tag-soup/createDOMParser.js

The fastest pure JS SAX/DOM XML/HTML parser.

import { resolveTokenizerOptions } from './createTokenizer.js';
import { ParserError, tokenizeMarkup } from './tokenizeMarkup.js';
import { CDATASection, Comment, Document, DocumentFragment, DocumentType, Element, ProcessingInstruction, Text, } from 'flyweight-dom';
/**
 * Parses text as a DOM.
 *
 * @example
 * import { createDOMParser, htmlTokenizerOptions } from 'tag-soup';
 *
 * const parser = createDOMParser(htmlTokenizerOptions);
 *
 * parser.parseFragment('Hello, <b>Bob</b>!');
 * // ⮕ DocumentFragment
 *
 * @param options Parser options.
 * @group DOM
 */
export function createDOMParser(options = {}) {
    const { decodeText } = options;
    const documentOptions = Object.assign(Object.assign({}, resolveTokenizerOptions(options)), { decodeText });
    const fragmentOptions = Object.assign(Object.assign({}, documentOptions), { isFragment: true });
    return {
        parseDocument(input) {
            return parseDOM(input, documentOptions);
        },
        parseFragment(input) {
            return parseDOM(input, fragmentOptions);
        },
    };
}
/**
 * Parses text as a DOM.
 *
 * @param input The text to parse.
 * @param options Parser options.
 * @returns The document or document fragment node.
 */
export function parseDOM(input, options = {}) {
    const { isStrict, isFragment, decodeText = identity } = options;
    const root = isFragment ? new DocumentFragment() : new Document();
    let parent = root;
    let attributeName;
    let piTarget;
    const tokenCallback = (token, startIndex, endIndex) => {
        switch (token) {
            case 'TEXT':
                parent.appendChild(new Text(decodeText(input.substring(startIndex, endIndex))));
                break;
            case 'START_TAG_NAME':
                parent.appendChild((parent = new Element(input.substring(startIndex, endIndex))));
                break;
            case 'START_TAG_CLOSING':
                break;
            case 'START_TAG_SELF_CLOSING':
            case 'END_TAG_NAME':
                parent = parent.parentNode;
                break;
            case 'ATTRIBUTE_NAME':
                attributeName = input.substring(startIndex, endIndex);
                break;
            case 'ATTRIBUTE_VALUE':
                parent.setAttribute(attributeName, decodeText(input.substring(startIndex, endIndex)));
                break;
            case 'CDATA_SECTION':
                parent.appendChild(new CDATASection(input.substring(startIndex, endIndex)));
                break;
            case 'COMMENT':
                parent.appendChild(new Comment(decodeText(input.substring(startIndex, endIndex))));
                break;
            case 'DOCTYPE_NAME':
                parent.appendChild(new DocumentType(input.substring(startIndex, endIndex)));
                break;
            case 'PROCESSING_INSTRUCTION_TARGET':
                piTarget = input.substring(startIndex, endIndex);
                break;
            case 'PROCESSING_INSTRUCTION_DATA':
                parent.appendChild(new ProcessingInstruction(piTarget, input.substring(startIndex, endIndex)));
                break;
        }
    };
    tokenizeMarkup(input, tokenCallback, options);
    if (isStrict && parent !== root) {
        throw new ParserError('Expected an end tag.', input, input.length);
    }
    return root;
}
function identity(value) {
    return value;
}