KoreanNumberFilter messes up offsets [LUCENE-10361]

It is a tokenfilter, tries to change offsets, so of course TestRandomChains finds bugs in it:
```
NOTE: reproduce with: gradlew test --tests TestRandomChains.testRandomChains -Dtests.seed=12BC606B774693E4 -Dtests.nightly=true -Dtests.slow=true -Dtests.locale=om-Latn-ET -Dtests.timezone=Australia/Yancowinna -Dtests.asserts=true -Dtests.file.encoding=UTF-8
```

```
org.apache.lucene.analysis.tests.TestRandomChains > test suite's output saved to /home/rmuir/workspace/lucene/lucene/analysis/integration.tests/build/test-results/test_16/outputs/OUTPUT-org.apache.lucene.analysis.tests.TestRandomChains.txt, copied below:
2> stage 0: 뱅<[0-1] +1> Ƒ<[1-2] +1> ė<[3-4] +1> 履<[6-7] +1> jEqyzUT<[8-15] +1>
2> stage 1: 000000<[0-1] +1> Ƒ<[1-2] +1> ė<[3-4] +1> 000000<[6-7] +1> 154300<[8-15] +1> 454300<[8-15] +0>
2> last stage: 0<[0-1] +1> Ƒ<[1-2] +1> ė<[3-4] +1> 000000<[6-7] +1> 454300<[8-15] +0>
2> TEST FAIL: useCharFilter=false text='\ubc45\u0191(\u0117\ud8ad\udf0a\uf9df jEqyzUT '
2> Exception from random analyzer:
2> charfilters=
2> org.apache.lucene.analysis.cjk.CJKWidthCharFilter(java.io.StringReader@17af5384)
2> org.apache.lucene.analysis.charfilter.MappingCharFilter(org.apache.lucene.analysis.charfilter.NormalizeCharMap@33e5bdbb, org.apache.lucene.analysis.cjk.CJKWidthCharFilter@1aafd271)
2> tokenizer=
2> org.apache.lucene.analysis.icu.segmentation.ICUTokenizer(org.apache.lucene.analysis.icu.segmentation.DefaultICUTokenizerConfig@4e6f4690)
2> filters=
2> Conditional:org.apache.lucene.analysis.phonetic.DaitchMokotoffSoundexFilter(OneTimeWrapper@34215eb7 term=,bytes=[],startOffset=0,endOffset=0,positionIncrement=1,positionLength=1,type=word,termFrequency=1,script=Common, false)
2> org.apache.lucene.analysis.ko.KoreanNumberFilter(ValidatingTokenFilter@7b4a2a5b term=,bytes=[],startOffset=0,endOffset=0,positionIncrement=1,positionLength=1,type=word,termFrequency=1,script=Common,keyword=false)
> java.lang.IllegalStateException: last stage: inconsistent startOffset at pos=3: 6 vs 8; token=454300
> at __randomizedtesting.SeedInfo.seed([12BC606B774693E4:2F5D490A30548E24]:0)
> at org.apache.lucene.test_framework@10.0.0-SNAPSHOT/org.apache.lucene.tests.analysis.ValidatingTokenFilter.incrementToken(ValidatingTokenFilter.java:138)
> at org.apache.lucene.test_framework@10.0.0-SNAPSHOT/org.apache.lucene.tests.analysis.BaseTokenStreamTestCase.checkAnalysisConsistency(BaseTokenStreamTestCase.java:1130)
> at org.apache.lucene.test_framework@10.0.0-SNAPSHOT/org.apache.lucene.tests.analysis.BaseTokenStreamTestCase.checkRandomData(BaseTokenStreamTestCase.java:1028)
> at org.apache.lucene.test_framework@10.0.0-SNAPSHOT/org.apache.lucene.tests.analysis.BaseTokenStreamTestCase.checkRandomData(BaseTokenStreamTestCase.java:922)
> at org.apache.lucene.analysis.tests@10.0.0-SNAPSHOT/org.apache.lucene.analysis.tests.TestRandomChains.testRandomChains(TestRandomChains.java:915)
```

---
Migrated from [LUCENE-10361](https://issues.apache.org/jira/browse/LUCENE-10361) by Robert Muir (@rmuir)

Start by running the supplied Gradle command for TestRandomChains.testRandomChains with the recorded seed, then inspect KoreanNumberFilter and TestRandomChains.java around line 915. Use the failure output and ValidatingTokenFilter to trace the inconsistent offset; done means the reproduced random chain no longer reports inconsistent startOffset values.

Written by the indexing model from the issue text.