@kentico-ericd/kontent-richtext-validation

Kontent Rich Text Validator

This package provides a function for validating the content of Rich Text elements. It can be used in conjunction with the Kontent Management JS SDK or our REST API to ensure the text is in the correct format before the data is sent to Kontent.

This can speed up the development process (e.g. when writing webhooks) and prevent errors while updating content, as errors can be resolved without the need for submitting repeated API calls, and could be integrated into testing scenarios.

Installation

Install via NPM:

npm i @kentico-ericd/kontent-richtext-validation --save

Usage

// Typescript + ES
import { KontentRichText } from '@kentico-ericd/kontent-richtext-validation'

// CommonJS
const { KontentRichText } = require('@kentico-ericd/kontent-richtext-validation');

The KontentRichText class provides a validate() function, which returns the following object:

{
  success: <boolean>,
  message: <string>
}

In the event that there is an error parsing the input, success will be false and message will contain the error and (in most cases) the exact position of the faulty text. On success, the message should contain the exact text that was passed to the function.

For example:

KontentRichText.validate('<p><a hrf="https://kontent.ai">Kentico Kontent</a></p>')

{
  success: false,
  message: "The A element must have a 'href' attribute or a 'data-email-address' attribute or a data attribute that identifies the link 
target ('data-item-id', 'data-item-external-id', 'data-asset-id' or 'data-asset-external-id'). (1, 30)"
}

We can see that the error occurred on line 1, col 30 which is the closing double-quote of the href attribute.

Example

This package could be integrated into the automatic translation webhook of the sample ExpressJS application. Perhaps we want to always return a 200 (success) response to Kontent when the webhook triggers, so that the webhook doesn't receive errors and stop processing. Or, we could validate the text and if there is some error, we can skip that item and continue with the rest of them.

Specifically, it could be added to this line where we get translated text from an external service. Maybe the external service accidentally changed some of the HTML? Let's check:

if (match.length > 0) {
  e.value = text.replace(/<br>/g, '<br/>');
  const validationResult = KontentRichText.validate(e.value);
  if(!validationResult.success) {
    const error = `Validation failed for item ${updatedVariant.data.item.id},`
      + ` lang ${targetLangCode}, element ${e.id}: ${validationResult.message}`;

    // We can log the error
    console.log(error);
    // We could write to some physical log file too
    writeToLogFile(error);
    // Send an email to devs?
    sendEmail('[e-mail]', 'Webhook error', error);
    // Skip this variant but continue with others
    stopProcessing = true;
    // or, maybe throw an error
    throw new Error(error);
  }
}