Fix StringUtils.substringAfter to return empty string when the separator is absent

[ns3154]

What is the purpose of the change?

StringUtils.substringAfter(String, int) is documented to return the empty string when the
separator is not found ("If nothing is found, the empty string is returned."), but it currently
returns the original string instead:

substringAfter("a:b", ':')  -> "b"     // ok
substringAfter("abc", ':')  -> "abc"   // expected "", actually returns "abc"

The not-found branch was copy-pasted from the neighbouring substringBefore (whose contract is
to return the original string), which left substringAfter inconsistent with both its own Javadoc
and its sibling substringAfterLast (which correctly returns the empty string). It also diverges
from commons-lang3's substringAfter.

This PR makes substringAfter return EMPTY_STRING when the separator is absent, and adds a unit
test (testSubstringAfter) covering the documented behaviour. The new test fails before the fix
(expected: <> but was: <abc>) and passes after it; the full StringUtilsTest suite stays green.

Fixes #16349

Checklist

• Make sure there is a GitHub_issue field for the change.
• Write a pull request description that is detailed enough to understand what the pull request does, how, and why.
• Write necessary unit-test to verify your logic correction. If the new feature or significant change is committed, please remember to add sample in dubbo samples project.
• Make sure gitHub actions can pass. Why the workflow is failing and how to fix it?

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 60.83%. Comparing base (b75adde) to head (5a0b52b).

Additional details and impacted files

@@             Coverage Diff              @@
##                3.3   #16350      +/-   ##
============================================
  Coverage     60.83%   60.83%              
- Complexity       15    11754   +11739     
============================================
  Files          1953     1953              
  Lines         89208    89208              
  Branches      13458    13458              
============================================
+ Hits          54270    54272       +2     
+ Misses        29365    29362       -3     
- Partials       5573     5574       +1     

Flag	Coverage Δ
integration-tests-java21	32.16% <0.00%> (-0.01%)	⬇️
integration-tests-java8	32.23% <0.00%> (-0.02%)	⬇️
samples-tests-java21	32.11% <0.00%> (-0.06%)	⬇️
samples-tests-java8	29.83% <0.00%> (-0.01%)	⬇️
unit-tests-java11	59.05% <100.00%> (-0.02%)	⬇️
unit-tests-java17	58.56% <100.00%> (+0.01%)	⬆️
unit-tests-java21	58.54% <100.00%> (+0.02%)	⬆️
unit-tests-java25	58.53% <100.00%> (+0.01%)	⬆️
unit-tests-java8	59.08% <100.00%> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[zrlw]

both substringBefore and substringAfter are not used, maybe we should delete them instead of fixing them.

[zrlw]

@oxsean

[zrlw]

LGTM

[zrlw]

LGTM

[ns3154]

Thanks for the suggestion, @zrlw! That's a fair point — both methods have no internal callers today. I chose to fix rather than delete because StringUtils is a public class and these are public methods: removing them could silently break any downstream project that happens to use them. Fixing the documented contract felt like the safer, non-breaking option. Happy to defer to the team's preference if deletion is favoured, though!