[BUG] global SnakeCase naming strategy is not applied to Java records

[artpaym]

Description

When fastjson2 global writer naming strategy is configured with PropertyNamingStrategy.SnakeCase, serialization applies snake_case naming to normal Java class fields, but not to Java records.

This causes DTOs implemented as records to serialize camelCase fields into final JSON even though global snake_case output is configured.

Environment

• Verified versions: 2.0.60, 2.0.61, 2.0.62
• Java: 21

Reproduction

import com.alibaba.fastjson2.JSON;
import com.alibaba.fastjson2.JSONFactory;
import com.alibaba.fastjson2.PropertyNamingStrategy;

record RecordDto(String currencyCode) {}

class ClassDto {
    public String currencyCode = "840";
}

JSONFactory.getDefaultObjectWriterProvider()
        .setNamingStrategy(PropertyNamingStrategy.SnakeCase);

System.out.println(JSON.toJSONString(new RecordDto("840")));
System.out.println(JSON.toJSONString(new ClassDto()));

Expected Result

Both DTOs serialize using snake_case:

{"currency_code":"840"}
{"currency_code":"840"}

Actual Result

Only the normal class is snake-cased. The record remains camelCase:

{"currencyCode":"840"}
{"currency_code":"840"}

Workaround

Use a normal Java class instead of a record, or annotate each affected record component with @JSONField(name = "snake_case_name").

Additional Notes

Also verified that @JSONType(naming = PropertyNamingStrategy.SnakeCase) and JSONWriter.Feature.FieldBased did not change the record output in the local probe.

[1919chichi]

Hi, I've submitted a fix for this issue in #7657.

The root cause was in ObjectWriterCreator: when resolving field names for records, the code called method.getName() directly instead of BeanUtils.getterName(method, kotlin, namingStrategy), which bypassed both the global naming strategy and the @JSONType(naming = ...) annotation entirely.

The fix makes records go through the same naming pipeline as regular classes.

[artpaym]

Thanks! @1919chichi