Principles of Computer Programming

• Computer hardware changes frequently - from room-filling machines with punch cards and tapes to modern laptops and tablets - and will continue doing so.
• With these changes, the capabilities of computers increase rapidly (storage, speed, graphics, etc.)
• Computer programming languages also change
  • Better programming language theory leads to new programming techniques
  • Improved programming language implementations
  • New languages are created, old ones updated
• There are hundreds of programming languages, why?
  • Different tools for different jobs
    • Some languages are better suited for certain jobs
    • For example, Python is best for scripting, Javascript is best for web pages, MySQL is best for databases, etc.
  • Personal preference and popularity
• This class is about “principles” of computer programming
  • Common principles behind all languages will not change, even though hardware and languages do
  • How to organize and structure data
  • How to express logical conditions and relations
  • How to solve problems with programs

Programming Language Concepts

We begin by discussing three categories of languages manipulated by computers. We will be studying and writing programs in high-level languages, but understanding their differences and relationships to other languages¹ is of importance to become familiar with them.

• Machine language
  • Computers are made of electronic circuits
    • Circuits are components connected by wires
    • Some wires carry data - e.g. numbers
    • Some carry control signals - e.g. do an add or a subtract operation
  • Instructions are settings on these control signals
    • A setting is represented as a 0 or 1
    • A machine language instruction is a group of settings - For example: 1000100111011000
  • Most CPUs use one of two languages: x86 or ARM
• Assembly language

  • Easier way for humans to write machine-language instructions

  • Instead of 1s and 0s, it uses letters and “words” to represent an instruction.

    • Example x86 instruction:

    MOV BX, AX

    which makes a copy of data stored in a component called AX and places it in one called BX

  • Assembler: Translates assembly language instructions to machine language instructions

    • For example: MOV BX, AX translates into 1000100111011000
    • One assembly instruction = one machine-language instruction
    • x86 assembly produces x86 machine code

  • Computers can only execute the machine code

• High-level language
  • Hundreds including C#, C++, Java, Python, etc.
  • Most programs are written in a high-level language since:
    • More human-readable than assembly language
    • High-level concepts such as processing a collection of items are easier to write and understand
    • Takes less code since each statement might be translated into several assembly instructions
  • Compiler: Translates high-level language to machine code
    • Finds “spelling” errors but not problem-solving errors
    • Incorporates code libraries – commonly used pieces of code previously written such as Math.Sqrt(9)
    • Optimizes high-level instructions – your code may look very different after it has been optimized
    • Compiler is specific to both the source language and the target computer
  • Compile high-level instructions into machine code then execute since computers can only execute machine code

A more subtle difference exists between high-level languages. Some (like C) are compiled (as we discussed above), some (like Python) are interpreted, and some (like C#) are in an in-between called managed.

• Compiled vs. Interpreted languages
  • Not all high-level languages use a compiler - some use an interpreter
  • Interpreter: Lets a computer “execute” high-level code by translating one statement at a time to machine code
  • Advantage: Less waiting time before you can execute the program (no separate “compile” step)
  • Disadvantage: Program executes slower since you wait for the high-level statements to be translated then the program is executed
• Managed high-level languages (like C#)
  • Combine features of compiled and interpreted languages
  • Compiler translates high-level statements to intermediate language instructions, not machine code
    • Intermediate language: Looks like assembly language, but not specific to any CPU
  • run-time executes by interpreting the intermediate language instructions - translates one at a time to machine code
    • Faster since translation is partially done already (by compiler), only a simple “last step” is done when executing the program
  • Advantages of managed languages:
    • In a “non-managed” language, a compiled program only works on one OS + CPU combination (platform) because it is machine code
    • Managed-language programs can be reused on a different platform without recompiling - intermediate language is not machine code and not CPU-specific
    • Still need to write an intermediate language interpreter for each platform (so it produces the right machine code), but, for a non-managed language, you must write a compiler for each platform
    • Writing a compiler is more complicated and more work than writing an interpreter thus an interpreter is a quicker (and cheaper) way to put your language on different platforms
    • Intermediate-language interpreter is much faster than a high-level language interpreter, so programs execute faster than an “interpreted language” like Python
  • This still executes slower than a non-managed language (due to the interpreter), so performance-minded programmers use non-managed compiled languages (e.g. for video games)

Software Concepts

• Flow of execution in a program
  • Program receives input from some source, e.g. keyboard, mouse, data in files
  • Program uses input to make decisions
  • Program produces output for the outside world to see, e.g. by displaying images on screen, writing text to console, or saving data in files
• Program interfaces
  • GUI or Graphical User Interface: Input is from clicking mouse in visual elements on screen (buttons, menus, etc.), output is by drawing onto the screen
  • CLI or Command Line Interface: Input is from text typed into “command prompt” or “terminal window,” output is text printed at same terminal window
  • This class will use CLI because it is simple, portable, easy to work with – no need to learn how to draw images, just read and write text

Programming Concepts

Programming workflow

The workflow of the programmer will differ a bit depending on if the program is written in a compiled or an intprepreted programming language. From the distance, both looks like what is pictured in the the flowchart demonstrating roles and tasks of a programmer, beta tester and user in the creation of programs, but some differences remain:

• Compiled language workflow
  • Writing down specifications
  • Creating the source code
  • Running the compiler
  • Reading the compiler’s output, warning and error messages
  • Fixing compile errors, if necessary
  • Executing and testing the program
  • Debugging the program, if necessary
• Interpreted language workflow
  • Writing down specifications
  • Creating the source code
  • Executing the program in the interpreter
  • Reading the interpreter’s output, determining if there is a syntax (language) error or the program finished executing
  • Editing the program to fix syntax errors
  • Testing the program (once it can execute with no errors)
  • Debugging the program, if necessary

Interpreted languages have

• Advantages: Fewer steps between writing and executing, can be a faster cycle
• Disadvantages: All errors happen when you execute the program, no distinction between syntax errors (compile errors) and logic errors (bugs in executing program)

(Integrated) Development Environment

Programmers can either use a collection of tools to write, compile, debug and execute a program, or use an “all-in-one” solution called an Integrated Development Environment (IDE).

• The “Unix philosophy” states that a program should do only one task, and do it properly. For programmers, this means that
  • One program will be needed to edit the source code, a text editor (it can be Geany, notepad, kwrite, emacs, sublime text, vi, etc.),
  • One program will be needed to compile the source code, a compiler (for C#, it will be either mono or Roslyn,
  • Other programs may be needed to debug, execute, or organize larger projects, such as makefile or MKBundle.

IDE “bundle” all of those functionality into a single interface, to ease the workflow of the programmer. This means sometimes that programmers have fewer control over their tools, but that it is easier to get started.

• Integrated Development Environment (IDE)
  • Combines a text editor, compiler, file browser, debugger, and other tools
  • Helps you organize a programming project
  • Helps you write, compile, and test code in one place

In particular, Visual Studio is an IDE, and it uses its own vocabulary:

• Solution: An entire software project, including source code, metadata, input data files, etc.
• “Build solution”: Compile all of your code
• “Start without debugging”: Execute the compiled code
• Solution location: The folder (on your computer’s file system) that contains the solution, meaning all your code and the information needed to compile and execute it