chair-flight Docsearch React

A clone of Algolia's React DocSearch, but it accepts any search backend, AND IS INCOMPATIBLE WITH IE, in order not to have to polyfill regeneratorRuntime.

Installation

yarn add @tefkah/chair-flight-docsearch-react
# or
npm install @tefkah/chair-flight-docsearch-react

Get started

The api for the component is pretty much the same as the original docsearch-react except that instead of providing an appId and apiKey you are supposed to pass an arbitrary search function:

import { DocSearch } from '@tefkah/chair-flight-docsearch-react';

import '@docsearch/css';

function App() {
  return (
    <DocSearch
      search={async (queryString) =>
        myCustomBackend.search(queryString).map(toDocSearchHit)
      }
      indexName="YOUR_INDEX_NAME_USED_FOR_LOCAL_CACHE_PURPOSES"
    />
  );
}

export default App;

The search function is expected to return an array of DocSearchHit, the internal type used by algolia documents. You will most likely have to write a function mapping your search results into this specific format:

export declare type DocSearchHit = {
  objectID: string;
  content: string | null;
  url: string;
  url_without_anchor: string;
  type: ContentType;
  anchor: string | null;
  hierarchy: {
    lvl0: string;
    lvl1: string;
    lvl2: string | null;
    lvl3: string | null;
    lvl4: string | null;
    lvl5: string | null;
    lvl6: string | null;
  };
  _highlightResult: DocSearchHitHighlightResult;
  _snippetResult: DocSearchHitSnippetResult;
  _rankingInfo?: {
    promoted: boolean;
    nbTypos: number;
    firstMatchedWord: number;
    proximityDistance?: number;
    geoDistance: number;
    geoPrecision?: number;
    nbExactWords: number;
    words: number;
    filters: number;
    userScore: number;
    matchedGeoLocation?: {
      lat: number;
      lng: number;
      distance: number;
    };
  };
  _distinctSeqID?: number;
};

Note: the search engine is fairly robust. Most of these params can be considered optional with no decrease of functionality.

Documentation

All other features are inline with docsearch, so I advise you to look in the original docs:

Read documentation →