Lookup Table Library

A library for defining lookup tables to your Provengo models.

Lookup tables are a popular way of creating when/then tests, such as:

When customer_age is "8" and customer_hobby is "cars", then recommended_item should be "red toy car".

This library also allows to definition of ranges of values, as well as generation of test cases based on table rows.

Example

Consider the test case of a recommendation system for products, based on customer properties.

(in) Age	(in) Hobby	(out) Item	(out) Color
0	n/a	plush toy	light green
1-10	cars	toy car	red
1-10	animals	frog doll	green
18-24	travel	backpack	blue
(all)	beige	bicycle	beige
(all)	(all)	bicycle	white

based on the above table, we expect the system to offer a frog doll when asked to recommend a product for a 7-year old who likes animals, a toy car for a similar kid who likes cars, and a white bicycle for a 13 year old.

The above table is captured in code like so:

const myLut = LUT.create({                                                                (1)
    name: "Recommendations",
     ins: ["age","hobby"],
    outs: ["item","color"]
});

myLut.addRow({ ins:[0, LUT.ANY                  ], outs:["plush toy", "light green" ]});  (2)
myLut.addRow({ ins:[LUT.Range(1,10),  "cars"    ], outs:["toy car",   "red"         ]});
myLut.addRow({ ins:[LUT.Range(1,10),  "animals" ], outs:["frog doll", "green"       ]});
myLut.addRow({ ins:[LUT.Range(18,24), "travel"  ], outs:["backpack",  "blue"        ]});
myLut.addRow({ ins:[LUT.ANY,          "beige"   ], outs:["bicycle",   "beige"       ]});

myLut.addRow({                                                                            (3)
    ins:{age:LUT.ANY, hobby:"beige"},
    outs:["bicycle", "white"]
});
myLut.addRow({
    ins:[LUT.ANY, LUT.ANY],
    outs:{item:"bicycle", color:"white"}
});

let useCase = request(myLut.generateLutEvents());                                         (4)
let expected = myLut.lookup(useCase.data.ins);                                            (5)

1. Defining the lookup table: name, input and output column headers

2. Adding rows using column structure (technically, arrays)

3. Adding rows using named parameters (technically, objects). Arrays and objects can be used together.

4. Generating test cases from the lookup table

5. Looking up expected data

Tip	The full sample code is available in this repository. See table definition and client code.

Figure 1. Scenario space generated by the above code

Note	Bug reports, suggestions, and pull-requests are very welcome 😀

In this Repo

This is a standard Provengo project folder, so you can expect the usual file structure. Listed below are files of interest:

data/recommendations.js

Sample lookup table, shown in the above code;

lib/lut.js

The library itself. Put this file in your project’s lib/ directory in order to use it.

spec/js/sample-recommendations.js

A sample client code for this library. Creates test cases based on the lookup table defined in data/recommendations.js.

meta-spec/book-writer.js

A manual test book creation script. Uses a method from the library to handle LUT library events.

Reference.adoc

API Reference of the library.

Installation

To install, place the library’s source file, lib/lut.js, in the lib folder of your project. To download directly from this site, write the following snippet at your project’s root directory:

curl https://raw.githubusercontent.com/Provengo/Lib-LuT/main/lib/lut.js > lib/lut.js

Documentation

• Library Documentation

• Provengo’s Documentation

• Provengo’s YouTube Channel

• Provengo’s Main Site

---

Enjoy, and we hope you find it useful!

 — Provengo Engineering