syntax: accept the empty for () condition (align up to cstage)

cstage parse.c:1027 accepts `for ()` — N_FOR with a nil cond, no
for-else arm, no test emitted (an unconditional loop; a deliberate
Go-flavored ww divergence from Hare, original to the toolchain
import). The wwstage parser had no RPAREN early-out and rejected it.
Selfhost checker and cgen already nil-guard n.cond, so the parser arm
is the whole gap. Graduates the held e2e row (pin 1731/1157/3462).

Adjacent asymmetries left filed, not fixed here (rule 11): the brace
form `for { }`, the empty middle clause of the 3-clause form, and
the REVERSE case (wwstage accepts 2-clause `for (cond; post)` which
cstage rejects).

lib/ww/syntax/stmt.ww

@@ -141,11 +141,22 @@ fn parsefor(p: *parser) *node = {
 	advance(p);	// past `for`
 	expecttok(p, tkind.TK_LPAREN, "expected '(' after for");
 
-	// Four forms (matching C parser):
+	// Five forms (matching C parser):
+	//   for ()                         — empty cond, unconditional loop
 	//   for (cond)                     — only cond
 	//   for (init; cond; post)         — C-style 3-clause
 	//   for (let x .. expr)            — Hare-style range, single binding
 	//   for (let (a, b) .. expr)       — range with tuple destructure
+
+	// Empty condition: cond stays nil (no test emitted downstream).
+	// Mirrors cmd/wcc/parse.c:1027 — returns WITHOUT the for-else arm,
+	// exactly as the cstage empty form does.
+	if (p.curkind == tkind.TK_RPAREN) {
+		advance(p);
+		let n = newnode(nkind.N_FOR, pf, pl, pc);
+		n.body = parseblock(p);
+		return n;
+	};
 	// Range and 3-clause both lead with `let`, so we commit to consuming
 	// `let` then disambiguate by looking at what follows.
 	if (p.curkind == tkind.TK_LET) {