earley-sgf/dist/sgfTokenize.js

Early algorithm used to parse SGF file

function classify(prefix) {
    switch (prefix) {
        case 'SZ': /* board size */
        case 'KM': /* komi */
        case 'DT': /* date */
        case 'B': /* move */
        case 'W':
        case 'PB': /* player */
        case 'PW':
        case 'TB': /* territory */
        case 'TW':
        case 'TE':
        case 'AB': /* setup */
        case 'AW':
        case 'AE':
        case 'RE': /* result */
            return prefix;
        default:
            return 'prefix';
    }
}
/*
 *
 * The reserved characters are the brackets and the semicolon
 */
const reserved = /[\]\[\(\);]/;
/*
* The tokens are :  the reserved characters,  strings between reserved characters.
*/
export function* sgfTokenize(data) {
    let previous = '';
    while (data.length > 0) {
        let regexp = (previous === '[') ? ']' : reserved;
        const b = data.search(regexp);
        if (b < 0) {
            throw new Error('Invalid SGF ' + data);
        }
        if (0 < b) {
            let text = data.slice(0, b).trim();
            if (0 < text.length) {
                if (previous === '[') {
                    yield { token: 'text', text };
                }
                else {
                    yield { token: classify(text), text };
                }
            }
        }
        previous = data.at(b);
        yield { token: previous, text: previous };
        data = data.slice(b + 1);
    }
}