all-stars

All-star lineup of Node-based FOSS contributors

all-stars is meant to be a lightweight data module containing pre-fetched author info for the maintainers of the top depended Node packages on npm.

Inspired by credits, all-stars was created to recognize some of the most prolific authors of Node-based FOSS, to honor them for the awesome work they've done that benefits us all.

One of the problems of giving credit where it is due is that any individual author may be known by several names. Sometimes authors change usernames or email addresses. Sometimes they use their real name, sometimes not. all-stars is designed to address this problem, combining pre-fetched, cross-referenced data with a simple API to resolve many possible identifiers to a known author.

all-stars can resolve any of the following to a known author with a single method:

• name or alias
• email address
• npm username
• GitHub username

If an author is resolved, an object representing that author is returned, containing the following information:

• all known names
• all known email addresses
• all known npm usernames
• all known GitHub usernames
• all known Twitter handles

Hopefully this can help normalize arbitrary string-based ids into a more usable structure (and reduce duplicates in credits).

Note that the data in this module was fetched from public resources that are freely available, but it could possibly be used for ill intent. Please be respectful and DBAD! Thanks.

Install

npm install --save all-stars

var allStars = require('all-stars')

Examples

Resolve author by string

var rvagg = allStars('Rod Vagg')
console.log(rvagg.toString())
//=> Rod Vagg <[email protected]> (npm: rvagg, GitHub: rvagg, Twitter: rvagg)

var othiym23 = allStars('[email protected]')
console.log(othiym23.summary())
//=> Forrest L Norvell <[email protected]>

var substack = allStars('substack')
console.log(substack.emails)
//=> [ '[email protected]', '[email protected]' ]

var jongleberry = allStars('jongleberry')
console.log(jongleberry.npmUsers)
//=> [ 'jongleberry', 'jonathanong' ]

var tj = allStars('tjholowaychuk')
console.log(tj.githubUsers)
//=> [ 'tj' ]

var bcoe = allStars('Benjamin E. Coe')
console.log(bcoe.names)
//=> [ 'Benjamin Coe', 'Ben Coe', 'Benjamin E. Coe' ]

var sindre = allStars('[email protected]')
console.log(sindre.twitters)
//=> [ 'sindresorhus' ]

Resolve author by array

Iterates over elements and returns first match.

var indutny = allStars([ 'Fedor', '[email protected]' ])
console.log('%s', indutny)
//=> Fedor Indutny <[email protected]> (npm: indutny, GitHub: indutny, Twitter: indutny)

Resolve author by object

Iterates over object property values and returns first match.

var domenic = allStars({ name: 'Domenic', email: '[email protected]' })
console.log(domenic.summary())
//=> Domenic Denicola <[email protected]>

API

allStars(query)

Synchronously resolve query to an individual author, if known.

• Returns: allStars.AllStar object or null
• query: string|array|object, identifier(s) to resolve

allStars.AllStar

Object returned from allStars() method. Contains the following:

• email: string (may be null), first email address from emails array
• emails: array of strings (may be null), all known email addresses
• githubUser: string (may be null), first GitHub username from array
• githubUsers: array of strings (may be null), all known GitHub usernames
• id: string, identifying key of this author, typically the preferred npm username
• name: string (may be null), first name from names array
• names: array of strings (may be null), all known names
• npmUser: string (may be null), first npm username from array
• npmUsers: array of strings (may be null), all known npm usernames
• subset: function, returns an object as a subset of this object. Useful for assigning or merging with another object. The subset object contains:
  • name: string or null, same as name
  • email: string or null, same as email
  • npm: string or null, same as npmUser
  • github: string or null, same as githubUser
  • twitter: string or null, same as twitter
• summary([all]): function, returns first value from each array, concatenated as string
  • all: boolean, true if you want all fields, false if you just want name and email
• toString(): function, alias for summary(true)
• twitter: string (may be null), first Twitter handle from array
• twitters: array of strings (may be null), all known Twitter handles

Data Fetching/Generation

The git repository contains CLI tools for fetching and generating the static data that is packaged within this module. The approach was intended to be as objective as possible, though some guidance for accurate cross-referencing or gap-filling proved necessary. Here's the basic process:

1. Fetch top depended packages, scraped from npm's website

   Executed via cli/genPackages.js. Number of packages to fetch is configurable but defaults to 150.

   Package list is written to generated/packages_YYYYMMDD_HHmmss.json, which can be copied to packages.json.

2. Fetch author info for a given set of packages

   Executed via cli/genAuthors.js. Package list to fetch authors for is configurable but defaults to the most recent generated packages json file.

   Attempts to collect names, emails, and usernames from the following locations:

   1. Package maintainer info, pulled from registry.npmjs.org

   2. Curated set of aliases from this module's source code (aliases.json)

   3. npm user profile, scraped from npm's site

   4. GitHub user profile, using GitHub's API

   Author info is written to generated/authors_YYYYMMDD_HHmmss.json, which can be copied to authors.json.

   Note that this process sends many concurrent requests that can accidentally DoS attack the resources (particularly npm). May need to run a few different times until no 503s are received. To avoid throttling by GitHub's API, you will need to provide a Personal Access Token, which can either be specified via CLI or exported as GITHUB_TOKEN env var.

After author data has been generated and copied over to authors.json, a prepublish script should be run to create an index_authors.json file, which is used by the main API. This is accomplished via npm run prepublish.

License

ISC © Andrew Goode and Contributors