Package spark.data.io

The Spark IO system contains classes for persisting and serializing Spark objects.

See:
Description

Interface Summary

SparkReader	This defines the interface for classes that can read Spark objects from some external form.
SparkWriter	Classes that implement this interface can be used to write Spark values out to some external form.
TypeFactory	Spark IO is used to create external forms of spark values.

Class Summary

JSONReader	This is a concrete implementation of SparkReader that will read text sources in the JSON format.
JSONWriter	This is a concrete implementation of SparkWriter that will write Spark data out in the JSON format.
SONReader	This is a concrete implementation of SparkReader that will read text sources in SON formation.
SONWriter	This is a concrete implementation of SparkWriter.
TypeFactoryReflect	This type factory uses reflection to create SOs that are subclasses of SOReflect.
TypeFactorySimple	This TypeFactory will always use SObj and SArray for its objects.

Package spark.data.io Description

The Spark IO system contains classes for persisting and serializing Spark objects. At the heart of the system are the interfaces SparkReader and SparkWriter. These define the standard mechanisms for reading and writing Spark objects. There are then a variety of subclasses of these two interfaces. These subclasses implement various formats in which Spark objects can be read and written.

The Spark IO maintains a relationship between the external form of an object and the class of object that it references. There are various ways in which this works. The key to this behavior is the TypeFactory interface. This defines how object classes are integrated with the external format.

The most important classes are:

• SparkReader This is an interface that defines how objects are read from text sources into Spark data.
• SparkWriter This is an interface that defines how Spark objects can be written to various text destinations.
• JSONReader This implementation of SparkReader will read a JSON specification from a variety of text sources and will return an SV object as its result. This is the primary mechanism for reading JSON.
• JSONWriter This implementation of SparkWriter will take an SV and will write its contents to a number of text outputs. The output is written in JSON format.
• SONReader This implementation of SparkReader will read a SON specification from a variety of text sources and will return an SV object as its result. This is the primary mechanism for reading SON. SON is virtually identical to JSON except that SON object specifications can be preceded by a class name. This greatly simplifies the ability to create textual representations of objects of specific classes.
• SONWriter This implementation of SparkWriter will take an SV and will write its contents to a number of text outputs. The output is written in SON format.