Introduction to Infosphere Streams 2.0 features: Using collections

You will also learn how to access the collection using the underlying c++++ Streams operator.

This article describes the 3 collection types currently available for Infosphere Streams 2.0 http://www.aliyun.com/zixun/aggregation/7155.html > developers. Set and mapped composite types are not only available for Streams version 2.0, but also provide flexibility for nesting of underlying and composite types. This article demonstrates the powerful features of the new type using both the Streams processing language and some examples of C + +.

Prerequisites

This article is written for Streams component developers and application programmers who have Streams programming language skills and C + + skills. You can use this article as a reference, or you can examine and run an example to demonstrate the technology described. To perform the sample, you should have a general understanding of Streams programming.

Running these samples requires a computer running Red Hat Enterprise linux® with Infosphere Streams V2.0 or later installed.

Understanding Collections

Streams Processing Language (SPL) supports both the underlying type (int and string, and so on), and also supports composite types (sets and mappings) that can be used to represent streaming data in an SPL application. A set of compound types, called collections, can be used to represent a complex arrangement of stream data, and support the use of the built-in SPL function to access and manipulate the contents of the collection.

In earlier versions of Infosphere Streams, the spade language supported a set of restricted composite types, including matrices and lists, that could be used to manage the underlying type groupings. Streams 2.0 gave up the matrix type. The list of supported collections now contains:

The collection has been further expanded to represent complex nesting of compound types, including lists of lists, which adds flexibility to the data managed by the collection.

Figure 1. Streams Foundation and composite types

Restricted and unrestricted types

For basic types such as strings, SPL also implements a collection type with a finite length. The size of any finite length collection needs to be defined using a compile-time constant. You can also declare a finite set of infinite types, such as a finite list of rstring. From a developer's perspective, the difference between finite and infinite types is small. However, there are some compile-time optimization advantages when using finite types.

Note that SQL restricts any collection to up to 231-1 elements, whether restricted or unrestricted.

The following sections describe the composite type in more detail.

Understanding List Collections