Koalati Results Validator

The ResultsValidator class can be used to ensure the validity of a Koalati tool's results. It is included and used by the @koalati/tool-template package, and to validate results before they are returned on Koalati.

To get started with the ResultsValidator, you can install it via NPM:

npm i @koalati/results-validator

Then, you can include it in your project and use it as such:

const ResultsValidator = require('@koalati/results-validator');

// ...

const results = yourResultsFunction();
const validator = new ResultsValidator();
const validationErrors = validator.checkResults(results);

// validator.checkResults returns an array of error messages (strings)
// if the array is empty, it means your results are valid
for (const error of validationErrors) {
    console.log('Results validation error: ' + error);
    // ... or other error handling of your choice
}

Tool result format and available properties

A tool's results are are simply an array of serializable objects. Each of those objects represent a test that the tool has carried out and contains the results for that test. Here are the properties that are allowed for those objects:

Property	Required	Allowed types	Description
uniqueName	Yes	string	A slug-like name for the test that is unique within a tool. This will be prefixed with the tool's name to generate a Koalati-wide unique name for the test.
title	Yes	string	A user-friendly title.
description	Yes	string	A user-friendly description of the test.
weight	Yes	float	The percentage of your tool's total score that comes from this test. Should be a float between 0 and 1.0
score	Yes	float	Score obtained by the tested page or website. Should be a float between 0 and 1.0
snippets	No	string[], ElementHandle[]	A list of strings to represent as code snippets in Koalati's results.
table	No	array[]	A two-dimensional array of data that will be represented as a table in Koalati's results. The first row should contain the column's headings.
recommendations	No	string, string[], array	Recommendation(s) telling the user what can be done to improve their results. If your recommendation contains dynamic value(s), you should use the array format and provide a message template (string) as the first element, and an object literal containing the values with the placeholders as keys. Ex.: ["Use XYZ to reduce page weight by %savings%.", { "%savings%": "6%" }]. A priority should also be provided as the third value of the array, with one one of the following values: null, "OPTIMIZATION", "ESSENTIAL", "ISSUE". If no prioority is specified, "OPTIMIZATION" is assumed.