@emartech/data-aggregator-language

Data Aggregator Language

Execute aggregation expressions on a given data set.

What it Does

Given the following input data structure,

const input = [{ 
  date: '2017-09-16',
  reservations: {
    silver: 12,
    gold: 5
  }  
}, {
  date: '2017-09-17',
  reservations: {
    silver: 3,
    gold: 2
  }  
}]

the parser allows mixing aggregation expressions over the input with common arithmetic operators.

For example,

• an aggregation expression: SUM reservations.silver would yield 12 + 3 = 15.
• an aggregation expression mixed with arithmetic: SUM reservations.silver + 3 would yield (12 + 3) + 3 = 18

Usage

const aggregator = require('@emartech/data-aggregator-language')(input);
const result = aggregator('SUM reservations.silver')

Supported Syntax in Aggregation Expressions

Examples below are for the input defined above.

Constants

• LENGTH
  • LENGTH (yields 2)
  • LENGTH + 3 (yields 5)
• Any number literal
  • 5 (yields 5)

Unary Operators

• SUM
  • SUM reservations.silver (yields 15)
• LAST
  • LAST reservations.silver (yields 3)
• AVERAGE
  • AVERAGE reservations.silver (yields 7.5)

Binary Operators

+, -, *, /

For example, (LAST reservations.silver + 3) * 2 / 2 (yields 6)

Using with Webpack

Because of the way the underlying Chevrotain library is implemented, name mangling interferes with the parser's operation. To resolve this issue, webpacked host projects need to disable name mangling for the token names used in this grammar.

To help with this, the module exports these names in a tokens array.

To configure webpack, include the following in your webpack.config.js:

const { tokens } = require('@emartech/data-aggregator-language');

// other config

module.exports.optimization = {
    minimizer: [
      new UglifyJsPlugin({
        uglifyOptions: {
          mangle: {
            reserved: tokens
          }
        }
      })
    ]
  };

Commit Convention

Semantic release depends on the Angular commit message conventions. Please follow these.

Credit

Uses SAP/chevrotain for parsing.