Return to Data Taxonomy

In programming, an Object represents a real-world object such as a car, an employee or a task. Ultimately, as the real-world Object is defined in terms of software, it is translated to a Data Structure. The Object Data (see Figure 1 is categorized into Characteristic Data, Definition Data, Field Data and Operation Data.

Figure 1: Components of an Object

Return to Top

Many DIDO Platforms refer to Object Data as Smart Contracts. These are similar too, but not the same as classes in Object-Oriented Programming (OOP) languages such as C++, Java, C#, Python, etc. In many ways, the differences are lexical in nature (i.e., Smart Contract versus Contract, method versus function, etc. However, there are some important differences which have to do with the distributed nature of DIDOs and the immutability of the field data (i.e., view and pure designators on methods.

The following discussion has to do with the more generic concept of Object Data but does include discussions of how it relates to DIDO Platforms in general and Ethereum's Solidity more specifically.

• 01_char
• 03_def
• 05_field
• 07_opers

Return to Top

Every Object has characteristics that contextually describe the object. For example, the Object's name, if it implements and interfaces or if it is an extension of another object.

• Name is a textual name that uniquely identifies the object with the context of the Namespace. For example, there are multiple objects with the name of Tank. The first is in the namespace (i.e., context) of the oil and gas industry and refers to a fuel tank. Another Tank object has a military namespace and refers to a military vehicle, and another Tank object has a clothing namespace and refers to a “tank top” tee-shirt. All the rest of the data for the object is dependent on context (i.e., namespace).
• Implements is a way to describe which Interfaces this object provides an implementation for. For example, the ERC-20 interface used for many Cryptocurrencies. Java defines standardized collection interfaces for things like Lists, Set, Queue, Maps and Iterators¹⁾. C++ has similar interfaces to those describe for Java ²⁾. The exact way that interfaces are implemented differs from language to language, and can even differ from versions of the same language.
• Extends is a way to describe a hierarchical tree for an object (i.e., its Inheritance). For example, a graphics program ma have a base object called a GeographicShape'. Other objects within the graphics program such as Square, Circle and Triangle are said to extend the original definition of GeographicShape.

Return to Top

Smart Contracts are akin to classes in Object-Oriented Programming (OOP) languages, and they can have Inheritance and Interfaces. There are two very well-known examples of interfaces in use within Ethereum:

• EIP 20: ERC-20 Token Standard
• EIP 721: ERC-721 Non-Fungible Token Standard

However, all many of the Ethereum: Ethereum Improvement Proposals (EIPs) define interfaces.

Interfaces are similar to abstract contracts, but they are limited to what the contract’s ABI can represent. In other words, you could convert an ABI into an interface, or vice versa, and no information would be lost. According to the Solidity docs they have a few additional restrictions. For example, Doug Crescenzi, 13 June 2018, Accessed 3 Novemebr 2021, https://medium.com/upstate-interactive/solidity-how-to-know-when-to-use-abstract-contracts-vs-interfaces-874cab860c56 )) outlines the following restrictions:

• Interfaces cannot have any functions implemented
• Interfaces cannot inherit other contracts or interfaces (contracts can however inherit interfaces just as they would inherit other contracts)
• Interfaces cannot define a constructor
• Interfaces cannot define variables
• Interfaces cannot define structs
• Interfaces cannot define enums

Interfaces are expressed using the interface keyword. Here’s an example:

pragma solidity ^0.4.24;
 
interface token 
  { function totalSupply() public view returns (uint256);
    function balanceOf(address who) public view returns (uint256);
    function transfer(address to, uint256 value) public returns (bool);
  } // End Token

Return to Top

An Object has the ability to Define Data that can be accessed throughout the rest of the Object and depending on the public/private attribution of the definitions outside the Object Definition. The kinds of components that can be defined are:

• Exceptions, or faults, are abnormal or unprecedented events occurring during or after the execution of an operation which adversely affects the flow of the operation or the flow of the operations that use the object. Although must runtime systems define some standard exceptions such as Array-out-of-bounds, divide-by-zero, or Runtime, each Data object may define its own exceptions based upon the unique characteristics of the operations. For example, a Thermostat object might have a Too-Cold or Too-Hot exception. It is a runtime error of an undesired result or event affecting normal program flow.
• Structures or Objects are closely related. The easiest way to think of the difference between Structures and Objects is that structures usually only define the actual data, and object (i.e., Class) defines the data structures as well as operations that can be performed on that data. Object Data is a recursive definition. However in C++, the main difference between a structure and a class (i.e., object) is the assumption of public/private attribution. Structures assume that all components are by default public while classes (i.e., objects) assume that all components are private. For example, a Thermostat object might define a TemperatureReading structure that is uses to report results.
• Data Type are definitions of type information associated with the object. Data Type information and be numeric that include ranges, enumerations, scalars or collections. For example, the Thermostat might include a type describing the legal range of numeric values, it might define an enumeration type for Fahrenheit, Celsius or Calvin. It might also define a type for array of temperature readings.

Note: These components are not required by any Object and the names used to describe these components can be different depending on the programming language.

Return to Top

Field Data, or Element Data, are the values stored within the Object and in many ways represent the reason for the Object (i.e., Class in C++ or Java) to exist. In Procedural Language) Field Data can be thought of as a Variable with some variables having their values set at compile time (i.e., constants) and those during the execution of the procedure. Note: Methods or Operations in Objects are procedural in nature, generally represented by a block of code that executes from the top to the bottom, but can be circuitous within the block of code having logical expressions, looping and exceptions to control the flow within the block.

Constant Data is a quality of the data to not be modifiable during the execution of the procedure. There are two broad categories of constants: Named Constants and Literal COnstants. Note: Constants can be scoped to the particular procedure, the package of procedures or during the execution of the program.

Often the two terms: Attribute and Property are used interchangeable in Computer Science because we often use them to describe the fields (i.e., elements) of an Object in an Object-Oriented Programming (OOP). However, it is useful to understand the difference between the two grammatically:

Attribute Data is a quality or object that we attribute to someone or something. For example, intelligence is an attribute of a person. You can not go to the store and buy intelligence, it is an attribute of the individual. Another example would be a planet belonging to a solar system. Planets do not usually exist outside a solar system.

Property Data is a quality that exists without any attribution. Geometric shapes are properties in their own right. They can be used to independently describe many things. For example, a planet or a ball can be described as a sphere³⁾. A definition of the sphere exists without the attribution to a planet or a ball. Therefore, we say that a planet or a ball have spherical properties.

In Unified Modeling Language (UML), an Attribute and a Property both represent an structural association between two entities. Attributes are most often represented as Composition⁴⁾, while Property is best represented by Aggregation ⁵⁾.

According to Lenny Delligatti⁶⁾, “UML Attributes are Data Types (e.g., Strings, Integers, etc.) whereas Systems Modeling Language (SysML) properties are ValueTypes (e.g., can be assigned computed values).”

There are different ways that Field Data can be stored. In most programming paradigms, the Field Data is Mutable, meaning the values simply replaced or overwritten with new values. The Object Accessor Methods and responsible for making sure the new values are syntactically and semantically correct. Figure 2 graphically illustrates the concept of mutable. Each time the setter is called, the contents of the Field Data within the Data Object is changed. Note: There can be setters for Attribute and Property data. In the example, the first time the setter is called, the value of the attribute is set to 2, the next time it is 4.

Figure 2: Each time the setter is called, the value of the Filed Data is changed.

However, in DIDOs, most of the Field Data is considered immutable, meaning that updates and modifications to the Field Data are made while not destroying the previous data, but a new entry is made for the Field Data that points back to the original value. See Section 2.1.6 Immutable Data Objects and Figure See Figure 3. Note: The concepts of mutable and immutable are different from those built into C/C++, which describes immutable more like constants.

Figure 3: The Immutable Data Chain where the current value of a Field Data points to the previous value.

Return to Top

Any particular operation within an Object is generally Imperative in nature. This means that the Operations are composed of a set of statements. As the program steps through the Operation and executes each statement (i.e., instruction) the state of the Object is changed. Some statements change the values of the Field Data while others change the flow of execution through Control Flow operations (i.e., IF / ELSE IF / ELSE / ENDIF, or FOR LOOPS | WHILE LOOPS /, etc.). As the statements are executed, the Program Counter for the Operation is also modified. Note: Rarely all the statements contained within a single operation, but relies on calling and executing other Operations. Therefore, most Operations are also considered a Procedural Language. Figure 4 provides a graphic of the difference between Imperative and Procedural Programs.

Figure 4: The difference between Imperative and Procedural Programming.

One of the interesting phenomenons about many of the DIDO Platforms, is the way they treat Operation Data. Many of the Platforms not only distribute the Field Data using Distributed Ledger or Journal, they also distribute the Operation Data in the same way (in essence, treating the software just as any other data). When changes are made to the Operation Data, a Transaction is created to be pushed to the Node Network.

Return to Top

A Constructor is a special group of Operators that are called when an Object is first created. In most Object-Oriented Programming (OOP) languages (i.e., C++, Java and C#, e.g.), the Constructor has the same name as the Object. It is the initial stage in the Lifecycle of Object Data (see 05_lifecycle). For example, for an Object called Vehicle, the Constructor would also be called Vehicle. The constructor is responsible for the setup of the Object: the initialization of Field Data, the allocation of memory from the Heap. For example, if there are dynamic Field Data, then the memory needed for those Fields is usually allocated from the Heap. Although Constants are generally managed and allocated by the compiler and are from the Stack, Static Field Data can be allocated from the Heap and can be set during construction. For example, the values of Field Data initialized by using initialization parameters on the Constructor or read from initialization or setup files.

There is always a default Constructor that required no parameters, however, there can be other Constructors allowing for the passing of values to be used during initialization. For example, the minimum or maximum values used for the Field Data.

Regardless of the number of Constructors, there is always a Constructor that is called when an object is created. Often the calling of a Constructor is automatic and used the default Constructor, but the programmer can use any of the Constructors defined for the Object. See: https://www.tutorialspoint.com/solidity/solidity_constructors.htm

When Data Object is deployed in Ethereum, the following occurs:

The contract is initialized using the Constructor method named: constructor()

Constructor is a special function declared using constructor keyword. It is an optional function and is used to initialize state variables of a contract. Following are the key characteristics of a constructor.

• A contract can have only one constructor.
• A constructor code is executed once when a contract is created, and it is used to initialize contract state.
• After a constructor code executed, the final code is deployed to blockchain. This code include public functions and code reachable through public functions. Constructor code or any internal method used only by constructor are not included in final code
• A constructor can be either public or internal.
• A internal constructor marks the contract as abstract
• In case, no constructor is defined, a default constructor is present in the contract

• Note: See https://www.tutorialspoint.com/solidity/solidity_constructors.htm

Destructor is a special method called automatically during the destruction of an object. Actions executed in the destructor include the following:

• Recovering Heap space allocated during the lifetime of an object (see 05_lifecycle)
• Releasing Shared or Network Resources such as printers, faxes, scanners or outside Voice over Internet Protocol (VOIP) resources
• Releasing Resource Lock, for example, Web Servers, banking systems read versus update, travel reservations, etc
• Closing connections such as database, file or service
• Other housekeeping tasks

• Note: Often the first use of the Destructor, Recovering Heap memory, is automated in many modern systems (i.e., Java, .NET, Python, JavaScript, etc.) but the remaining reasons for a destructor remain in place, and it is up to the architect to determine what resources need to be freed upon the end of a Data Object (see 05_lifecycle).

When Data Object is destoryed in Ethereum, the following occurs:

• A contract can be destructed via the selfdestruct() method or the deprecated suicide() method

• Note: See https://www.tutorialspoint.com/solidity/solidity_constructors.htm

Return to Top

An Accessor Method is an Operation defined by Object-Oriented Programming (OOP) providing access to Field Data. There are generally two Accessor Methods defined: get and set. In some Object-Oriented (OO) paradigms, the Accessor methods are further defined as:

• Accessor Methods provide Read-Only access to the Field Data. Accessor Methods provide a way to obtain the state of an object's Field Data from outside the Object.
• Mutator Methods provide a Write-Only access to the Field Data. A Mutator not only does simple checking of the data being set such as type and bound checking; it can also check the semantics of the Field Data to ensure they are valid and consistent. For example, using a Mutex Method to deduct a payment, might also set the date of the payment at the same time even though they are separate Fields within the Object.

Note: There may or may not be any MUTual EXclusion (mutex) or Resource Lock associated with the Field Data.

Note: Often both the Accessor and the Mutator are simply referred to as the Accessors. “ Note: Both the Accessor and the Mutator are designated as public, but they can also be declares as private or protected. Those that are designated Private can only be used from within the Object. Public designations allow for the methods to be used from inside and outside the Object. Protected allows the internal Object Methods to access the methods, but also allows any Objects derived from the Object to access the methods.

Some benefits of using a Mutator Methods include:

• The prevention of data corruption by directly accessing the private Field Data of an Object
• The flexibility in modifying the internal representation of the Data Fields of an Object without breaking the Interface and consequently code that depends on the Interface.
• The ability to include additional processing logic such as validation of a value set, triggering of events, etc. during mutation of the Field Data or adding Data Logging of acccesses.
• The ability to add MUTual EXclusion (mutex) or Resource Lock for synchronizing multithreading or multiprocessor scenarios.
• The ability to define Accessor and Mutator methods in derived Data Objects from the base Data Object.

Return to Top

Ethereum's Solidity View can attribute Methods as View. methods ensure they do not modify the Data Object state.

A method can be declared as view. The following statements if present in the method are considered modifying the state and the compiler throws warning in such cases.

• Modifying state variables
• Emitting events
• Creating other contracts
• Using selfdestruct
• Sending via calls
• Calling any method which is not marked view or pure
• Using low-level calls
• Using inline assembly containing certain opcodes

Getter Method (i.e., Accessor Methods are by default view methods.

• Note: See: https://www.tutorialspoint.com/solidity/solidity_view_functions.htm

Ethereum's Solidity can attribute Methods as Pure.

Pure methods ensure that they not read or modify the state. A method can be declared as pure. The following statements if present in the method are considered reading the state and compiler will throw warning in such cases.

• Reading state variables.
• Accessing address(this).balance or <address>.balance.
• Accessing any of the special variable of block, tx, msg (msg.sig and msg.data can be read).
• Calling any method not marked pure.
• Using inline assembly that contains certain opcodes.

Pure methods can use the revert() and require() functions to revert potential state changes if an error occurs.

• Note: See: https://www.tutorialspoint.com/solidity/solidity_pure_functions.htm

Return to Top

An Abstract Method, or Virtual Method, are methods that are declared but have no underlying implementation (i.e., no body). The primary purpose of these methods is to form a base, template or guide for subsequent Data Objects derived from this Data Object. The derived Data Objects all have the same behavior defined by Abstract Methods of the base object. The original base Data Object is considered as Abstract or Virtual since without an implementation for these methods, the Data Object remains a “concept”.

Almost all Object-Oriented Programming (OOP) languages provide for the concept of Abstract or Virtual methods. In recent years, the concept of Virtual or Abstract Data Objects have lost favor to the use of Generics or Templates which can be instantiated for a particular type. For example, a double-linked-lst template, or a name-value pair template.

There are benefits of using Abstract Data Objects⁷⁾:

Figure 5: Relationship of Abstract Interfaces with realized Data Objects

Return to Top

Doug Crescenzi ⁸⁾ does an excellent job in summarizing the use of Abstract smart_contracts and Interfaces.

Ethereum allows for both Interfaces and for Abstract Contracts (i.e., Data Objects). An Abstract Smart Contract is any Smart Contract that has at least one method specified that does not have a corresponding body (implementation). This means that the Abstract Data Object acts like a Base Class, then the Derived Class MUST provide an implementation for the abstract method(s).

pragma solidity ^0.4.24;
 
contract Person 
{ function gender() public returns (bytes32);
}
 
contract Employee is Person 
{ function gender() public returns (bytes32) 
    { return "female"; }
}

Return to Top

[char]New Section -- review