detective-amd

What is detective-amd?

The detective-amd package is designed for analyzing AMD (Asynchronous Module Definition) style modules. It allows developers to parse AMD module definitions to extract dependencies, providing a way to understand and manipulate the dependency tree of projects using this module format.

What are detective-amd's main functionalities?

Extracting dependencies from AMD modules

This feature allows you to pass AMD module source code as a string to the detective function, which returns an array of the module's dependencies. It's useful for analyzing and understanding module dependencies.

const detective = require('detective-amd');
const src = "define(['a', 'b'], function (a, b) {});";
const dependencies = detective(src);
// dependencies would be ['a', 'b']

Other packages similar to detective-amd

madge

Madge is a tool for generating a visual graph of module dependencies, or finding circular dependencies in JavaScript and TypeScript projects. It supports AMD, CommonJS, and ES6 module formats. Compared to detective-amd, Madge offers a broader range of functionalities including visualization and support for multiple module formats.

dependency-tree

dependency-tree is a utility for generating a tree of dependencies for a given file. It supports various module formats including AMD, CommonJS, and ES6. Unlike detective-amd, which focuses on AMD modules, dependency-tree provides a more general approach to dependency analysis across different module types.

README

Detective-AMD

Returns a list of dependencies for a given JavaScript file or AST using any of the AMD module syntaxes.

Inspired by substack/node-detective but built for AMD.

npm install detective-amd

• Supports JSX code via node-source-walk.

Usage

Let's say we have the following file definitions:

// a.js
define(['./b', './c'], function (b, c) {
  console.log(b, c);
});

// b.js
define({
  name: 'foo'
});

// c.js
define(function () {
  return 'bar';
});

Here's how you can grab the list of dependencies of a.js synchronously.

const detective = require('detective-amd');

const srca = fs.readFileSync('a.js', 'utf8');

// Pass in the source code or an AST (if you've already parsed the file)
console.log(detective(srca)); // prints ['./b', './c']

You may also (optionally) configure the detective via a second object argument detective(src, options) that supports the following options:

• skipLazyLoaded: (Boolean) whether or not to omit inner requires in the list of extracted dependencies.

• Note: this does not affect the REM form since those inner requires are not "lazily" fetched.

Syntax Support

Supports the 4 forms of AMD module syntax:

• "named": define('name', [deps], func)
• "dependency list": define([deps], func)
• "factory": define(func(require))
• "no dependencies": define({})

Extra forms:

• "driver script" (or entry-point) syntax: require([deps], func)
• "REM" (or CommonJS-like) form: define(function(require, exports, module) {}).

Also handles dynamically loaded dependencies (ex: inner requires).

Supports driver scripts

You can also find the dependencies from a script that has a top-level require (an app initialization/driver/entry-point script):

require([
  './a'
], function (a) {
  // My app will get booted up from here
});

Expression-based requires

If there's a require call that doesn't have a string literal but an expression, a string (escodegen-generated) representation will be returned.

For example, if a.js was of the "factory" form and contained a dynamic module name:

// a.js

define(function (require) {
  // Assume str is some variable that gets set to a string dynamically
  // const str = ...

  const b = require('./' + str);
  const c = require('./c');

  console.log(b, c);
});

The dependency list will be: [ '\'./\' + str', './c' ]

• Even though that string representation isn't incredibly useful, it's still added to the list to represent/count that dependency