Ngraminator

3.0.2 · active · verified Sun Apr 19

Ngraminator is a lightweight JavaScript utility designed to generate ngrams from arrays of words. It supports various module environments, including Node.js (CommonJS and ES Modules) and browsers (UMD), making it versatile for both backend and frontend text processing tasks. The current stable version is 3.0.2. The package underwent a significant architectural change in v3.0.1, migrating from webpack to Rollup for bundling, which introduced breaking changes in how the library is imported via CommonJS and accessed globally in UMD environments. Its primary differentiator is its small footprint and broad environment compatibility for generating n-grams of specified lengths from input arrays, without additional complex dependencies. It maintains a moderate release cadence, with recent updates focusing on build system improvements and module compatibility.

Common errors

```
TypeError: ngraminator is not a function
```
cause Attempting to use `require('ngraminator')` as a direct function call in CommonJS environments since v3.0.1, where it's now a named export.

fix Change your import from `const ngraminator = require('ngraminator')` to `const { ngraminator } = require('ngraminator')`.
```
ReferenceError: ngraminator is not defined
```
cause Accessing the `ngraminator` function directly in a browser environment after loading the UMD build (e.g., via `<script>`) for versions 3.0.1 and later.

fix Access the function via the `ngrm` global object: `ngrm.ngraminator(...)`.

Warnings

breaking The CommonJS `require()` syntax changed from a default import to a named import. Previously `const ngraminator = require('ngraminator')` worked, but now it requires destructuring.
Fix: Update your CommonJS import statements to `const { ngraminator } = require('ngraminator')`.
breaking For browser environments using the UMD build via a script tag, the global object has changed. The `ngraminator()` function is no longer directly accessible; it's now nested under a new global variable.
Fix: Access the function via `ngrm.ngraminator()` instead of `ngraminator()` in your browser scripts.
gotcha Version 3.0.1 and later changed the internal bundler from webpack to Rollup and the test framework from Tape to Ava/Playwright. While not directly affecting API usage, it signifies a major internal refactor that could lead to subtle behavioral differences or impact bundling toolchains.
Fix: Review your existing build processes if they interact with `ngraminator`'s internals or expect specific bundle structures (e.g., source maps, tree-shaking behavior).
deprecated A security update was released in version 2.0.3. Users on older v2.x versions should upgrade to at least 2.0.3 or preferably to the latest v3.x branch to ensure security patches are applied.
Fix: Upgrade to `npm install ngraminator@latest` or specifically `npm install ngraminator@^2.0.3` for v2.x series.

Install

npm install ngraminator npm
yarn add ngraminator yarn
pnpm add ngraminator pnpm

Imports

ngraminator
wrong:
```
import ngraminator from 'ngraminator'
```
correct:
```
import { ngraminator } from 'ngraminator'
```
ESM modules correctly export `ngraminator` as a named export since v3.0.1.
ngraminator
wrong:
```
const ngraminator = require('ngraminator')
```
correct:
```
const { ngraminator } = require('ngraminator')
```
Since v3.0.1, CommonJS imports require destructuring as `ngraminator` is a named export, not a default export.
ngrm.ngraminator
wrong:
```
// In browser script tag: ngraminator(wordArray, ngramLengthArray)
```
correct:
```
// In browser script tag: ngrm.ngraminator(wordArray, ngramLengthArray)
```
For UMD builds (e.g., via script tag), the global object has changed from `ngraminator` to `ngrm.ngraminator` since v3.0.1.

Quickstart

Demonstrates how to import `ngraminator` and generate various n-gram lengths from a string.

import { ngraminator } from 'ngraminator';

const text = "The quick brown fox jumps over the lazy dog";
const words = text.split(' ');

// Get unigrams (n=1), bigrams (n=2), and trigrams (n=3)
const ngrams = ngraminator(words, [1, 2, 3]);

console.log('Generated N-grams:');
ngrams.forEach(ngram => console.log(ngram.join(' ')));

// Example of specific ngram lengths and transformations
const str = "mary had a little lamb it's fleece";
const result = ngraminator(str.split(' '), [1, 2, 5]).map(item => item.join(' '));

console.log('\nSpecific N-grams and stringified output:');
console.log(result);

view raw JSON →