`BodyKeySanitizer` applying to body with escaped unicode results in failure to match

Occurring on commit `b5320b85f1303024252651f5075fc45c9946f78d` of [revise-common-sanitizers](https://github.com/samvaity/azure-sdk-for-java/tree/revise-common-sanitizers) branch.

Go after `TextAnalyticsClientTest#recognizePiiEntitiesEmojiWithSkinToneModifier` to repro. This should be transferred to a unit test while integrating the fix.

[Existing recording](https://github.com/Azure/azure-sdk-assets/blob/java/textanalytics/azure-ai-textanalytics_fbd1568119/java/sdk/textanalytics/azure-ai-textanalytics/src/test/resources/session-records/TextAnalyticsClientTest.recognizePiiEntitiesEmojiWithSkinToneModifier%5B1%5D.json)

AFAICT it is [this LoC](https://github.com/Azure/azure-sdk-tools/blob/04333b37b738c9a0ca13ef480728fe09cde566eb/tools/test-proxy/Azure.Sdk.Tools.TestProxy/Sanitizers/BodyKeySanitizer.cs#L104) breaking the serialization. Then during matching the bytes don't line up.