X-Git-Url: https://git.kianting.info/?a=blobdiff_plain;f=src%2Findex.js;h=15e18628b9bd679b10e00ed1215069f029b1dfe2;hb=d4dfd2e99f564ca880d40172687986ea3ce757f0;hp=0e3686ca2732f5843159c62a956aa15071329fda;hpb=5f0943539e27b3e0d08e70559ea39017e6df330e;p=clo diff --git a/src/index.js b/src/index.js index 0e3686c..15e1862 100644 --- a/src/index.js +++ b/src/index.js @@ -23,195 +23,184 @@ var __importStar = (this && this.__importStar) || function (mod) { return result; }; Object.defineProperty(exports, "__esModule", { value: true }); -exports.m1TType = void 0; +exports.tkTreeToSExp = void 0; var fs = require('fs'); -const node_process_1 = require("node:process"); -const tk = __importStar(require("./tokenize.js")); const util = __importStar(require("util")); +const p = __importStar(require("typescript-parsec")); /** - * debug reprensenting + * + * # REPRESENTATION */ -let repr = (x) => { return util.inspect(x, { depth: null }); }; /** - * concated 2 `tkTree`s - * @param x the array to be concated - * @param y the item or array to ve concated - * @returns concated tkTree array, or thrown error if can't be concated. + * convert a `tkTree` AST to S-expr string + * @param t the `tkTree` + * @returns S-expr String */ -function concat(x, y) { - if (Array.isArray(x)) { - return x.concat(y); - } - else { - throw new Error("the tkTree can't be concated, because it's not an array."); - } -} -function slice(x, index, end) { - if (Array.isArray(x)) { - return x.slice(index, end); +function tkTreeToSExp(t) { + var str = ""; + if (Array.isArray(t)) { + let strArray = t.map((x) => tkTreeToSExp(x)); + str = "(" + strArray.join(" ") + ")"; } else { - throw new Error("the tkTree can't be concated, because it's not an array."); - } -} -/** - * @description - * match one token type. - * - * it returns a function which test if the type of first token of the `remained` part of - * the argument of the function is `typ` , if it's true, update the `TokenMatcheePair` wrapped - * in `Some`. Otherwise, it returns `None`. - * * @param typ : the type to be test. - * @returns the updated `TokenMatcheePair` wrapped in `Some(x)` or `None`. - */ -function m1TType(typ) { - return (m) => { - if (m.remained.length == 0) { - return { _tag: "None" }; - } - /** - * token to be matched - * */ - const ttbm = m.remained[0]; - if (ttbm.type == typ) { - let new_matched = m.matched.concat(ttbm); - let result = { - _tag: "Some", value: { - matched: new_matched, - remained: m.remained.slice(1), - ast: ([ttbm]), - } - }; - return result; + if (t === undefined) { + str = "%undefined"; } else { - return { _tag: "None" }; + str = t; } - }; + } + return str; } -exports.m1TType = m1TType; -; +exports.tkTreeToSExp = tkTreeToSExp; +/**inspect the inner of the representation. */ +let repr = (x) => { return util.inspect(x, { depth: null }); }; +var TokenKind; +(function (TokenKind) { + TokenKind[TokenKind["Seperator"] = 0] = "Seperator"; + TokenKind[TokenKind["Semicolon"] = 1] = "Semicolon"; + TokenKind[TokenKind["Number"] = 2] = "Number"; + TokenKind[TokenKind["Op"] = 3] = "Op"; + TokenKind[TokenKind["ExprMark"] = 4] = "ExprMark"; + TokenKind[TokenKind["ExcapeAt"] = 5] = "ExcapeAt"; + TokenKind[TokenKind["Paren"] = 6] = "Paren"; + TokenKind[TokenKind["SpaceNL"] = 7] = "SpaceNL"; + TokenKind[TokenKind["Id"] = 8] = "Id"; + TokenKind[TokenKind["Str"] = 9] = "Str"; + TokenKind[TokenKind["Comment"] = 10] = "Comment"; +})(TokenKind || (TokenKind = {})); /** - * type int + * Parsing */ -let tInt = m1TType(tk.TokenType.INT); -let tAdd = m1TType(tk.TokenType.I_ADD); -let tSub = m1TType(tk.TokenType.I_SUB); -let tMul = m1TType(tk.TokenType.I_MUL); -let tDiv = m1TType(tk.TokenType.I_DIV); -let tLParen = m1TType(tk.TokenType.L_PAREN); -let tRParen = m1TType(tk.TokenType.R_PAREN); -node_process_1.argv.forEach((val, index) => { - console.log(`${index}=${val}`); -}); +const lexer = p.buildLexer([ + [true, /^\d+(\.\d+)?/g, TokenKind.Number], + [true, /^\\\@/g, TokenKind.ExcapeAt], + [true, /^\/\*([^/]|\/[^*])*\*\//g, TokenKind.Comment], + [true, /^\;/g, TokenKind.Semicolon], + [true, /^[-][-][-]/g, TokenKind.Seperator], + [true, /^[\+\-\*\/\&\|\!\^\<\>\~\=\?]+/g, TokenKind.Op], + [true, /^\@/g, TokenKind.ExprMark], + [true, /^[()\[\]{}]/g, TokenKind.Paren], + [true, /^["]([\"]|[\\].)*["]/g, TokenKind.Str], + [true, /^[']([\']|[\\].)*[']/g, TokenKind.Str], + [true, /^[()\[\]{}]/g, TokenKind.Paren], + [true, /^[^\/\\\@\s\n\t\r;]+/g, TokenKind.Id], + [true, /^(\s|\n|\r|\t)+/g, TokenKind.SpaceNL], +]); /** - * like `m ==> f` in ocaml - * @param m matchee wrapped - * @param f matching function - * @returns wrapped result + * + * # TEST */ -function thenDo(m, f) { - if (m._tag == "None") { - return m; +const inputTxt = `import a as b; /*bacourt*/ +/* ba choir +ipsum lorem*/ + +import you as john; +--- + +臺中市\\\@ + +公園 +@1+2==3; + +console.log("122");@ + +山頂 +`; +const PROG = p.rule(); +const SEGMENT = p.rule(); +const IMPORT = p.rule(); +const IMPORTS = p.rule(); +const SEMICOLON = p.rule(); +const EXCAPE_AT = p.rule(); +const NOT_AT_TEXT = p.rule(); +const CONTENT = p.rule(); +let doubleMinus = { type: 'Punctuator', value: '--' }; +let doubleMinus2 = p.str('--'); +const TERM = p.rule(); +function applySegment(input) { + let unpackedInnerExprs = input[1].map((x) => { return x.text; }); + return ["%exprs", unpackedInnerExprs]; +} +function applySemiColon(value) { + return value.text; +} +function applyParts(first, second) { + return ["%clo", first, second[1]]; +} +function applyComment(value) { + return [value.text]; +} +function applyImport(input) { + let importTail = input[1].map(x => x.text); + return ["import"].concat(importTail); +} +; +/* +function applyImportComment(input: [Token,Token[], + tkTree, Token]) : tkTree{ + let importTail = input[1].map(x=>x.text); + let comment = [input[3].text]; + return ["import"].concat(importTail).concat(comment); +};*/ +function applyImports(input) { + let resultBody = [input[0]].concat(input[1]); + let resultWrapper = ["%import", resultBody]; + return resultWrapper; +} +; +function applyNotAtText(value) { + if (value.text == "\\\@") { + return '@'; } else { - var a = f(m.value); - if (a._tag == "Some") { - a.value.ast = concat(m.value.ast, a.value.ast); - } - return a; + return value.text; } } -/** - * like `f1 | f2` in regex - * @param f1 the first tried function - * @param f2 the second tried function - * @returns wrapped result - */ -function orDo(f1, f2) { - return (x) => { - let res1 = f1(x); - if (res1._tag == "Some") { - return res1; - } - else { - let res2 = f2(x); - return res2; - } - }; +; +function applyText(input) { + return ["%text", input]; +} +; +function applyContent(input) { + return ["%content", input]; +} +; +function applySpaceNL(value) { + return value.text; } /** - * aux function for midfix operator - * @param f function - * @param signal the rule name - * @returns + * IMPORTEE: Number, Op, Paren, Id, Str, Comment, */ -let midfix = (f, signal) => (x) => { - var a = f(x); - if (a._tag == "Some") { - let ast_tail = slice(a.value.ast, a.value.ast.length - 3); - let new_ast = [ast_tail]; - a.value.ast = new_ast; - console.log("+" + signal + "+" + repr(a)); - } - return a; -}; -let circumfix = (f, signal) => (x) => { - var a = f(x); - if (a._tag == "Some") { - let inner = a.value.ast[a.value.ast.length - 2]; - console.log("AST====" + repr(a.value.ast)); - let ast_middle = [inner]; - let new_ast = [ast_middle]; - a.value.ast = new_ast; - console.log("+" + signal + "+" + repr(a)); - } - return a; -}; -/** fac1 = "(" expr ")" */ -let fac1 = circumfix((x) => thenDo(thenDo(thenDo(tk.toSome(x), tLParen), expr), tRParen), "fac1"); -let fac2 = tInt; -let fac = orDo(fac1, fac2); +let IMPORTEE = p.alt(p.tok(TokenKind.Number), p.tok(TokenKind.Op), p.tok(TokenKind.Paren), p.tok(TokenKind.Id), p.tok(TokenKind.Str), p.tok(TokenKind.SpaceNL), p.tok(TokenKind.Comment)); +let NOT_AT = p.alt(p.tok(TokenKind.Seperator), p.tok(TokenKind.Semicolon), p.tok(TokenKind.Number), p.tok(TokenKind.ExcapeAt), p.tok(TokenKind.Op), p.tok(TokenKind.Paren), p.tok(TokenKind.SpaceNL), p.tok(TokenKind.Id), p.tok(TokenKind.Str), p.tok(TokenKind.Comment)); /** - * - * term1 = fac (MUL | DIV) fac + * PROG : IMPORTS '---' CONTENT; */ -let term1 = midfix((x) => thenDo(thenDo(thenDo(tk.toSome(x), fac), orDo(tMul, tDiv)), fac), "term1"); +PROG.setPattern(p.lrec_sc(IMPORTS, p.seq(p.str('---'), CONTENT), applyParts)); /** - * - * term2 = int MUL int + * NOT_AT_TEXT : NOT_AT */ -let term2 = fac; +NOT_AT_TEXT.setPattern(p.apply(NOT_AT, applyNotAtText)); +IMPORTS.setPattern(p.apply(p.seq(IMPORT, p.rep(IMPORT)), applyImports)); /** - * term = term1 | term2 + * IMPORT : + * 'import' IMPORTEE* SEMICOLON | + * COMMENT | */ -let term = orDo(term1, term2); +IMPORT.setPattern(p.alt(p.apply(p.seq(p.str('import'), p.rep_sc(IMPORTEE), SEMICOLON), applyImport), p.apply(p.tok(TokenKind.Comment), applyComment), p.apply(p.tok(TokenKind.SpaceNL), applySpaceNL))); /** - * - * expr1 = term ADD term + * SEMICOLON : ';'; */ -let expr1 = midfix((x) => thenDo(thenDo(thenDo(tk.toSome(x), term), orDo(tAdd, tSub)), term), "expr1"); +SEMICOLON.setPattern(p.apply(p.tok(TokenKind.Semicolon), applySemiColon)); /** - * expr2 = term + * SEGMENT : '@' NOT_AT* '@' | + * (NOT_AT_TEXT | EXCAPE_AT)* */ -let expr2 = term; +SEGMENT.setPattern(p.alt(p.apply(p.rep_sc(NOT_AT_TEXT), applyText), p.apply(p.seq(p.str('@'), p.rep(NOT_AT), p.str('@')), applySegment))); /** - * expr = expr1 | expr2 + * CONTENT : SEGMENT* */ -let expr = orDo(expr1, expr2); -let tokens = tk.tokenize("(4-(3/4))"); //tk.tokenize(argv[2]); -let tokensFiltered = tokens.filter((x) => { - return (x.type != tk.TokenType.NL - && x.type != tk.TokenType.SP); -}); -let wrappedTokens = tk.toSome({ - matched: [], - remained: tokensFiltered, - ast: [] -}); -let beta = expr({ - matched: [], - remained: tokensFiltered, - ast: [] -}); -console.log(repr(beta)); +CONTENT.setPattern(p.apply(p.rep(SEGMENT), applyContent)); +let tree = p.expectSingleResult(p.expectEOF(PROG.parse(lexer.parse(inputTxt)))); +console.log("RESULT=" + tkTreeToSExp(tree));