fix: improve upstream cache management and add recovery test for empty events

[Aias00]

Make sure that:

• You have read the contribution guidelines.
• You submit test cases (unit or integration tests) that back your changes.
• Your local test passed ./mvnw clean install -Dmaven.javadoc.skip=true.

[Copilot]

Pull request overview

This PR improves upstream cache behavior when an empty upstream event is submitted, and aligns Upstream.hashCode() with the existing equals() semantics to prevent inconsistent collection behavior.

Changes:

• Fix Upstream.hashCode() to hash only (protocol, url) to match equals().
• Update UpstreamCacheManager.submit() empty-list handling to avoid leaving stale entries in UPSTREAM_MAP.
• Add/extend unit tests to cover recovery after an empty upstream submit and verify updated equality/hash behavior.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 1 comment.

File	Description
shenyu-loadbalancer/src/main/java/org/apache/shenyu/loadbalancer/entity/Upstream.java	Makes hashCode() consistent with equals() by removing weight from hashing.
shenyu-loadbalancer/src/main/java/org/apache/shenyu/loadbalancer/cache/UpstreamCacheManager.java	Changes empty-upstream submission handling to remove selector entries from UPSTREAM_MAP.
shenyu-loadbalancer/src/test/java/org/apache/shenyu/loadbalancer/entity/UpstreamTest.java	Expands tests to validate equality/hashCode consistency across different weights.
shenyu-loadbalancer/src/test/java/org/apache/shenyu/loadbalancer/cache/UpstreamCacheManagerTest.java	Adds a recovery test ensuring submit works after an empty upstream event.

---

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

[Copilot]

When upstreamList is empty, this branch removes upstreams based on UPSTREAM_MAP, but UPSTREAM_MAP is populated from task.getHealthyUpstreamListBySelectorId(...) (i.e., only healthy upstreams). Any entries that only exist in the task's unhealthyUpstream map will not be cleared, so a selector can retain stale unhealthy state after an empty event. Consider clearing the task state directly (e.g., task.triggerRemoveAll(selectorId)) in addition to (or instead of) iterating over existUpstreamList, and then remove the selectorId from UPSTREAM_MAP.

Suggested change

	List<Upstream> existUpstreamList = UPSTREAM_MAP.get(selectorId);
	if (Objects.nonNull(existUpstreamList)) {
	removeAllUpstreams(selectorId, existUpstreamList);
	}
	UPSTREAM_MAP.remove(selectorId);
	// Clear both cached upstreams and all health-check state for this selector.
	removeByKey(selectorId);