1.1.2 The DDS Object

The DDS is used to store information about the organization of the data set and its variables. It contains information about the type and shape of variables. While the DDS is similar to the DAS in that it is used to store information about the data set, it is used quite differently by both the client and server components of DODS. The DAS is a stand-alone object and is used solely for the purpose of storing attributes of variables and the dataset. The DDS, however, stores type information about a data set's variables by storing actual instances of those variables.

The DODS data access protocol variable objects have methods that can be used to read values from a data set or transfer the variable's value over the network. This makes it convenient to use the DDS object itself to hold data, and on the server side, the DDS object is used by both the DDS filter program and the data filter program. (See Section 4.1 for more information about the structure of the DODS server.)  

• A String object used to store the name of the dataset to which this DDS refers.
• A singly-linked list to store the variables in the data set. Each variable is stored in an instance of one of the BaseType class's descendants, and can be a simple or compound data type.
• A "printer" method to create a textual representation of the DDS. Among other uses, this is used to transmit the DDS from server to client.
• A parser and scanner to read a text DDS and convert it into its C++form. This is used to read DDS information from a disk, and also to receive it over the network.
• A singly-linked list of parsed constraint expression clauses. The clauses wait here to be evaluated.
• A constraint expression parser and scanner to extract constraint expression clauses from the input constraint expression. This component is responsible for creating the list of constraint expression clauses.

The DDS object provides methods to access and operate each of these components. The two lists can be traversed with a g++ Pix object.

The DDS is `lexically scoped' so that two Structure variables may have components with the same names; each component will be referred to using the Structure name and the dot operator. So for example, if a DDS called ralph contains two structures, vitals and new_hip, and each structure contains a variable called age, you can differentiate the two by referring to one as ralph.vitals.age and the other as ralph.new_hip.age.