Support naming strategy between ES index mapping and Java object

[weweibuy]

Sometimes ES index mapping naming strategy is SNAKE_CASE like:
{ "mapping": { "properties": { "customer_id": { "type": "long" }, "order_no": { "type": "keyword" } } } }
but Java object is LOWER_CAMEL_CASE like:

`public class Order {

        private Long customerId;

        private String orderNo;
    }`

In spring-data-elasticsearch 3x, we can use @JsonNaming or replace implements of org.springframework.data.elasticsearch.core.EntityMapper with SNAKE_CASE naming strategy ObjectMapper to map or read Java object;

In future spring-data-elasticsearch , Can it support:

• config global naming strategy between ES index mapping and Java object
• customized config for each ES index, like @JsonNaming in spring-data-elasticsearch 3x

[sothawo]

Implemented with #1565, available since version 4.2.0.M3

[sothawo]

additional info: this is an application wide configuration currently. To have it configurable for individual entites would be something new.