parseCopyrightNotice.js

import { preferScalar } from '@kellnerd/es-utils/array/scalar.js';
import { getUniqueElementsByJSON } from '@kellnerd/es-utils/array/unique.js';
import { transform } from '@kellnerd/es-utils/string/transform.js';

const copyrightRE = /([©℗](?:\s*[&+]?\s*[©℗])?)(?:.+?;)?\s*(\d{4}(?:\s*[,&/+]\s*\d{4})*)?(?:[^,.]*\sby|\sthis\scompilation)?\s+/;

const legalInfoRE = /((?:(?:licen[sc]ed?\s(?:to|from)|(?:distributed|manufactured|marketed)(?:\sby)?)(?:\sand)?\s)+)/;

/** @type {CreditParserOptions} */
export const parserDefaults = {
	nameRE: /.+?(?:,?\s(?:LLC|LLP|(?:Corp|Inc|Ltd)\.?|Co\.(?:\sKG)?|(?:\p{Letter}\.){2,}))*/,
	nameSeparatorRE: /[/|](?=\s|\w{2})|\s[–-]\s/,
	terminatorRE: /$|(?=,|(?<!Bros)\.(?:\W|$)|\sunder\s)|(?<=(?<!Bros)\.)\W/,
};

/**
 * Extracts all copyright and legal information from the given text.
 * @param {string} text 
 * @param {Partial<CreditParserOptions>} [customOptions]
 */
export function parseCopyrightNotice(text, customOptions = {}) {
	// provide default options
	const options = {
		...parserDefaults,
		...customOptions,
	};

	/** @type {CopyrightItem[]} */
	const copyrightInfo = [];
	const namePattern = options.nameRE.source;
	const terminatorPattern = options.terminatorRE.source;

	// standardize copyright notice
	text = transform(text, [
		[/\(C\)/gi, '©'],
		[/\(P\)/gi, '℗'],

		// remove a-tisket's French quotes
		[/«(.+?)»/g, '$1'],

		// simplify region-specific copyrights
		[/for (.+?) and (.+?) for the world outside (?:of )?\1/g, '/ $2'],

		// simplify license text
		[/as licen[sc]ee for/gi, 'under license from'],

		// drop confusingly used ℗ symbols and text between ℗ symbol and year
		[/℗\s*(under\s)/gi, '$1'],
		[/(?<=℗\s*)digital remaster/gi, ''],

		// split © & ℗ with different years into two lines
		[/([©℗]\s*\d{4})\s*[&+]?\s*([©℗]\s*\d{4})(.+)$/g, '$1$3\n$2$3'],
	]);

	const copyrightMatches = text.matchAll(new RegExp(
		String.raw`${copyrightRE.source}(?:\s*[–-]\s+)?(${namePattern}(?:\s*/\s*${namePattern})*)(?:${terminatorPattern})`,
		'gimu'));

	for (const match of copyrightMatches) {
		const names = match[3].split(options.nameSeparatorRE).map((name) => name.trim());
		const types = match[1].split(/[&+]|(?<=[©℗])\s*(?=[©℗])/).map(cleanType);
		const years = match[2]?.split(/[,&/+]/).map((year) => year.trim());

		names.forEach((name) => {
			// skip fake copyrights which contain the release label
			if (/an?\s(.+?)\srelease/i.test(name)) return;

			copyrightInfo.push({
				name,
				types,
				year: preferScalar(years),
			});
		});
	}

	const legalInfoMatches = text.matchAll(new RegExp(
		String.raw`${legalInfoRE.source}(?:\s*[–-]\s+)?(${namePattern})(?:${terminatorPattern})`,
		'gimu'));

	for (const match of legalInfoMatches) {
		const types = match[1].split(/\sand\s/).map(cleanType);
		copyrightInfo.push({
			name: match[2],
			types,
		});
	}

	return getUniqueElementsByJSON(copyrightInfo);
}

/**
 * Cleans and standardizes the given free text copyright/legal type.
 * @param {string} type 
 */
function cleanType(type) {
	return transform(type.toLowerCase().trim(), [
		[/licen[sc]ed?/g, 'licensed'],
		[/(distributed|manufactured|marketed)(\sby)?/, '$1 by'],
	]);
}