Abstract: Pieces of an entity's data set can be related with each other using a canonical Uniform Resource Locator (URL). If a server returns pieces of an entity's data record to a client within multiple separate REST-based responses, the client can discern from the canonical URL that those pieces relate to the same entity. In response to each REST-based request from a client, a server returns, with the client-requested data, a canonical URL that uniquely identifies that data's entity. A client can receive the canonical URL with the data that the server returns. If the canonical URLs returned along with data items in separate requests match, then the client determines that those data items pertain to the same entity. If the client determines that separately received data items pertain to the same entity, then the client can merge those data items together into a unified record that the client stores locally.

Abstract: In accordance with an embodiment, described herein is a JSON schema generator that parses a JSON stream, determines a data type of each record of the JSON stream, and selects a corresponding record schema generation logic component based on the determined data type to build a record schema for that record. In building a record schema, each element in a record and each sub-element in one or more nested layers of the record can be examined, to determine their data types. A corresponding parser can be selected for each element and sub-element based on a determined data type to generate a schema node. The generated schema nodes can be joined together to create a record schema for the record. Each record schema can be dynamically merged together to create a schema for the JSON stream.

Abstract: In accordance with an embodiment, described herein is a JSON schema generator that parses a JSON stream, determines a data type of each record of the JSON stream, and selects a corresponding record schema generation logic component based on the determined data type to build a record schema for that record. In building a record schema, each element in a record and each sub-element in one or more nested layers of the record can be examined, to determine their data types. A corresponding parser can be selected for each element and sub-element based on a determined data type to generate a schema node. The generated schema nodes can be joined together to create a record schema for the record. Each record schema can be dynamically merged together to create a schema for the JSON stream.

Abstract: Pieces of an entity's data set can be related with each other using a canonical Uniform Resource Locator (URL). If a server returns pieces of an entity's data record to a client within multiple separate REST-based responses, the client can discern from the canonical URL that those pieces relate to the same entity. In response to each REST-based request from a client, a server returns, with the client-requested data, a canonical URL that uniquely identifies that data's entity. A client can receive the canonical URL with the data that the server returns. If the canonical URLs returned along with data items in separate requests match, then the client determines that those data items pertain to the same entity. If the client determines that separately received data items pertain to the same entity, then the client can merge those data items together into a unified record that the client stores locally.