Small/big int representation for starlark Int (#155)

diff --git a/starlark/int_test.go b/starlark/int_test.go
new file mode 100644
index 0000000..985bc38
--- /dev/null
+++ b/starlark/int_test.go
@@ -0,0 +1,72 @@
+// Copyright 2017 The Bazel Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package starlark
+
+import (
+	"fmt"
+	"math"
+	"math/big"
+	"testing"
+)
+
+// TestIntOpts exercises integer arithmetic, especially at the boundaries.
+func TestIntOpts(t *testing.T) {
+	f := MakeInt64
+	left, right := big.NewInt(math.MinInt32), big.NewInt(math.MaxInt32)
+
+	for i, test := range []struct {
+		val  Int
+		want string
+	}{
+		// Add
+		{f(math.MaxInt32).Add(f(1)), "80000000"},
+		{f(math.MinInt32).Add(f(-1)), "-80000001"},
+		// Mul
+		{f(math.MaxInt32).Mul(f(math.MaxInt32)), "3fffffff00000001"},
+		{f(math.MinInt32).Mul(f(math.MinInt32)), "4000000000000000"},
+		{f(math.MaxUint32).Mul(f(math.MaxUint32)), "fffffffe00000001"},
+		{f(math.MinInt32).Mul(f(-1)), "80000000"},
+		// Div
+		{f(math.MinInt32).Div(f(-1)), "80000000"},
+		{f(1 << 31).Div(f(2)), "40000000"},
+		// And
+		{f(math.MaxInt32).And(f(math.MaxInt32)), "7fffffff"},
+		{f(math.MinInt32).And(f(math.MinInt32)), "-80000000"},
+		{f(1 << 33).And(f(1 << 32)), "0"},
+		// Mod
+		{f(1 << 32).Mod(f(2)), "0"},
+		// Or
+		{f(1 << 32).Or(f(0)), "100000000"},
+		{f(math.MaxInt32).Or(f(0)), "7fffffff"},
+		{f(math.MaxUint32).Or(f(0)), "ffffffff"},
+		{f(math.MinInt32).Or(f(math.MinInt32)), "-80000000"},
+		// Xor
+		{f(math.MinInt32).Xor(f(-1)), "7fffffff"},
+		// Not
+		{f(math.MinInt32).Not(), "7fffffff"},
+		{f(math.MaxInt32).Not(), "-80000000"},
+		// Shift
+		{f(1).Lsh(31), "80000000"},
+		{f(1).Lsh(32), "100000000"},
+		{f(math.MaxInt32 + 1).Rsh(1), "40000000"},
+		{f(math.MinInt32 * 2).Rsh(1), "-80000000"},
+	} {
+		if got := fmt.Sprintf("%x", test.val); got != test.want {
+			t.Errorf("%d equals %s, want %s", i, got, test.want)
+		}
+		if test.val.small < math.MinInt32 || math.MaxInt32 < test.val.small {
+			t.Errorf("expected big, %d %s", i, test.val)
+		}
+		if test.val.big == nil {
+			continue
+		}
+		if test.val.small != 0 {
+			t.Errorf("expected 0 small, %d %s with %d", i, test.val, test.val.small)
+		}
+		if b := test.val.big; b.Cmp(left) >= 0 && b.Cmp(right) <= 0 {
+			t.Errorf("expected small, %d %s", i, test.val)
+		}
+	}
+}