This is a Node module for loading structured data from a directory in the style of Jekyll's _data. Install it with:

npm install data-directory

The data-directory module exports two top-level functions:

1. load(dirname, callback) asynchronously loads all of the data in a directory and calls the callback function with the entire data structure when finished.

var datadir = require('data-directory');
datadir.load('path/to/data', function(error, data) {
  // your data here
});

1. proxy(dirname [, data]) creates a proxy object that loads data synchronously, and only as needed, from the same directory structure.

var datadir = require('data-directory');
var data = datadir.proxy('path/to/data');
// reads path/to/data/foo.{csv,json,tsv,ya?ml}
// or, if path/to/data/foo is a directory, returns a new proxy
var foo = data.foo;

⚠️ In order to use the proxy feature, you'll need to run Node with the --harmony-proxies flag.

Your directory should contain one or more files with the following extensions:

• .csv for comma-separated values
• .json for JSON
• .tsv for tab-separated values
• .yaml or .yml for YAML

You can read them all into a single data structure like this:

var datadir = require('data-directory');
datadir.load('_data', function(error, data) {
  if (error) return console.error('error:', error);
  console.log('data:', JSON.stringify(data, null, '  '));
});

Nested directories will introduce new levels in the data structure. For instance, if you data directory looks like this:

├─ bar.json
└─ baz
   └─ qux.csv

Then it should parse into a JSON structure that looks like:

{
  "bar": {
    // contents of bar.json
  },
  "baz": {
    "qux": [
      // rows in baz/qux.csv
    ]
  }
}