+/**
+ * @description "or", like the regex `( f1 | f2 )` .
+ * It returns a function `f` of which the argument is`x`.
+ * if `f1(x)` is None, then `f` returns `f2(x)`. Otherwise,
+ * `F` returns `f1(x)`.
+ * @param f1 : 1st function to be compared
+ * @param f2 : 2nd function to be compared
+ * @returns:the combined function
+ */
+function orDo(f1, f2) {
+ return (x) => {
+ let f1x = (f1(x));
+ {
+ if (f1x._tag == "None") {
+ return f2(x);
+ }
+ else {
+ return f1x;
+ }
+ }
+ };
+}
+exports.orDo = orDo;
+/**
+* @description repeating matching function `f`
+* zero or more times, like the asterisk `*` in regex `f*` .
+* @param f : the function to be repeated 0+ times.
+* @returns:the combined function
+*/
+function zeroOrMoreDo(f) {
+ return (x) => {
+ var wrapped_old_x = { _tag: "Some", value: x };
+ var wrapped_new_x = wrapped_old_x;
+ while (wrapped_new_x._tag != "None") {
+ wrapped_old_x = wrapped_new_x;
+ wrapped_new_x = thenDo(wrapped_old_x, f);
+ }
+ ;
+ return wrapped_old_x;
+ };
+}
+exports.zeroOrMoreDo = zeroOrMoreDo;
+/**
+* @description Not. like the `^` inside regex of [^f].
+* returns a function `F(x)` such that if `f(x)` is `None`,
+* returns the x consuming a char; if `f(x)` is not None, F(x)
+* returns `None`.
+* @param f: the function forbidden to be matched.
+* @returns: combined function `F`.
+*/
+function notDo(f) {
+ return (x) => {
+ let wrapped_x = {
+ _tag: "Some",
+ value: x
+ };
+ let f_x = thenDo(wrapped_x, f);
+ if (f_x._tag != "None") {
+ return { _tag: "None" };
+ }
+ else {
+ return thenDo(wrapped_x, matchAny);
+ }
+ };
+}
+exports.notDo = notDo;
+/**
+ * if `x` is matched by `f` once, returns `f(x)`. Otherwise,
+ * returns x
+ * similar to `?` in regex `f?`.
+ * @param f : the function to be matched
+ * @returns return wrapped f(x)
+ */
+function zeroOrOnceDo(f) {
+ return (x) => {
+ var wrapped_old_x = { _tag: "Some", value: x };
+ var wrapped_new_x = thenDo(wrapped_old_x, f);
+ if (wrapped_new_x._tag != "None") {
+ return wrapped_new_x;
+ }
+ else {
+ return wrapped_old_x;
+ }
+ };
+}
+exports.zeroOrOnceDo = zeroOrOnceDo;
+function tokenize(input) {
+ var input_matchee_pair = toSome({ matched: "",
+ remained: input });
+ /**
+ * generate a parser of a basic term (b_term)
+ * @param pattern : the pattern parser
+ * @param token_type : the returning token type
+ * @returns a wrapped parser.
+ */
+ function bTerm(pattern, token_type) {
+ return (x) => {
+ let wrapped_x = toSome(x);
+ let result = pattern(wrapped_x);
+ if (result._tag == "Some") {
+ result.value.matched_type = token_type;
+ }
+ return result;
+ };
+ }
+ let d = matchRange('0', '9'); // \d
+ // [+-]
+ let plusMinus = orDo(match1Char('+'), match1Char('-'));
+ let s_aux = orDo(match1Char(' '), match1Char('\t')); // (" " | "\t")
+ // integer = ([+]|[-])?\d\d*
+ let integer = bTerm((x) => thenDo(thenDo(thenDo(x, zeroOrOnceDo(plusMinus)), d), zeroOrMoreDo(d)), TokenType.INT);
+ // space = [ \t]+
+ let space = bTerm((x) => thenDo(thenDo(x, s_aux), zeroOrMoreDo(s_aux)), TokenType.INT);
+ // newline = \r?\n
+ let newline = bTerm((x) => thenDo(thenDo(x, zeroOrOnceDo(match1Char('\r'))), match1Char('\n')), TokenType.NL);
+ // [_A-Za-z]
+ let idHead = orDo(orDo(matchRange('a', 'z'), matchRange('A', 'Z')), match1Char('_'));
+ let idRemained = orDo(idHead, matchRange('0', '9')); // [_A-Za-z0-9]
+ // id = [_A-Za-z][_A-Za-z0-9]*
+ let id = bTerm((x) => thenDo(thenDo(x, idHead), zeroOrMoreDo(idRemained)), TokenType.ID);
+ let doublequote = match1Char("\"");
+ // [\\][\"]
+ let escapeReverseSlash = (x) => thenDo(thenDo(toSome(x), match1Char("\\")), doublequote);
+ // ([\\]["]|[^\"])*
+ let stringInnerPattern = zeroOrMoreDo(orDo(escapeReverseSlash, notDo(match1Char("\""))));
+ // str = ["]([\\]["]|[^"])*["]
+ let str = bTerm((x) => thenDo(thenDo(thenDo(x, doublequote), stringInnerPattern), doublequote), TokenType.STR);
+ // float = [+-]?\d+[.]\d+
+ function floatPattern(x) {
+ return thenDo(thenDo(thenDo(thenDo(thenDo(thenDo(x, zeroOrOnceDo(plusMinus)), d), zeroOrMoreDo(d)), match1Char(".")), d), zeroOrMoreDo(d));
+ }
+ ;
+ let float = bTerm(floatPattern, TokenType.FLO);
+ // operators
+ // +.
+ let floatAdd = bTerm((x) => thenDo(thenDo(x, match1Char("+")), match1Char(".")), TokenType.F_ADD);
+ // +.
+ let floatSub = bTerm((x) => thenDo(thenDo(x, match1Char("-")), match1Char(".")), TokenType.F_SUB);
+ // *.
+ let floatMul = bTerm((x) => thenDo(thenDo(x, match1Char("*")), match1Char(".")), TokenType.F_MUL);
+ // /.
+ let floatDiv = bTerm((x) => thenDo(thenDo(x, match1Char("/")), match1Char(".")), TokenType.F_DIV);
+ // ==
+ let eq = bTerm((x) => thenDo(thenDo(x, match1Char("=")), match1Char("=")), TokenType.EQ);
+ // >=
+ let ge = bTerm((x) => thenDo(thenDo(x, match1Char(">")), match1Char("=")), TokenType.GE);
+ // <=
+ let le = bTerm((x) => thenDo(thenDo(x, match1Char("<")), match1Char("=")), TokenType.LE);
+ // ->
+ let rightArrow = bTerm((x) => thenDo(thenDo(x, match1Char("-")), match1Char(">")), TokenType.R_ARROW);
+ /**
+ * unary operator : generating the pattern of basic unary operator
+ * @param char : uniry char for the operator
+ * @param token_type : the corresponding token_type
+ */
+ function unaryOp(char, token_type) {
+ return bTerm((x) => thenDo(x, match1Char(char)), token_type);
+ }
+ ;
+ let intAdd = unaryOp('+', TokenType.I_ADD);
+ let intSub = unaryOp('-', TokenType.I_SUB);
+ let intMul = unaryOp('*', TokenType.I_MUL);
+ let intDiv = unaryOp('/', TokenType.I_DIV);
+ let lParen = unaryOp('(', TokenType.L_PAREN);
+ let rParen = unaryOp(')', TokenType.R_PAREN);
+ let lBracket = unaryOp('[', TokenType.L_BRACK);
+ let rBracket = unaryOp(']', TokenType.R_BRACK);
+ let lBrace = unaryOp('{', TokenType.L_BRACE);
+ let rBrace = unaryOp('}', TokenType.R_BRACE);
+ let comma = unaryOp(',', TokenType.COMMA);
+ let dot = unaryOp('.', TokenType.DOT);
+ let colon = unaryOp(':', TokenType.COLON);
+ let semicolon = unaryOp(';', TokenType.SEMI_C);
+ let at = unaryOp('@', TokenType.AT);
+ let hash = unaryOp('#', TokenType.HASH);
+ let set = unaryOp('=', TokenType.SET);
+ let greaterthan = unaryOp('>', TokenType.GT);
+ let lessthan = unaryOp('<', TokenType.LE);
+ let term = (token_list, x) => {
+ var ln = 1;
+ var col = 0;
+ var old_x = x;
+ let term_list = [float, newline, space, integer, str, id,
+ floatAdd, floatSub, floatMul, floatDiv,
+ intAdd, intSub, intMul, intDiv,
+ eq, ge, le, rightArrow,
+ lParen, rParen, lBracket, rBracket, lBrace, rBrace,
+ comma, dot, colon, semicolon, at, hash,
+ set, greaterthan, lessthan];
+ let term_aux = term_list.reduce((x, y) => orDo(x, y));
+ var new_x = thenDo(old_x, term_aux);
+ while (new_x._tag != "None") {
+ if (new_x.value.matched_type != TokenType.NL) {
+ col += new_x.value.matched.length;
+ token_list.push({ text: new_x.value.matched,
+ type: new_x.value.matched_type,
+ ln: ln,
+ col: col });
+ }
+ else {
+ col = 0;
+ ln += 1;
+ token_list.push({ text: new_x.value.matched,
+ type: new_x.value.matched_type,
+ ln: ln,
+ col: col });
+ }
+ old_x = toSome({ matched: "",
+ remained: new_x.value.remained });
+ new_x = thenDo(old_x, term_aux);
+ }
+ if (old_x.value.remained.length) {
+ console.log(token_list);
+ throw new Error("the code can't be tokenized is near Ln. " + ln + ", Col." + col
+ + ", starting with " + old_x.value.remained.substring(0, 10));
+ }
+ return token_list;
+ };
+ console.log(term([], input_matchee_pair));
+ // TODO: id, string, space, basic operator, 3 marks: @, {, }.
+}
+exports.tokenize = tokenize;