const myConsonant = "\u1000-\u1021"; // "က-အ"

const enChar = "a-zA-Z0-9";

// "ဣဤဥဦဧဩဪဿ၌၍၏၀-၉၊။!-/:-@[-`{-~\s"
const otherChar = "\u1023\u1024\u1025\u1026\u1027\u1029\u102a\u103f\u104c\u104d\u104f\u1040-\u1049\u104a\u104b!-/:-@\\[-`\\{-~\\s";

const ssSymbol = "\u1039";

const ngaThat = "\u1004\u103a";

const aThat = "\u103a";

// Regular expression pattern for Myanmar syllable breaking
// *** a consonant not after a subscript symbol AND a consonant is not
// followed by a-That character or a subscript symbol
const BREAK_PATTERN = new RegExp("((?!" + ssSymbol + ")[" + myConsonant + "](?![" + aThat + ssSymbol + "])" + "|[" + enChar + otherChar + "])", "mg");

export function _segment(raw: string): string[] {
    const outputs: string[] = raw.replace(BREAK_PATTERN, "\uD835\uDD4A$1").split("\uD835\uDD4A");
    if (outputs.length > 0) {
        outputs.shift();
    }
    return outputs;
}

export function segment(text: string, separator: string): string{
    return text.replace(BREAK_PATTERN, separator + "$1");
}