges capabilities of different applications like Discover, Visbuilder, and Event Analytics into a singular platform.

Key Features

1. Unified Data Exploration: Data Explorer acts as a consolidated platform for all data exploration tasks, aiming to provide users with a seamless and efficient environment.
2. Extensibility: Provides an architecture that allows existing exploration apps to migrate with minimal changes.
3. Shared Utilities: Offers components and utilities that can be shared across different views.

Architecture and Integration

Data Explorer, at its core, is a shell for data exploration views. Here's a breakdown of the architecture and how it manages responsibilities:

Data Explorer Responsibilities:

1. Data Source: Acts as the central point for the data source being explored.
2. View Registry: Allows apps to register themselves as views and manages their display based on user selection.
3. State Management: Provides shared state management and hooks for underlying apps to register their state reducers.
4. Shared Utilities: Contains components and utilities that can be reused by various views.

View Responsibilities:

1. Metadata Storage: Handles the logic for storing metadata and its retrieval (Saved objects).
2. Data Fetching: Manages the logic for fetching data from the data source.
3. View specific state management: Handles view-specific state management and hooks into Data Explorer's state management.
4. Nav Options & Search/Query Bar: Manages the navigation options, time filter, and search bar.
5. View Specific Logic: Contains view-specific rendering and application logic.
6. Embeddables: Responsible for registering their embeddables.

Migrating Existing Applications:

Existing applications can migrate their data exploration views to Data Explorer. Such migrations involve:

1. Registering the application as a view.
2. Using Data Explorer's state management and data source.
3. Modifying routes to utilize Data Explorer's routes.
4. Adapting the UI to match Data Explorer's panel and canvas components.

Routing:

Existing routes for each view are expected to redirect to new routes prefixed with /data_explorer. E.g., existing Discover route will redirect to /data_explorer/discover.

View Registry:

For an application to be registered as a view within Data Explorer, it needs to adhere to the following data model:

interface ViewDefinition<T = any> {
  readonly id: string;
  readonly title: string;
  readonly ui?: {
    defaults: DefaultViewState | (() => DefaultViewState) | (() => Promise<DefaultViewState>);
    slice: Slice<T>;
  };
  readonly Canvas: LazyExoticComponent<(props: ViewProps) => React.ReactElement>;
  readonly Panel: LazyExoticComponent<(props: ViewProps) => React.ReactElement>;
  readonly Context: LazyExoticComponent<
    (props: React.PropsWithChildren<ViewProps>) => React.ReactElement
  >;
  readonly defaultPath: string;
  readonly appExtentions: {
    savedObject: {
      docTypes: [string];
      toListItem: (obj: { id: string; title: string }) => ViewListItem;
    };
  };
  readonly shouldShow?: (state: any) => boolean;
}

---

Original proposal: opensearch-project#4165