Fix pow overflow (#784)

library/src/compute/calc.rs

@@ -108,7 +108,11 @@ pub fn pow(
     };
 
     let result = match (base, exponent.v) {
-        (Num::Int(a), Num::Int(b)) if b >= 0 => Num::Int(a.pow(b as u32)),
+        (Num::Int(a), Num::Int(b)) if b >= 0 => a
+            .checked_pow(b as u32)
+            .map(Num::Int)
+            .ok_or("the result is too large")
+            .at(args.span)?,
         (a, Num::Int(b)) => Num::Float(a.float().powi(b as i32)),
         (a, b) => Num::Float(a.float().powf(b.float())),
     };

tests/typ/compute/calc.typ

@@ -90,6 +90,10 @@
 // Error: 14-31 exponent is too large
 #calc.pow(2, 10000000000000000)
 
+---
+// Error: 10-25 the result is too large
+#calc.pow(2, 2147483647)
+
 ---
 // Error: 14-36 exponent may not be infinite, subnormal, or NaN
 #calc.pow(2, calc.pow(2.0, 10000.0))