Cole-Greer commented on code in PR #3153:
URL: https://github.com/apache/tinkerpop/pull/3153#discussion_r2226308237
##########
gremlin-core/src/main/java/org/apache/tinkerpop/gremlin/util/NumberHelper.java:
##########
@@ -721,6 +723,86 @@ public static Number coerceTo(final Number a, final
Class<? extends Number> claz
return a;
}
+ /**
+ * Casts the given number to the specified numeric type if it can fit into
it.
+ * Otherwise, throw.
+ *
+ * @param a the number to be cast
+ * @param numberToken the number token denoting the desired type to cast
+ * @return the number cast to the specified type
+ * @throws IllegalArgumentException if the specified numeric type is
unsupported
+ * @throws ArithmeticException if the number overflows
+ */
+ public static Number castTo(final Number a, final N numberToken) {
+ Class<?> clazz = numberToken.getType();
+ if (a.getClass().equals(clazz)) {
+ return a;
+ } else if (clazz.equals(Integer.class)) {
+ Long val = getLong(a, numberToken);
+ if (val >= Integer.MIN_VALUE && val <= Integer.MAX_VALUE) {
+ return a.intValue();
+ }
+ } else if (clazz.equals(Long.class)) {
+ return getLong(a, numberToken);
+ } else if (clazz.equals(Float.class)) {
+ // BigDecimal to double will overflow into Infinity, we want to
throw instead of passing through
+ if (!a.getClass().equals(BigDecimal.class) &&
+ (Double.isInfinite(a.doubleValue()) ||
Double.isNaN(a.doubleValue()))) {
+ return a.floatValue();
+ }
+ if (a.doubleValue() >= -Float.MAX_VALUE && a.doubleValue() <=
Float.MAX_VALUE) {
+ return a.floatValue();
+ }
+ } else if (clazz.equals(Double.class)) {
+ // BigDecimal to double will overflow into Infinity, we want to
throw instead of passing through
+ if (!a.getClass().equals(BigDecimal.class) &&
+ (Double.isInfinite(a.doubleValue()) ||
Double.isNaN(a.doubleValue()))) {
+ return a.doubleValue();
+ }
+ if (!Double.isInfinite(a.doubleValue())) {
+ // float losses precision, use string intermediate
+ return a.getClass().equals(Float.class) ?
Double.parseDouble(a.toString()) : a.doubleValue();
+ }
+ } else if (clazz.equals(Byte.class)) {
+ Long val = getLong(a, numberToken);
+ if (val >= Byte.MIN_VALUE && val <= Byte.MAX_VALUE) {
+ return a.byteValue();
+ }
+ } else if (clazz.equals(Short.class)) {
+ Long val = getLong(a, numberToken);
+ if (val >= Short.MIN_VALUE && val <= Short.MAX_VALUE) {
+ return a.shortValue();
+ }
+ } else if (clazz.equals(BigInteger.class)) {
+ return NumberHelper.bigIntegerValue(a);
+ } else if (clazz.equals(BigDecimal.class)) {
+ // float losses precision, use string intermediate
+ return a.getClass().equals(Float.class) ? new
BigDecimal(a.toString()) : NumberHelper.bigDecimalValue(a);
+ } else {
+ throw new IllegalArgumentException("Unsupported number type token:
" + numberToken);
+ }
+
+ throw new ArithmeticException(String.format("Can't convert number of
type %s to %s due to overflow.",
+ a.getClass().getSimpleName(), numberToken));
+ }
+
+ private static Long getLong(final Number num, final N numberToken) {
+ // Explicitly throw when converting floating point infinity and NaN to
whole numbers
+ if (Double.isNaN(num.doubleValue())) {
+ throw new ArithmeticException(String.format("Can't convert NaN to
%s.", numberToken));
+ }
+ if (Double.isInfinite(num.doubleValue())) {
+ throw new ArithmeticException(String.format("Can't convert
floating point infinity to %s.", numberToken));
+ }
+ try {
+ return num.getClass().equals(BigInteger.class) ? ((BigInteger)
num).longValueExact() : num.longValue();
Review Comment:
We should also use
[BigDecimal.longValueExact()](https://docs.oracle.com/javase/8/docs/api/java/math/BigDecimal.html#longValueExact--)
as `BigDecimal.longValue()` may result in some big numbers appearing small:
```
new BigDecimal(Long.MAX_VALUE+"1").longValue()
==>-9
gremlin> g.inject(new BigDecimal(Long.MAX_VALUE+"1")).asNumber(N.nshort)
==>-9
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
