Handling Large Integer IDs in Java and JavaScript: Preventing Precision Loss with Jackson Configuration

When sensitive data IDs are generated with a Snowflake‑style algorithm, the resulting values are stored as Java long (max 9223372036854775807). In browsers, especially Chrome, these large numbers cannot be displayed correctly because JavaScript's safe integer limit (2^53‑1) is much smaller.

The article shows screenshots of the discrepancy and points out that JavaScript integers lose precision for the higher bits of a Java long , leading to incorrect values on the front end.

To solve the problem, the back‑end can serialize Long values as strings so that the front‑end receives the exact identifier without precision loss. A Spring Boot solution is provided by customizing Jackson's object mapper.

@Configuration
public class Jackson2Customizer {
    public static final String DATE_FORMAT = "yyyy-MM-dd";
    public static final String DATE_TIME_FORMAT = "yyyy-MM-dd HH:mm:ss";

    @Bean
    public Jackson2ObjectMapperBuilderCustomizer jackson2ObjectMapperBuilderCustomizer() {
        return jacksonObjectMapperBuilder -> {
            // Fix front‑end JS precision issue
            jacksonObjectMapperBuilder.serializerByType(Long.class, ToStringSerializer.instance);
            jacksonObjectMapperBuilder.serializerByType(Long.TYPE, ToStringSerializer.instance);
            jacksonObjectMapperBuilder.simpleDateFormat(DATE_TIME_FORMAT);
        };
    }
}

Alternatively, individual fields can be annotated directly:

@JsonSerialize(using = ToStringSerializer.class)

By converting Long values to strings during JSON serialization, the front‑end receives the full identifier without truncation, eliminating the display issue.