The high-impact (popular) packages of npm.
- What is this?
- When should I use this?
- Install
- Use
- API
- Data
- Scripts
- Types
- Compatibility
- Related
- Contribute
- Security
- License
This package exposes the names of popular packages on the public npm registry. The definition of the term popular here is the same as what npm itself calls high-impact. They classify packages as having a high impact on the ecosystem when a package meets one or more of the following conditions:
- download count of 1 000 000 or more per week
- depended on by 500 or more other packages
Please use this for fun experiments when researching the npm registry.
This package is ESM only. In Node.js (version 16+), install with npm:
npm install npm-high-impact
In Deno with esm.sh
:
import {npmHighImpact} from 'https://esm.sh/npm-high-impact@1'
In browsers with esm.sh
:
<script type="module">
import {npmHighImpact} from 'https://esm.sh/npm-high-impact@1?bundle'
</script>
import {npmHighImpact} from 'npm-high-impact'
console.log(npmHighImpact.length)
console.log(npmHighImpact)
8133
[
'ansi-styles',
'supports-color',
'semver',
'chalk',
'debug',
'has-flag',
'tslib',
'color-convert',
'color-name',
'ms',
// …
]
This package exports the identifiers npmHighImpact
,
npmTopDependents
, and
npmTopDownloads
.
There is no default export.
List of top package names (Array<string>
).
Sorted by most downloaded first.
Includes (unique) packages from npmTopDependents
and npmTopDownloads
.
List of package names that are depended on a lot (Array<string>
).
Sorted by most dependents first.
List of package names that are downloaded a lot (Array<string>
).
Sorted by most downloaded first.
👉 Note: not all of these packages are popular. There are some false-positives, such that download counts can be gamed, and that
libraries.io
sometimes thinks that a fork of webpack or so is actually webpack.
This repo includes several scripts to crawl different services.
script/crawl-packages.js
follows an append-only database to find all the
changes to things in the npm registry.
We filter duplicates out, but still end up with ±2.5m “things”, which aren’t
all proper packages.
Later scripts will have to deal with them being missing.
The script takes like 12-18 hours to run (it finished somewhere at night). But the good news is that it’s additive: so the next time you run it, it’ll only pull in everything that changed since you last ran in.
It crawls replicate.npmjs.com
.
script/crawl-top-download-unscoped.js
and
script/crawl-top-download-scoped.js
look for download counts of all ±4.1m
packages on the registry.
Later scripts can filter the complete list to get the top packages.
The script takes like 30 hours to run.
About 10 hours is spent on ±3.2m unscoped packages.
Another 8 or so on ±900k scoped packages (yes, sad).
After filtering, the interesting data would result in about 6k packages.
It crawls the npm package download count API. Unscoped packages are crawled using the batch API to get 128 per request. Scoped packages are crawled with 20 HTTP requests at a time, as there is no batch API, and higher rates are limited by npm.
script/crawl-top-dependent.js
looks for packages that are depended on by
500 or more other packages.
The script takes like 30 minutes to run and currently gets about 3 000
packages.
It crawls the libraries.io
project search API, whose
results can also be browsed on the web.
Crawling stops paginating when a package is seen that is depended on by less
than 500 other packages.
You need an API key for libraries.io
, see their API docs for more info.
This package is fully typed with TypeScript. It exports no additional types.
This package is at least compatible with all maintained versions of Node.js. As of now, that is Node.js 16+. It also works in Deno and modern browsers.
npm-high-impact-cli
— find the popular npm packages someone maintainsnpm-esm-vs-cjs
— data on the share of ESM vs CJS on the public npm registry
Yes please! See How to Contribute to Open Source.
This package is safe.