deepsearchjs

Search for key in deeply-nested objects.

Installation

$ yarn add deepsearchjs

Usage

search<T>(object: T, any> | Array<any>, searchTerm: string): P;

| Name | Description | Required | Type | | ------------ | -------------------------------------------------------------------------------------------------------------------------------- | -------- | ----------------------------------------------------------------------------------------------------------------- | | object | Object to be searched | True | T = Record<any, any> \| Array<any> | | searchTerm | string to search for, a function that determines if an object key matches the query, or a regular expression to use for matching | True | string or ((key: string, value: any) => boolean) or RegExp | | output | Output object | --- | P = Record<key: string, any>Where, key is a valid object path(eg. a.b.c[0].d.e, etc..) |

Parameters

• obj (T): The object to search. It is also generic, meaning it can accept any type of object.
• query (string | ((key: string, value: any) => boolean) | RegExp): The query to search for. This can be a string, a function, or a regular expression.
  • If it is a string, the function will search for keys that end with that string.
  • If it is a function, it will determine if the key matches the query.
  • If it is a regular expression, it will test the key against that regular

Example

index.ts

import { search } from "deepsearchjs";
import data from "./data.json";

// string as query
search(data, "city");

// call back function as query
search(data, (key: string, value: string): boolean => /city/gi.test(key));

// regex as query
search(data, /city/gi);

/*
 * All three calls will have the same output:
 * {
 *   "users[0].address.city": "Jenkinsboro",
 *   "users[0].address.previousAddresses[0].city": "Folsom",
 *   ...
 * }
 */

TODO

• [ ] ci/cd: Automate versioning and releases
• [ ] feat: Optimize search algorithm to handle oversized datasets
• [ ] feat: Add a third parameter to the function to customize the search parameters ( leaf-node results only, etc... )
• [ ] feat: Search across values??

Benchmark

• text-based search x 22,188 ops/sec ±0.20% (94 runs sampled)
• regex-based search x 29,531 ops/sec ±0.11% (97 runs sampled)
• cb-based search x 9,243 ops/sec ±0.21% (96 runs sampled)
• Fastest method ( for the current release ): regex-based search :sparkles::tada:!!