Fix model hierarchy backward compatibility

[fjtirado]

Provide a default asNumber(Class targerNumberClass) implementation

[Copilot]

Pull request overview

This PR restores backward compatibility in the AbstractWorkflowModel hierarchy by providing a default implementation for number subclass conversions, reducing the need for each concrete model to duplicate asNumber(Class<N>) logic.

Changes:

• Add shared asSubclass(Number, Class<N>) conversion helper to AbstractWorkflowModel.
• Provide a default protected <N extends Number> Optional<N> asNumber(Class<N>) implementation in AbstractWorkflowModel.
• Simplify JavaModel by delegating number subclass conversion to the shared helper and removing now-unneeded imports/logic.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
impl/core/src/main/java/io/serverlessworkflow/impl/AbstractWorkflowModel.java	Introduces shared number-conversion helper and default asNumber(Class<N>) implementation for model subclasses.
experimental/model/src/main/java/io/serverlessworkflow/impl/model/func/JavaModel.java	Removes duplicated number conversion logic and delegates to the new shared helper.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

asSubclass converts to BigDecimal via BigDecimal.valueOf(num.doubleValue()), which (a) can throw NumberFormatException for NaN/Infinity inputs and (b) loses precision for large integral values (e.g., Long > 2^53). Consider using type-specific exact conversions (e.g., BigDecimal.valueOf(long) for integral types, new BigDecimal((BigInteger) num) for BigInteger, BigDecimal.valueOf(double) only for floating types) and returning Optional.empty() when conversion is not representable (including non-finite doubles).

Suggested change

	return Optional.of(targetNumberClass.cast(BigDecimal.valueOf(num.doubleValue())));
	BigDecimal result;
	if (num instanceof BigDecimal) {
	result = (BigDecimal) num;
	} else if (num instanceof BigInteger) {
	result = new BigDecimal((BigInteger) num);
	} else if (num instanceof Byte
	|| num instanceof Short
	|| num instanceof Integer
	|| num instanceof Long) {
	result = BigDecimal.valueOf(num.longValue());
	} else if (num instanceof Float || num instanceof Double) {
	double d = num.doubleValue();
	if (Double.isNaN(d) || Double.isInfinite(d)) {
	return Optional.empty();
	}
	result = BigDecimal.valueOf(d);
	} else {
	result = new BigDecimal(num.toString());
	}
	return Optional.of(targetNumberClass.cast(result));

[Copilot]

Pull request overview

Copilot reviewed 4 out of 4 changed files in this pull request and generated 2 comments.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

This change fixes exit-code propagation for ProcessReturnType.ALL and CODE, but there’s no regression test covering the failure case when return: all is used (i.e., ensuring ProcessResult.code is non-zero for a failing command). Adding a test workflow that fails with return: all would help prevent this from regressing again.

[Copilot]

asSubclass converts to BigDecimal via BigDecimal.valueOf(num.doubleValue()) and to BigInteger via BigInteger.valueOf(num.longValue()), which is lossy for large integral values and for BigInteger/BigDecimal sources outside long/double precision. Since this helper is now shared across models as the default number-conversion path, consider using non-lossy conversions where possible (e.g., BigInteger -> BigDecimal via constructor, BigDecimal -> BigInteger via toBigInteger, and avoiding doubleValue() for integral inputs).