Migration to v2.0.0

Breaking changes

Programmatic API restructuring

The library has been refactored to provide a more direct and modern API. The previous version exported modules as namespaces, whereas v2.0.0 exports functions and constants directly.

Before (v1):

import { rules, validation } from "@rviscomi/capo.js";

const weight = rules.getWeight(element);
const isValid = validation.isValidElement(element);

After (v2):

import { getWeight, isValidElement, BrowserAdapter } from "@rviscomi/capo.js";

const adapter = new BrowserAdapter();
const weight = getWeight(element, adapter);
const isValid = isValidElement(element, adapter);

Removal of WebPageTest and BigQuery support

Support for WebPageTest and BigQuery has been removed from the core library to reduce maintenance burden and focus on the core analysis engine. This includes the removal of the webpagetest/ and bigquery/ directories.

Removal of capo.js snippet

The standalone capo.js snippet has been removed. Users are encouraged to use the browser extension or the programmatic API.

New features and improvements

New adapter system

capo.js now supports a flexible adapter system, allowing it to run on different DOM implementations.

• Built-in adapter: The BrowserAdapter is provided for standard browser DOM environments.
• Custom adapters: Consumers can implement their own adapters by extending the base AdapterInterface class. This enables capo.js to work with different parsers (like JSDOM or ASTs), unlocking its use in static analysis tools.

import { BrowserAdapter, analyzeHead } from "@rviscomi/capo.js";

const adapter = new BrowserAdapter();
const results = analyzeHead(document.head, adapter);

New validation rules

• Font preload missing crossorigin: capo.js now validates that font preloads include the crossorigin attribute.

Virtual console

The demo page now includes a virtual console, making it easier to see Capo output right in the docs, without needing to open the browser console.