add long.__float__ attribute, check overflow in PyLong_AsDouble

src/runtime/long.cpp

@@ -338,7 +338,15 @@ extern "C" long PyLong_AsLongAndOverflow(Box* vv, int* overflow) noexcept {
 extern "C" double PyLong_AsDouble(PyObject* vv) noexcept {
     RELEASE_ASSERT(PyLong_Check(vv), "");
     BoxedLong* l = static_cast<BoxedLong*>(vv);
-    return mpz_get_d(l->n);
+
+    double result = mpz_get_d(l->n);
+
+    if (std::isinf(result)) {
+        PyErr_SetString(PyExc_OverflowError, "long int too large to convert to float");
+        return -1;
+    }
+
+    return result;
 }
 
 /* Convert the long to a string object with given base,
@@ -688,6 +696,19 @@ Box* longInt(Box* v) {
         return new BoxedInt(n);
 }
 
+Box* longFloat(BoxedLong* v) {
+    if (!isSubclass(v->cls, long_cls))
+        raiseExcHelper(TypeError, "descriptor '__float__' requires a 'long' object but received a '%s'",
+                       getTypeName(v));
+
+    double result = PyLong_AsDouble(v);
+
+    if (result == -1)
+        checkAndThrowCAPIException();
+
+    return new BoxedFloat(result);
+}
+
 Box* longRepr(BoxedLong* v) {
     if (!isSubclass(v->cls, long_cls))
         raiseExcHelper(TypeError, "descriptor '__repr__' requires a 'long' object but received a '%s'", getTypeName(v));
@@ -1388,6 +1409,7 @@ void setupLong() {
     long_cls->giveAttr("__rshift__", new BoxedFunction(boxRTFunction((void*)longRshift, UNKNOWN, 2)));
 
     long_cls->giveAttr("__int__", new BoxedFunction(boxRTFunction((void*)longInt, UNKNOWN, 1)));
+    long_cls->giveAttr("__float__", new BoxedFunction(boxRTFunction((void*)longFloat, UNKNOWN, 1)));
     long_cls->giveAttr("__repr__", new BoxedFunction(boxRTFunction((void*)longRepr, STR, 1)));
     long_cls->giveAttr("__str__", new BoxedFunction(boxRTFunction((void*)longStr, STR, 1)));
     long_cls->giveAttr("__hex__", new BoxedFunction(boxRTFunction((void*)longHex, STR, 1)));

test/tests/float.py

@@ -74,3 +74,17 @@ try:
     0.0 ** (-1.0)
 except ZeroDivisionError as e:
     print e.message
+
+print float(1l)
+print float(0l)
+print float(-1l)
+print (1l).__float__()
+l = 1024 << 1024
+try:
+    float(l)
+except OverflowError as e:
+    print e.message
+try:
+    float(-l)
+except OverflowError as e:
+    print e.message