gofront 0.0.4 → 0.0.5

package/src/minifier.js

@@ -0,0 +1,698 @@
+// GoFront built-in minifier
+// Operates on JS string output (post-CodeGen), no AST needed.
+
+const JS_KEYWORDS = new Set([
+	"break",
+	"case",
+	"catch",
+	"class",
+	"const",
+	"continue",
+	"debugger",
+	"default",
+	"delete",
+	"do",
+	"else",
+	"enum",
+	"export",
+	"extends",
+	"false",
+	"finally",
+	"for",
+	"function",
+	"if",
+	"import",
+	"in",
+	"instanceof",
+	"let",
+	"new",
+	"null",
+	"of",
+	"return",
+	"super",
+	"switch",
+	"this",
+	"throw",
+	"true",
+	"try",
+	"typeof",
+	"var",
+	"void",
+	"while",
+	"with",
+	"yield",
+	"async",
+	"await",
+	"static",
+]);
+
+// Keywords that must keep a space after them when followed by an identifier/literal
+const SPACE_AFTER_KW = new Set([
+	"return",
+	"let",
+	"const",
+	"var",
+	"typeof",
+	"new",
+	"case",
+	"throw",
+	"void",
+	"delete",
+	"in",
+	"of",
+	"instanceof",
+	"function",
+	"class",
+	"extends",
+	"import",
+	"export",
+	"async",
+	"await",
+	"yield",
+	"static",
+]);
+
+// ── Tokenizer ────────────────────────────────────────────────
+// Splits JS source into tokens preserving strings, template literals, and regexes.
+
+function tokenize(code) {
+	const tokens = [];
+	let i = 0;
+	const len = code.length;
+
+	while (i < len) {
+		const ch = code[i];
+
+		// Line comment
+		if (ch === "/" && code[i + 1] === "/") {
+			let end = code.indexOf("\n", i);
+			if (end === -1) end = len;
+			tokens.push({ type: "comment", value: code.slice(i, end) });
+			i = end;
+			continue;
+		}
+
+		// Block comment
+		if (ch === "/" && code[i + 1] === "*") {
+			const end = code.indexOf("*/", i + 2);
+			const closeAt = end === -1 ? len : end + 2;
+			tokens.push({ type: "comment", value: code.slice(i, closeAt) });
+			i = closeAt;
+			continue;
+		}
+
+		// Regex literal — only after certain tokens
+		if (ch === "/") {
+			const prev = lastNonWhitespaceToken(tokens);
+			const prevVal = prev?.value;
+			const isRegexContext =
+				!prev ||
+				prev.type === "op" ||
+				prevVal === "(" ||
+				prevVal === "[" ||
+				prevVal === "," ||
+				prevVal === ";" ||
+				prevVal === "!" ||
+				prevVal === "=" ||
+				prevVal === "==" ||
+				prevVal === "===" ||
+				prevVal === "!=" ||
+				prevVal === "!==" ||
+				prevVal === ":" ||
+				prevVal === "return" ||
+				prevVal === "case" ||
+				prevVal === "typeof" ||
+				prevVal === "{" ||
+				prevVal === "&&" ||
+				prevVal === "||";
+			if (isRegexContext) {
+				let j = i + 1;
+				let escaped = false;
+				let inCharClass = false;
+				while (j < len) {
+					const rc = code[j];
+					if (escaped) {
+						escaped = false;
+					} else if (rc === "\\") {
+						escaped = true;
+					} else if (rc === "[") {
+						inCharClass = true;
+					} else if (rc === "]") {
+						inCharClass = false;
+					} else if (rc === "/" && !inCharClass) {
+						j++;
+						break;
+					}
+					j++;
+				}
+				// consume flags
+				while (j < len && /[gimsuy]/.test(code[j])) j++;
+				tokens.push({ type: "regex", value: code.slice(i, j) });
+				i = j;
+				continue;
+			}
+		}
+
+		// String literals (double or single quote)
+		if (ch === '"' || ch === "'") {
+			let j = i + 1;
+			while (j < len) {
+				if (code[j] === "\\" && j + 1 < len) {
+					j += 2;
+				} else if (code[j] === ch) {
+					j++;
+					break;
+				} else {
+					j++;
+				}
+			}
+			tokens.push({ type: "string", value: code.slice(i, j) });
+			i = j;
+			continue;
+		}
+
+		// Template literal
+		if (ch === "`") {
+			let j = i + 1;
+			let depth = 1;
+			let braceDepth = 0;
+			while (j < len && depth > 0) {
+				if (code[j] === "\\" && j + 1 < len) {
+					j += 2;
+					continue;
+				}
+				if (braceDepth === 0 && code[j] === "`") {
+					depth--;
+					j++;
+					continue;
+				}
+				if (code[j] === "$" && code[j + 1] === "{" && braceDepth === 0) {
+					braceDepth = 1;
+					j += 2;
+					continue;
+				}
+				if (braceDepth > 0) {
+					if (code[j] === "{") braceDepth++;
+					else if (code[j] === "}") {
+						braceDepth--;
+					}
+				}
+				j++;
+			}
+			tokens.push({ type: "template", value: code.slice(i, j) });
+			i = j;
+			continue;
+		}
+
+		// Whitespace
+		if (/\s/.test(ch)) {
+			let j = i + 1;
+			while (j < len && /\s/.test(code[j])) j++;
+			tokens.push({ type: "ws", value: code.slice(i, j) });
+			i = j;
+			continue;
+		}
+
+		// Identifier or keyword
+		if (/[a-zA-Z_$]/.test(ch)) {
+			let j = i + 1;
+			while (j < len && /[a-zA-Z0-9_$]/.test(code[j])) j++;
+			const word = code.slice(i, j);
+			tokens.push({
+				type: JS_KEYWORDS.has(word) ? "kw" : "ident",
+				value: word,
+			});
+			i = j;
+			continue;
+		}
+
+		// Number
+		if (
+			/[0-9]/.test(ch) ||
+			(ch === "." && i + 1 < len && /[0-9]/.test(code[i + 1]))
+		) {
+			let j = i + 1;
+			// hex
+			if (ch === "0" && j < len && (code[j] === "x" || code[j] === "X")) {
+				j++;
+				while (j < len && /[0-9a-fA-F]/.test(code[j])) j++;
+			} else {
+				while (j < len && /[0-9]/.test(code[j])) j++;
+				if (j < len && code[j] === ".") {
+					j++;
+					while (j < len && /[0-9]/.test(code[j])) j++;
+				}
+				if (j < len && (code[j] === "e" || code[j] === "E")) {
+					j++;
+					if (j < len && (code[j] === "+" || code[j] === "-")) j++;
+					while (j < len && /[0-9]/.test(code[j])) j++;
+				}
+			}
+			tokens.push({ type: "num", value: code.slice(i, j) });
+			i = j;
+			continue;
+		}
+
+		// Multi-char operators
+		const two = code.slice(i, i + 2);
+		const three = code.slice(i, i + 3);
+		if (
+			three === "===" ||
+			three === "!==" ||
+			three === "..." ||
+			three === ">>>" ||
+			three === "**="
+		) {
+			tokens.push({ type: "op", value: three });
+			i += 3;
+			continue;
+		}
+		if (
+			two === "==" ||
+			two === "!=" ||
+			two === "<=" ||
+			two === ">=" ||
+			two === "&&" ||
+			two === "||" ||
+			two === "++" ||
+			two === "--" ||
+			two === "+=" ||
+			two === "-=" ||
+			two === "*=" ||
+			two === "/=" ||
+			two === "=>" ||
+			two === "**" ||
+			two === "??" ||
+			two === "?." ||
+			two === ">>" ||
+			two === "<<"
+		) {
+			tokens.push({ type: "op", value: two });
+			i += 2;
+			continue;
+		}
+
+		// Single-char operator/punctuation
+		tokens.push({ type: "op", value: ch });
+		i += 1;
+	}
+
+	return tokens;
+}
+
+function lastNonWhitespaceToken(tokens) {
+	for (let i = tokens.length - 1; i >= 0; i--) {
+		if (tokens[i].type !== "ws" && tokens[i].type !== "comment")
+			return tokens[i];
+	}
+	return null;
+}
+
+// ── Stage 1 & 2: Compress ────────────────────────────────────
+
+function isIdentLike(type) {
+	return type === "ident" || type === "kw" || type === "num";
+}
+
+function needsSpaceBetween(left, right) {
+	// Keywords that need space after when followed by ident/num/string/keyword
+	if (left.type === "kw" && SPACE_AFTER_KW.has(left.value)) {
+		if (
+			right.type === "ident" ||
+			right.type === "kw" ||
+			right.type === "num" ||
+			right.type === "string" ||
+			right.type === "template" ||
+			right.type === "regex"
+		) {
+			return true;
+		}
+		// e.g., `return(` or `return!` — no space needed
+		if (
+			right.type === "op" &&
+			(right.value === "(" ||
+				right.value === "[" ||
+				right.value === "!" ||
+				right.value === "-" ||
+				right.value === "+")
+		) {
+			// "return(" is fine, "delete(" is fine, "new(" is fine
+			// But "function(" needs no space, "class{" needs no space... actually they do need space before name
+			return false;
+		}
+	}
+
+	// Two adjacent ident-like tokens need a space
+	if (isIdentLike(left.type) && isIdentLike(right.type)) return true;
+
+	// ident followed by string/template/regex needs space in some cases
+	// but generally not for GoFront output
+
+	// Prevent ++ or -- from merging with adjacent + or -
+	if (
+		(left.value === "+" && right.value === "+") ||
+		(left.value === "-" && right.value === "-") ||
+		(left.value === "+" && right.value === "++") ||
+		(left.value === "-" && right.value === "--") ||
+		(left.value === "++" && right.value === "+") ||
+		(left.value === "--" && right.value === "-")
+	) {
+		return true;
+	}
+
+	return false;
+}
+
+function compress(code) {
+	const tokens = tokenize(code);
+	let out = "";
+
+	for (let i = 0; i < tokens.length; i++) {
+		const tok = tokens[i];
+
+		// Stage 1: strip comments
+		if (tok.type === "comment") continue;
+		// Stage 1: strip whitespace (handled by needsSpaceBetween)
+		if (tok.type === "ws") continue;
+
+		// Preserve strings, templates, regexes verbatim
+		if (
+			tok.type === "string" ||
+			tok.type === "template" ||
+			tok.type === "regex"
+		) {
+			if (out.length > 0) {
+				const lastChar = out[out.length - 1];
+				// Need space if previous was ident-like
+				if (/[a-zA-Z0-9_$]/.test(lastChar) && tok.type !== "regex") {
+					// Check if preceding token is a keyword needing space
+					const prevTok = findPrevNonWs(tokens, i);
+					if (
+						prevTok &&
+						prevTok.type === "kw" &&
+						SPACE_AFTER_KW.has(prevTok.value)
+					) {
+						out += " ";
+					}
+				}
+			}
+			out += tok.value;
+			continue;
+		}
+
+		// For all other tokens, check if space is needed
+		if (out.length > 0) {
+			const prevTok = findPrevNonWs(tokens, i);
+			if (prevTok && needsSpaceBetween(prevTok, tok)) {
+				out += " ";
+			}
+		}
+
+		out += tok.value;
+	}
+
+	return out;
+}
+
+function findPrevNonWs(tokens, index) {
+	for (let i = index - 1; i >= 0; i--) {
+		if (tokens[i].type !== "ws" && tokens[i].type !== "comment")
+			return tokens[i];
+	}
+	return null;
+}
+
+// ── Stage 3: Identifier mangling ─────────────────────────────
+
+function mangle(code) {
+	const tokens = tokenize(code);
+
+	// Collect local variable declarations and their scopes
+	const nameGen = shortNameGeneratorFn();
+	const renameMap = new Map();
+
+	// Find all identifiers that are local declarations (let, const, var, function params)
+	// and not class/method names or properties
+	const declaredLocals = new Set();
+	const protectedNames = new Set();
+
+	for (let i = 0; i < tokens.length; i++) {
+		const tok = tokens[i];
+		if (tok.type !== "kw" && tok.type !== "op") continue;
+
+		// class Name or extends Name — protect
+		if (tok.value === "class" || tok.value === "extends") {
+			const next = nextNonWs(tokens, i);
+			if (next && next.tok.type === "ident") {
+				protectedNames.add(next.tok.value);
+			}
+		}
+
+		// Property access: .name — protect
+		if (tok.value === ".") {
+			const next = nextNonWs(tokens, i);
+			if (next && next.tok.type === "ident") {
+				protectedNames.add(next.tok.value);
+			}
+		}
+
+		// Method definitions: name( in class body — protect
+		// We approximate: ident followed by ( at class level
+
+		// let/const/var name — local
+		if (tok.value === "let" || tok.value === "const" || tok.value === "var") {
+			const next = nextNonWs(tokens, i);
+			if (next && next.tok.type === "ident") {
+				declaredLocals.add(next.tok.value);
+			}
+			// Destructuring: let [a, b] or let {a, b}
+			if (next && (next.tok.value === "[" || next.tok.value === "{")) {
+				collectDestructured(tokens, next.idx, declaredLocals);
+			}
+		}
+
+		// Function params
+		if (tok.value === "(") {
+			const prev = findPrevNonWsToken(tokens, i);
+			if (
+				prev &&
+				(prev.value === "function" ||
+					prev.type === "ident" ||
+					prev.value === ")")
+			) {
+				collectParams(tokens, i, declaredLocals);
+			}
+		}
+
+		// Arrow function params handled by ( above
+	}
+
+	// Remove protected names from locals
+	for (const name of protectedNames) {
+		declaredLocals.delete(name);
+	}
+
+	// Also protect known globals and short names
+	const globals = new Set([
+		"console",
+		"document",
+		"window",
+		"Math",
+		"JSON",
+		"String",
+		"Number",
+		"Boolean",
+		"Array",
+		"Object",
+		"Error",
+		"Date",
+		"RegExp",
+		"Map",
+		"Set",
+		"Promise",
+		"Symbol",
+		"parseInt",
+		"parseFloat",
+		"isNaN",
+		"isFinite",
+		"undefined",
+		"null",
+		"NaN",
+		"Infinity",
+		"TextEncoder",
+		"TextDecoder",
+		"setTimeout",
+		"setInterval",
+		"clearTimeout",
+		"clearInterval",
+		"prototype",
+		"constructor",
+		"this",
+		"super",
+		"arguments",
+	]);
+	for (const name of globals) declaredLocals.delete(name);
+
+	// Don't mangle __ prefixed helpers or single-char names
+	for (const name of declaredLocals) {
+		if (name.startsWith("__") || name.length <= 1) {
+			declaredLocals.delete(name);
+		}
+	}
+
+	// Assign short names
+	for (const name of declaredLocals) {
+		const short = nameGen();
+		renameMap.set(name, short);
+	}
+
+	// Apply renaming
+	let out = "";
+	for (let i = 0; i < tokens.length; i++) {
+		const tok = tokens[i];
+		if (tok.type === "ident" && renameMap.has(tok.value)) {
+			// Don't rename if preceded by dot (property access)
+			const prev = findPrevNonWsToken(tokens, i);
+			if (prev && prev.value === ".") {
+				out += tok.value;
+			} else {
+				out += renameMap.get(tok.value);
+			}
+		} else {
+			out += tok.value;
+		}
+	}
+
+	return out;
+}
+
+function nextNonWs(tokens, index) {
+	for (let i = index + 1; i < tokens.length; i++) {
+		if (tokens[i].type !== "ws" && tokens[i].type !== "comment") {
+			return { tok: tokens[i], idx: i };
+		}
+	}
+	return null;
+}
+
+function findPrevNonWsToken(tokens, index) {
+	for (let i = index - 1; i >= 0; i--) {
+		if (tokens[i].type !== "ws" && tokens[i].type !== "comment")
+			return tokens[i];
+	}
+	return null;
+}
+
+function collectDestructured(tokens, startIdx, locals) {
+	const open = tokens[startIdx].value;
+	const close = open === "[" ? "]" : "}";
+	let depth = 1;
+	for (let i = startIdx + 1; i < tokens.length && depth > 0; i++) {
+		const t = tokens[i];
+		if (t.value === open) depth++;
+		else if (t.value === close) depth--;
+		else if (depth === 1 && t.type === "ident") {
+			locals.add(t.value);
+		}
+	}
+}
+
+function collectParams(tokens, openIdx, locals) {
+	let depth = 1;
+	for (let i = openIdx + 1; i < tokens.length && depth > 0; i++) {
+		const t = tokens[i];
+		if (t.value === "(") depth++;
+		else if (t.value === ")") depth--;
+		else if (depth === 1 && t.type === "ident") {
+			// Check if not a default value expression
+			const prev = findPrevNonWsToken(tokens, i);
+			if (
+				!prev ||
+				prev.value === "(" ||
+				prev.value === "," ||
+				prev.value === "[" ||
+				prev.value === "{"
+			) {
+				locals.add(t.value);
+			}
+		}
+	}
+}
+
+function* shortNameGenerator() {
+	const chars = "abcdefghijklmnopqrstuvwxyz";
+	let index = 0;
+	while (true) {
+		if (index < 26) {
+			yield chars[index++];
+		} else {
+			const base = index - 26;
+			yield chars[base % 26] + (Math.floor(base / 26) + 1);
+			index++;
+		}
+	}
+}
+
+// Make it callable (not an iterator interface)
+function shortNameGeneratorFn() {
+	const gen = shortNameGenerator();
+	return () => gen.next().value;
+}
+
+// ── Stage 4: Literal folding ─────────────────────────────────
+
+function foldLiterals(code) {
+	// Fold patterns like: <number> <op> <number> where both are literal constants
+	return code.replace(
+		/\b(\d+(?:\.\d+)?)\s*(\+|-|\*|\/|%)\s*(\d+(?:\.\d+)?)\b/g,
+		(match, left, op, right) => {
+			const l = Number(left);
+			const r = Number(right);
+			let result;
+			switch (op) {
+				case "+":
+					result = l + r;
+					break;
+				case "-":
+					result = l - r;
+					break;
+				case "*":
+					result = l * r;
+					break;
+				case "/":
+					result = r !== 0 ? l / r : null;
+					break;
+				case "%":
+					result = r !== 0 ? l % r : null;
+					break;
+				default:
+					return match;
+			}
+			if (result === null || !Number.isFinite(result)) return match;
+			// If result is integer, emit without decimal point
+			const s = Number.isInteger(result) ? String(result) : String(result);
+			return s;
+		},
+	);
+}
+
+// ── Public API ───────────────────────────────────────────────
+
+export function minify(code, options = {}) {
+	const { mangle: doMangle = false } = options;
+
+	// Stage 1 + 2: compress (strips comments, whitespace, compresses tokens)
+	let result = compress(code);
+
+	// Stage 4: literal folding
+	result = foldLiterals(result);
+
+	// Stage 3: mangling (optional)
+	if (doMangle) {
+		result = mangle(result);
+		// Re-compress after mangling to clean up any leftover whitespace
+		result = compress(result);
+	}
+
+	return result;
+}