[b84330808]

What is the purpose of the change?

Several places turn identifiers and config keys into other forms with toLowerCase()/toUpperCase(), which follow the JVM default locale. Under a Turkish default locale "I".toLowerCase() is a dotless "ı" and "i".toUpperCase() is a dotted "İ", so these keys get mangled. This PR uses Locale.ROOT in all of them, as EnvironmentConfiguration, FixedParamValue and SwitchLogLevel already do:

• AbstractConfig: property names derived from getters/setters. appendParameters() wrote ıothreads=8 instead of iothreads=8, and refresh() looked up dubbo.protocol.ıothreads, so dubbo.protocol.iothreads was silently ignored.
• AbstractConfigManager: dubbo.config.mode=override became OVERRİDE, so ConfigMode.valueOf threw and startup failed.
• ServiceConfig: protocol tri looked up TRİ_DUBBO_PORT_TO_BIND / TRİ_DUBBO_IP_TO_BIND; also the random-port map keys.
• StringUtils: camelToSplitName, snakeToSplitName, isSnakeCase, toURLKey, toOSStyleKey (dubbo.protocol.iothreads became env var DUBBO_PROTOCOL_İOTHREADS).
• DubboBeanDefinitionParser: setIothreads() mapped to ıothreads, so the iothreads XML attribute was ignored.
• Same pattern in ReflectUtils, MethodUtils, PojoUtils, JavaBeanSerializeUtil, ExtensionLoader and EnvironmentAdapter (bean property, setter and extension names).

Regression tests set the default locale to tr-TR (in AbstractConfigTest, ConfigManagerTest, ServiceConfigTest, StringUtilsTest, DubboBeanDefinitionParserTest, ReflectUtilsTest, MethodUtilsTest, PojoUtilsTest, JavaBeanSerializeUtilTest). Each test fails without its source change and passes with it. ExtensionLoader and EnvironmentAdapter have no dedicated test; they are the same one-line change.

I used Claude Code to write this change.

Checklist

• Make sure there is a GitHub_issue field for the change. (no separate issue; the description above covers it)
• 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.
• Make sure gitHub actions can pass.

[codecov-commenter]

Codecov Report

❌ Patch coverage is 94.28571% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 60.95%. Comparing base (843b98b) to head (b8a2927).
⚠️ Report is 3 commits behind head on 3.3.

Files with missing lines	Patch %	Lines
...apache/dubbo/common/extension/ExtensionLoader.java	66.66%	0 Missing and 1 partial ⚠️
.../java/org/apache/dubbo/common/utils/PojoUtils.java	75.00%	1 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff              @@
##                3.3   #16550      +/-   ##
============================================
+ Coverage     60.90%   60.95%   +0.05%     
- Complexity       15    11754   +11739     
============================================
  Files          1953     1953              
  Lines         89280    89282       +2     
  Branches      13476    13477       +1     
============================================
+ Hits          54376    54422      +46     
+ Misses        29323    29268      -55     
- Partials       5581     5592      +11     

Flag	Coverage Δ
integration-tests-java21	32.07% <60.00%> (-0.01%)	⬇️
integration-tests-java8	32.20% <60.00%> (+0.05%)	⬆️
samples-tests-java21	32.10% <57.14%> (-0.03%)	⬇️
samples-tests-java8	29.78% <57.14%> (+<0.01%)	⬆️
unit-tests-java11	59.20% <94.28%> (+0.09%)	⬆️
unit-tests-java17	58.68% <94.28%> (+0.07%)	⬆️
unit-tests-java21	58.68% <94.28%> (+0.05%)	⬆️
unit-tests-java25	58.65% <94.28%> (+0.04%)	⬆️
unit-tests-java8	59.24% <94.28%> (+0.08%)	⬆️

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.

[EarthChen]

The bug is real: under a Turkish-like default locale (tr, az), "Iothreads".substring(0, 1).toLowerCase() gives "ı", so keys such as iothreads are derived wrongly in appendParameters and refresh. The AbstractConfig changes and tests look correct, but the same root cause remains elsewhere, some of it worse:

1. AbstractConfigManager (~line 124): configModeStr.toUpperCase() turns strict into STRİCT, and ConfigMode.valueOf throws.
2. ServiceConfig (~lines 1173/1182/1187): protocol name triple becomes TRİPLE.
3. StringUtils (~lines 1005/1013/1056/1060): snakeToSplitName, isSnakeCase, toURLKey, toOSStyleKey.
4. DubboBeanDefinitionParser (~lines 333/398): substring(3, 4).toLowerCase(), the same pattern as this PR.

Please cover these in this PR, or state in the description that the scope is AbstractConfig only and open a follow-up issue. (Line numbers are from a static review on 3.3.)

[b84330808]

Thanks, good catch! Covered the four spots in b8a2927 (AbstractConfigManager, ServiceConfig, StringUtils, DubboBeanDefinitionParser), each with a tr-TR test. While grepping dubbo-common I also fixed the same getter/setter-name pattern in ReflectUtils, MethodUtils, PojoUtils, JavaBeanSerializeUtil, ExtensionLoader and EnvironmentAdapter.

Left out on purpose: AbortPolicyWithReport lowercases os.name before checking for "win" (Turkish Windows gives wındows), but that's not a key, so I'd keep it separate. There are also ~25 similar calls in other modules (rpc, remoting, registry, plugins). Happy to do a follow-up for those if you'd like.