Go Programming Language

In-depth internals, my personal notes, example codes and projects. Includes - Thousands of codes, OOP, Concurrency, Parallelism, Goroutines, Mutexes & Wait Groups, Testing in Go, Go tool chain, Backend web development, Some projects including Log file parser using bufio.Scanner, Spam Masker, Retro led clock, Console animations, Dictionary programs, Social Network built using Go and GopherJS, Database Connectivity and working (MySQL, MongoDB, Redis), GopherJS and lot more..

Author

Aditya Hajare (Linkedin).

Current Status

WIP (Work In Progress)!

License

Open-sourced software licensed under the MIT license.

Important Notes

• Uber Go Style Guide
• Concurrency Using Golang
• Protocol Buffers Using Golang
• REST Microservice | DDD | Gin Gonic | Testing
• REST Microservice | DDD | Gorilla/Mux
• REST Microservice | MVC | Bad Practices
• Go Configurations

    + Environment Configurations
    + VS Code Configurations

• Basics

    + Packages
        - Executable Packages
        - Library Packages
    + Function init()
    + Scopes
    + Renaming Imports
    + Exporting
    + Data Types
        - Basic Data Types
    + Variables
        - Zero Values
        - Unused variables
        - Multiple Declarations
        - Type Inference
        - Short Declaration
        - Multiple Short Declarations
        - Redeclarations With Short Declarations
    + Blank Identifier
    + fmt.Printf and fmt.Sprintf Formatting
    + Slice Vs. Array - Performance
    + Composite Types In Go

• Naming Conventions In Go
• Type System In Go

    + Important Links
    + Predeclared Types
    + Defined Types
    + Aliased Types

• Constants

    + Important Links
    + Constant Types
    + Multiple Constants Declaration
    + Typeless Or Untyped Constants
    + Default Types
    + IOTA
    + Common Abbreviations Used In Go

• Error Handling

    + nil

• Strings Runes And Bytes

    + Important Links
    + Strings Runes And Bytes 101

• Maps In Go

    + Maps 101

• Structs In Go

    + Inheritance vs. Composition
    + Structs 101

• OOP In Go With Methods And Interfaces

    + Methods
    + Pointer Receivers
    + Attaching Methods To Any Types
    + Interfaces
    + Type Assertion
    + Empty Interface
    + Type Switch

• Concurrency And Parallelism

    + Concurrency
    + Parallelism
    + Concurrency vs. parallelism

• Goroutines

    + Advantages of Goroutines over Threads

• Channels

    + Buffered Channels

• Mutexes And Wait Groups From GoSync Package

    + Mutexes
    + Wait Groups

• Go Vet
• Go Documentation Server On Local Machine

Go Configurations

+ Environment Configurations

• Open up .profile or .zshrc or .bashrc depending on our OS and add/edit following:

    #!/bin/bash
    # Specifies where the Go destribution is installed on the system.
    export GOROOT=/usr/local/go
    # Specifies top-level directory containing source code for all our Go projects.
    # Inside this directory, we need to create 3 more directories viz. "src", "pkg" and "bin".
    export GOPATH=~/adiwork/go  # This directory is also known as Go Workspace.
    # "src" directory inside Workspace represents where all the Go source code will be stored.
    # "pkg" directory inside Workspace represents where the compiled Go packages will be stored.
    # "bin" directory inside Workspace represents where the produced Go compiled binaries will be stored.
    # Specifies where Go should install compiled binaries.
    export GOBIN=${GOPATH}/bin
    # Attaching GOROOT and GOBIN to shell environment's path variable.
    export PATH=${PATH}:/usr/local/bin:${GOROOT}/bin:${GOBIN}

• Execute following command to get stringer:

    go get -u golang.org/x/tools/cmd/stringer

+ VS Code Configurations

• My VS Code configs for Go:

    {
        "go.lintTool": "golangci-lint",
        "go.formatTool": "goimports",
        "go.useLanguageServer": true,
        "go.lintOnSave": "package",
        "go.vetOnSave": "package",
        "go.vetFlags": [
            "-all",
            "-shadow"
        ]
    }

Basics

• Go is a strongly typed language. Because of that, it helps Go compiler to identify many types of errors at compile time even before our program is run.

+ Packages

• All package files, should be in the same (single) directory. i.e. all package source code files should be located in a one single directory.
• All files in a specific folder should belong to a one single package. It’s a convention, not a rule.

  There are 2 kinds of packages in Go: Executable Packages and Library Packages.

• To make a package executable, name that package main. It’s a special package.
• package clause can be used only once per file and it should be the first line in .go source file.
• Package contains multiple Go files belonging to same folder.
• Any package that is intended to run on a command-line, must declare package main.

• To alias a package upon importing:

    package main
    import fm "fmt"  // Package "fmt" has been aliased as "fm"
    func main() {
        //
    }

- Executable Packages

• It’s name should always be package main.
• Executable Package should also contain main() function and that too only once.
• These are created only for running it as a Go program.
• These cannot be imported into a Go program.
• Package name should be main.

- Library Packages

• Almost all Go Standard Library Packages are of type Library Packages.
• They are reusable packages.
• They are not executable packages. So we can’t run them.
• We can only import them.
• These are created only for reusability purposes.
• Package name can have any name.
• Doesn’t need to have function named main(). To avoid confusion, it’s better not to have function named main() in a reusable package.

+ Function init()

• The init() function is used to initialize the state of a package.
• Go automatically calls init() function before calling command-line package’smain() function.

+ Scopes

• Same name cannot be declared again inside a same scope.
• There are following types of scopes in Go:
  1. package: Each Go package has it’s own scope. For e.g. declared funcs are only visible to the files belonging to same package.
  2. file: Imported packages are only visible to the importing file. Each file has to import external packages on it’s own.
  3. func.
  4. block.

+ Renaming Imports

• We can rename an imported package name with following syntax:

    package main
    import "fmt"
    import adi "fmt"    // Imported "fmt" package and renamed it to "adi"
    func main() {
        adi.Println("नमस्ते आदित्य")    // This will print "नमस्ते आदित्य"
    }

• We can import packages with the same name into same file by giving one of them imports a new name.

+ Exporting

• To export a name in Go, just make it’s first letter an uppercase letter.

• For e.g.

    package aditest
    func Adi() { // 'Adi()' will be exported and will be available throughout 'aditest' package
        // Code..
    }
    func adiNew() { // 'adiNew()' will not be exported since it's name doesn't start with uppercase letter.
        // Code
    }

+ Data Types

• literal means the value itself. Unline variable, a literal doesn’t have a name.
• There are following data types in Go:
  • Basic type: Numbers, strings, and booleans come under this category.
  • Aggregate type: Array and structs come under this category.
  • Reference type: Pointers, slices, maps, functions, and channels come under this * category.
  • Interface type

- Basic Data Types

• Following are the basic data types in Go:

  • Numeric:

      // Integer Types
      uint8   // Unsigned 8-bit integers (0 to 255)
      uint16  // Unsigned 16-bit integers (0 to 65535)
      uint32  // Unsigned 32-bit integers (0 to 4294967295)
      uint64  // Unsigned 64-bit integers (0 to 18446744073709551615)
      int8    // Signed 8-bit integers (-128 to 127)
      int16   // Signed 16-bit integers (-32768 to 32767)
      int32   // Signed 32-bit integers (-2147483648 to 2147483647)
      int64   // Signed 64-bit integers (-9223372036854775808 to 9223372036854775807)
      // +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
      // Floating Types
      float32     // IEEE-754 32-bit floating-point numbers
      float64     // IEEE-754 64-bit floating-point numbers
      complex64   // Complex numbers with float32 real and imaginary parts
      complex128  // Complex numbers with float64 real and imaginary parts
      // +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
      // Other Numeric Types
      byte     // same as uint8
      rune     // same as int32
      uint     // 32 or 64 bits
      int      // same size as uint
      uintptr  // an unsigned integer to store the uninterpreted bits of a pointer value

  • Boolean:

      bool    // Represents 'true' or 'false'

  • String:
    • In Go language, strings are different from other languages like Java, C++, Python, etc.
    • Strings can’t be null in Go.
    • It is a sequence of variable-width characters where each and every character is represented by one or more bytes using UTF-8 Encoding.
    • In Go, a string is in effect is a read-only slice of bytes (immutable).
    • Or in other words, strings are the immutable chain of arbitrary bytes (including bytes with zero value) and the bytes of the strings can be represented in the Unicode text using UTF-8 encoding.
    • String literals can be created in 2 ways:
      • Using double quotes
      • Using backticks

+ Variables

• Variables in Go Lang
• In Go, we have to declare a variable before we can use it. This is required and necessary for the compile time safety.
• Variables are not created at compile time. They are created at run time.
• The unnamed variables are pointers (like in C).
• Once we declare a type for a variable, it cannot be changed later. It is static.

- Zero Values

• When a variable is declared and it isn’t assigned any value at the time of declaration, Go will assign a zero value to it based on it’s variable type.
• Type of a variable decides what zero value it will take initially when declared (and if it isn’t assigned any value at the time of declaration).

    // Zero Values assigned to variables by Go when they are declared and not assigned any values at the time of declaration.
    var adiBool bool          // false
    var adiInt int            // 0
    var adiFloat float64      // 0
    var adiStr string         // ""
    var adiPointer *string    // nil | 'nil' means it doesn't point to any memory location

- Unused variables

• Unused variables in blocked scope are not allowed in Go since they cause maintenance nightmares. If we declare a variable in blocked scope then we must use it or else completely remove it from the block. We cannot have unused variables declared in blocked scope dangling in our source codes. Go throws unused variable errors at compile time only.
• We should avoid using package level variables. Go doesn’t throw unused variable errors at compile time for variables declared at package level.

- Multiple Declarations

• Sometimes it is also called as parallel variable declarations.

• Declaring multiple variables with different types in a single statement:

    package main
    func main() {
        var (
            adiBool bool
            adiInt int
            adiFloat float64
            adiStr string
            adiPointer *string
        )
    }

• Declaring multiple variables with same type in a single statement:

    package main
    func main() {
        var foo, bar, baz int
    }

- Type Inference

• Type Inference means Go can figure out the type of a variable automatically from it’s assigned value.
• If we are assigning value to a variable at the time of it’s declaration, we can ommit it’s type specification.

• For e.g.

    package main
    main() {
        var someFlag = true // We are not specifying type of 'someFlag' as bool here.
    }

- Short Declaration

• With Type Inference, Go can figure out variable type based off it’s assigned value.
• In Short Declaration, we can declare variable by completely ommitting var keyword along with it’s variable type.
• It declares and initializes the variable.
• We cannot use Short Declaration syntax to declare variables in Package Scope.
• At Package Scope, all declarations should start with a keyword. Since Short Declaration syntax doesn’t have any keyword in it, it doesn’t work at Package Scope.

• For e.g.

    package main
    main() {
        someFlag := true // 'var' keyword and 'variable type' is not specified. It works!
    }

- Multiple Short Declarations

• We can declare and initialize multiple variables of different types using short declaration syntax:

    package main
    main() {
        someFlag, age, name := true, 30, "आदित्य" // Multiple variables of different types.
    }

• In this type of declaration, number of values and number of names must be the same. Otherwise it will result in error.

- Redeclarations With Short Declarations

• Short Declaration can initialize new variables and assign to existing variables at the same time.
• At least one of the variable in Short Declaration Redeclaration must be a new variable.

• For e.g.

    package main
    main() {
        var someFlag bool
        // someFlag := true // Error! At least one variable must be new to make this work.
        someFlag, age := true, 30 // This works! Because 'age' is a new variable being declared in the same statement. someFlag will be set (redeclared) to true.
    }

+ Blank Identifier

“There are only two hard things in Computer Science: cache invalidation and naming things”. Tim Bray quoting Phil Karlton

• Go doesn’t allow unused variables in blocked scope.
• To ignore a variable, Blank Identifier (_) is used as a variable name in Go.
• Go compiler will not throw unsed variable error if a blocked scope variable is named _.
• We cannot use value assigned to _.
• It is like a black hole that swallows variable.
• Detailed information and usage of Blank Identifier

+ fmt.Printf and fmt.Sprintf Formatting

• Following formatting can be used with fmt.Printf as well as fmt.Sprintf:

    // String and slice of bytes
    %s  // the uninte­rpreted bytes of the string or slice
    %q  // a double­-quoted string safely escaped with Go syntax
    %x  // base 16, lower-­case, two characters per byte
    %X  // base 16, upper-­case, two characters per byte
    // +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    // Boolean
    %t  // the word true or false
    // +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    // General
    %v  // The value in a default format. When printing structs, the plus flag (%+v) adds field names.
    %#v // a Go-syntax repres­ent­ation of the value
    %T  // a Go-syntax repres­ent­ation of the type of the value
    %%  // a literal percent sign; consumes no value
    // +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    // Integer
    %b  // base 2
    %c  // the character repres­ented by the corres­ponding Unicode code point
    %d  // base 10
    %o  // base 8
    %q  // a single­-quoted character literal safely escaped with Go syntax
    %x  // base 16, with lower-case letters for a-f
    %X  // base 16, with upper-case letters for A-F
    %U  // Unicode format: U+1234; same as "­U+%­04X­"
    // +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    // The default format for %v
    bool // %t
    int, int8 // %d
    uint, uint8 // %d, %x if printed with %#v
    float32, complex64 // %g
    string // %s
    chan // %p
    pointer // %p
    // +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    // Floati­ng-­point and complex consti­tuents
    %b  // decima­lless scientific notation with exponent a power of two, in the manner of strcon­v.F­orm­atFloat with the 'b' format, e.g. -12345­6p-78
    %e  // scientific notation, e.g. -1.234­456e+78
    %E  // scientific notation, e.g. -1.234­456E+78
    %f  // decimal point but no exponent, e.g. 123.456
    %F  // synonym for %f
    %g  // %e for large exponents, %f otherwise
    %G  // %E for large exponents, %F otherwise
    // +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    // Floati­ng-­point Precision
    %f      // default width, default precision
    %9f     // width 9, default precision
    %.2f    // default width, precision 2
    %9.2f   // width 9, precision 2
    %9.f    // width 9, precision 0
    // +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    // Pointer
    %p  // base 16 notation, with leading 0x
    // +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    // Other flags
    +   // always print a sign for numeric values; guarantee ASCII-only output for %q (%+q).
    -   // pad with spaces on the right rather than the left (left-­justify the field).
    #   // alternate format: add leading 0 for octal (%#o), 0x for hex (%#x); 0X for hex (%#X); suppress 0x for %p (%#p); for %q, print a raw (backq­uoted) string if strcon­v.C­anB­ack­quote returns true;
    ' ' (space)  // leave a space for elided sign in numbers (% d); put spaces between bytes printing strings or slices in hex (% x, % X).
    0   // pad with leading zeros rather than spaces; for numbers, this moves the padding after the sign.

+ Slice Vs. Array - Performance

• Slice operations are cheap!
• Slicing: Creates a new slice header.
• Assigning a Slice to another Slice or, passing it to a function: Only copies the slice header.
• Slice header has a fixed size and it doesn't change even if we have got millions of elements.
• Array can be expensive as compared to Slice.
• Assigning an array to another array or passing it to a function: Copies all the elements of it.

+ Composite Types In Go

• Following are the Composite Types in Go:
  • Arrays: Collection of elements. Indexable and Fixed Length.
  • Slices: Collection of elements. Indexable and Dynamic Length.
  • Strings: Byte Slices. ASCII and UNICODE.
  • Maps: Collection of Indexable Key-Value Pairs.
  • Structs: Groups different types of variables together.

Type System In Go

• At compile time, a Go compiler can catch overflow errors.
• In runtime, when overflows occurs:
  • integer wrap arounds and go to their minimum and maximum values.
  • float wrap arounds to positive infinity or negative infinity.

+ Important Links

• What is Binary?
• What’s a Bit
• What’s a Byte
• How old school graphics worked? PART 1
• How old school graphics worked? PART 2
• Stackoverflow: What actually is a Byte Stream?
• Why Byte but not uint8?

+ Predeclared Types

• A predeclared type is a built-in type that we can use everywhere without importing any package.
• A built-in type means it’s a core feature of Go i.e. it comes with compiler itself.
• A predeclared type has a name and we can use it in any scope.
• We don’t have to declare a predeclared type before using it.
• It has a type representation i.e. how Go see it and how we can use it. In other words, what values a type can represent.
• It has a size in bytes i.e. how much space it needs in memory and it also determines the range of values it can represent.
• Go cannot catch overflow errors in runtime. For e.g. A variable belongs to runtime and it’s value cannot be known at the compile time.
• In Go, when variable values overflow, they gets wrapped around i.e. They get reassigned to the minimum value their variable type can represent.

• Examples of Predeclared Types:

    bool // 'bool' is a predeclared type and it has following characteristics:
    // Name: bool
    // Representation: 'true' or 'false'
    // Size: 1 byte
    int // 'int' is a predeclared type and it has following characteristics:
    // Name: int
    // Representation: -1, 0, 1, 1000000000000000
    // Size: 8 byte

+ Defined Types

• A Defined Type is also called as Named Type.
• A Defined Type can only be created from another existing Type.
• We need to give a new name to newly created type.
• Newly Defined Type can optionally have it’s own methods.
• A type can be converted to another type if they share the same underlying type and vice versa.
• A defined type and it’s source type share the same underlying type.

• For e.g.

    // 'Duration' is the 'defined type' or 'named type'.
    // 'int64' is the 'underlying type'.
    type Duration int64
    // Type conversion
    var microSeconds int64   // 'microSeconds' variable is of type 'int64'
    var nanoSeconds Duration // 'nanoSeconds' variable is of type 'Duration'
    nanoSeconds = microSeconds // ERROR! This won't work. To make it work:
    nanoSeconds = Duration(microSeconds) // Works! We are converting 'microSeconds' to 'Named Type' we have created above i.e. 'Duration'
    microSeconds = int64(nanoSeconds) // This also works!

+ Aliased Types

• byte and uint8 are exactly the same types just with diferent names.
• rune and int32 are exactly the same types just with diferent names. i.e. rune is an alias of int32. The rune type is used to represent unicode characters.
• Type Alias declaration is not for everyday usage. It is mainly used in very huge codebase refactors.

Constants

• Constants belong to compile time. They must be initilized with value when they are declared.
• Constants are created at compile time. In the run time, Go just transforms it into a value.
• Unnamed constants: All basic literals are unnamed constants. Following are examples of basic literals:

    // Unnamed constants
    1
    3.14
    "hello"
    true
    false

• Named Constants: All named constants will be replaced to their values in runtime. They need to be declared first.
• Untyped Constants: Constants may or may not have a type.
• If the value is’nt going to change throughout our program’s lifetime and we already know the value (if it belongs to compile time) then we should go for named constants.
• Constants are immutable i.e. we cannot change their values.
• We cannot initialize a constant to a runtime value.
• We can use expressions while initializing constants.

+ Important Links

• Go’s typed and untyped constants
• Go enums and iota — Full of tips and tricks with runnable code examples

+ Constant Types

• We can declare constants using non-numeric types as well.
• Constants don’t have to be only numeric values.
• We don't have to declare the types of constants.

• For e.g.

    func main() {
        // Below works..
        const min       int     = 1
        const pi        float64 = 3.14
        const version   string  = "2.0.3"
        const debug     bool    = true
        // Declaring constants without types also works.
        const min     = 1
        const pi      = 3.14
        const version = "2.0.3"
        const debug   = true
        // We can use expressions while initializing constants.
        const min     = 1 + 1               //2
        const pi      = 3.14 * min          // 6.28
        const version = "2.0.3" + "-beta"   // 2.0.3-beta
        const debug   = !true               // false
    }

+ Multiple Constants Declaration

• Constants get their types and expressions from the previous constant.

• We can declare multiple constants in a single go as below:

    func main() {
        // Multiple constants of same type in one go
        const min, max int = 1, 1000
        // Declaring in group
        const (
            min int = 1
            max int = 1000
        )
        // Constants get their types and expressions from the previous constant
        const (
            min int = 1000     // 1000
            max                // 1000
        )
    }

+ Typeless Or Untyped Constants

• When we declare a constant without a type, it becomes untyped constant (typeless constant).
• All basic literals are also typeless. They all are typeless constant values.
• A constant with a type can only be used with a value of the same type.
• The untyped numeric constant can be used with all numeric values together.

• For e.g.

    func main() {
        const min = 42
        var i int      =  min  // Type of constant 'min' = int
        var f float64  =  min  // Type of constant 'min' = float64
        var b byte     =  min  // Type of constant 'min' = byte
        var j int32    =  min  // Type of constant 'min' = int32
        var r rune     =  min  // Type of constant 'min' = rune
    }

+ Default Types

• Conversion only happens when a type is needed.
• Go converts a typeless constant to a typed value when a type is needed.

• For e.g.

    func main() {
        const min int32 = 1000
        max := 5 + min // Type of 'max' is 'int32'
                       // Internally this happens: max := int32(5) + min
    }

• Go implicitly converts the typeless constant to a typed value.

• For e.g.

    func main() {
        const min = 1000
        max := 5 + min // Type of 'max' is 'int'
                       // Internally this happens: max := int(5) + int(min)
    }

• An untyped constant has a default type.
• Go evaluates the expression then it converts the resulting typeless value to its default value.

+ IOTA

• IOTA is nothing but a number generator for constants. In other words, it is ever increasing automatic counter.
• IOTA is built-in constant generator which generates ever increasing numbers.
• IOTA starts at 0.
• We can use expressions with IOTA. So, the other constants will repeat the expressions.
• We can use blank identifier (_) to adjust the values of constants:

    func main() {
        const (
            EST = -(5 + iota)   // -5
            _                   // -6 | Discarded/skipped due to blank identifier
            MST                 // -7
            PST                 // -8
        )
    }

+ Common Abbreviations Used In Go

• Following are some of the common Abbreviations used in Go standard libraries:

    var s string        // string
    var i int           // index
    var num int         // number
    var msg string      // message
    var v string        // value
    var val string      // value
    var fv string       // flag value
    var err error       // error value
    var args []string   // arguments
    var seen bool       // has seen?
    var parsed bool     // parsing ok?
    var buf []byte      // buffer
    var off int         // offset
    var op int          // operation
    var opRead int      // read operation
    var m int           // another number
    var c int           // capacity
    var c int           // character
    var sep string      // separator
    var src int         // source
    var dst int         // destination
    var b byte          // byte
    var b []byte        // buffer
    var buf []byte      // buffer
    var w io.Writer     // writer
    var r io.Reader     // reader
    var pos int         // position
    // ...list goes on and on...

• Use the complete words in larger scopes such as package scope.
• Use mixedCaps.
• Use all capital letters for common acronyms such as API.
• Do not use underscores in names.

Error Handling

• In Go, nil value is extensively used for Error Handling.

• For e.g.

    func main() {
        data, err := someFunc()
        if err != nil {
            fmt.Println("Error occurred")
            return
        } else {
            fmt.Println("Success")
        }
    }

+ nil

• nil is a predeclared identifier like true, false, len(), int32, float64 etc.
• Since it is a predeclared identifier, it can be used anywhere without importing any package.
• nil value means that the value is not initialized yet.
• It is similar to following identifiers in other languages:

    null    // JavaScript
    None    // Python
    null    // Java
    nil     // Ruby

• The zero value of all pointer-based types in Go is nil. Following are the pointer-based types in Go:

    pointers
    slices
    maps
    interfaces
    channels

• In Go, nil value can be untyped or typed depending on the context.

Strings Runes And Bytes

+ Important Links

• Representing letters with numbers - Overview of ASCII and Unicode
• Characters in a computer - Advanced technical videos about the underlyings of ASCII and Unicode
  • The 3rd video is especially important because it talks about UTF-8 encoding and decoding.
• Hexadecimal Number System - Hexadecimal numbers are important when working with bytes
• Go Blog: Strings

+ Strings Runes And Bytes 101

• A string value is nothing but a series of bytes.

• We can represent a string value as a byte slice. For e.g.

    "hey"                   // String value
    []byte{104, 101, 121}   // Representing string "hey" in byte slice
    []byte("hey")                   // Converting string "hey" into byte slice
    string([]byte{104, 101, 121})   // Converting byte slice into string value

• Instead of numbers (byte slice), we can also represent string characters as rune literals.
• Numbers and Rune Literals are the same thing.
• In Go, Unicode Code Points are called Runes.
• A Rune literal is a typeless integer literal.
• A Rune literal can be of any integer type. for e.g. byte (uint8), rune (int32) or any other integer type.
• In short, Rune is a Unicode Code Point that is represented by an Integer Value.
• Using UTF-8 we can represent Unicode Code Points between 1 byte and 4 bytes.
• We can represent any Unicode Code Point using the Rune Type because it can store 4 bytes of data. For e.g.

    char := '🍺'

• String values are read-only byte slices i.e.

    string value ----> read-only []byte

• String to Byte Slice conversion creates a new []byte slice and copies the bytes of the string to a new slice’s backing array. They don’t share the same backing array.
• In short, String is an immutable byte slice and we cannot change any of it’s elements. However, we can convert string to a byte slice and then we can change that new slice.
• A string is a data structure that points to a read-only backing array.
• UTF-8 is a variable length encoding (for efficiency). So each rune may start at a different index.
• for range loop jumps over the runes of a string, rather than the bytes of a string. Each index returns the starting index of the next rune.
• Runes in a UTF-8 encoded string can have a different number of bytes because UTF-8 is a variable byte-length encoding.
• Especially in scripting languages, we can manipulateUTF-8 strings by indexes easily. However, Go doesn’t allow us to do so by default because of efficiency reasons.
• Go never hides the cost of doing something.
• []rune(string) creates a new slice, and copies each rune to new slice’s backing array. This is inefficient way of indexing strings.
• A string value usually use UTF-8 so it can be more efficient because each rune on the other hand uses 1 to 4 bytes (variable-byte length).
• Each rune in []rune (Rune Slice) has the same length i.e. 4 bytes. It is inefficient because the rune type is an alias to int32.
• In Go, if our source code file is encoded into utf-8 then String Literals in our file are automatically encoded into utf-8.
• When we’re working with bytes, continue working with bytes. Do not convert a string to []byte (Byte Slice) or vice versa, unless necessary. Prefer working with []byte (Byte Slice) whenever possible. Bytes are more efficient and used almost everywhere in Go standard libraries.

Maps In Go

+ Maps 101

• Maps allows us to quickly access to an element/valueusing a unique key.
• Map keys must be unique because otherwise it can’t find the corresponding values/elements.
• The types of Map Keys and Values in Maps can be different.
• A Map Key must be a comparable type.
• All Map Keys and Map Values must belong to their corresponding types. They can't be mixed up.
• A Map Variable (or a Value) is nothing but a pointer to a Map Header Value in the memory.
• A Map Value only contains the memory address of a Map Header.

Structs In Go

+ Inheritance vs. Composition

+ Structs 101

• Structs are blueprints — They are fixed at compile-time.
• It’s like a class in OOP languages. Groups related data in a single type.
• Struct types are created at compile-time.
• A struct may store different types of data.
• Struct fields are declared at compile-time. However, struct values fill them in runtime.
• The field names and types are declared at compile-time. They are fixed and cannot change in runtime.
• Field values belong to runtime. We can change them in runtime.
• Structs cannot dynamically grow but they can have different set of types.
• Struct example:

    type VideoGame struct {
        Title string Genre string
        Published bool
    }

• Two structs are equal if all their fields are equal.
• Anonymous Fields: When the field names conflict theparent type takes priority.

OOP In Go With Methods And Interfaces

+ Methods

• Methods enhance types with additional behavior.
• Methods of the type are called Method Set.

• To attach method to a type:

    // Syntax
    // "varName Type" is called a "receiver"
    func (varName Type) funcName() {
        // Code
    }
    // Example
    // "book" is a struct here
    func (b book) printBook() {
        fmt.Println(b.title, b.price)
    }

• A receiver is nothing but method’s input parameters written before a method name.
• A method belongs to a single type.
• Methods on different types can have the same names.
• Method Expressions allows us to call methods through types. For e.g.

    // "game" is a struct type
    game.print(cod)
    game.print(battlefield)

• Behind the scenes, a method is a function that takes receiver as it’s first argument.

+ Pointer Receivers

• We can define methods on types using Pointer Receivers.
• The only difference between method and a function is that a method belongs to a type, whereas a function belongs to a package.
• Consistent Design Tip: When one of the methods in any type are using pointer receiver, it is better to convert all method receivers of that type to pointer receivers.
• We (must) use a pointer receiver when we want to make changes to a receiver variable. In other words, use a pointer receiver when the received value into method is going to be very large.

+ Attaching Methods To Any Types

• We can attach methods to any type in Go. For e.g.

    // Basic Types
    int
    string
    float64
    // Bare Types
    array
    struct
    // -----------------------
    // Do not use "Pointer Receivers" with below types since they already carry a pointer with themselves.
    // i.e. slice, map, chan, func
    // -----------------------
    // Pointer Bearing Types
    slice
    map
    chan // Channels
    // We can also attach methods to:
    func

+ Interfaces

• We declare an Interface much like as we define a user defined type.
• Interfaces decouple different types from each other so we can create more maintainable programs.
• An Interface is a Protocol, a Contract.
• Bigger the Interface the weaker the abstraction. —> Rob Pike
• It’s an abstract type. It doesn’t have any implementation. It only describes the expected behavior.
• The opposite of Abstract Type is Concrete Type.
• All the types in Go except Interface are of Concrete Type.
• For e.g. Following are Concrete Types:

    // Concrete Types
    int
    string
    float64
    array
    struct
    slice
    map
    chan
    func

• The Interface only defines the expected behavior.
• Go does not have an implements keyword.
• A Type satisfies an Interface automatically when it has all the methods of the Interface without explicitely specifying it.
• Interface values are comparable.
• Go interfaces are implicit. The implementing types don’t need to specify that they implement an interface.
• Interface declaration example:

    type MyInterface interface {
        foo() int
        bar() float64
        baz() string
    }

+ Type Assertion

• Type Assertion allows us to extract the dynamic value from Interface.
• It can also be used to check (assert) whether the Interface Value provides the method we want.

+ Empty Interface

• Do not use Empty Interface unless really necessary.
• Every type in Go implements the empty interface.
• An Interface Value has 2 parts:
  • A dynamic Value.
  • A dynamic Type.
• Empty Interface is the one which doesn’t have any methods.
• Every Type satisfies the Empty Interface.
• It can represent any Type of Value.
• We can’t directly use the dynamic value of an empty interface value.

• Example of Empty Interface:

    type someInterface interface {
    }

• To use a value from Empty Interface, we first need to extract it using Type Assertion.
• Empty Interface Slice contains the Empty Interface Values.
• Example use cases of empty interfaces:
  • A function that returns a value of interface{} can return any type.
  • We can store heterogeneous values in an array, slice, or map using the empty interface{} type.

+ Type Switch

• Type Switch allows us to detect and extract dynamic values from Interface Values using Switch Statement.
• When we have lots of conditions to check, we can use Type Switch.
• Example of Type Switch statement:

    // "v"     ---> Interface Value
    // "type"  ---> Extracts type from the Interface Value "v"
    // "e"     ---> Extracted value will be assigned to variable "e". It changes depending on the extracted value.
    switch e := v.(type) {
        case int:
            // "e" is an "int" here..
        case string:
            // "e" is an "string" here..
        default:
            // "e"'s type equals to "v"'s type..
    }

• Unlike regular Switch which compares values, the Type Switch compares types of the values.

Concurrency And Parallelism

+ Concurrency

• The composition of independently executing tasks.
• Applied when dealing with handling lots of things at once.
• The focus is on how to structure a solution to solve a problem which may or may not be solved in a parallel manner.
• Concurrency is a way to structure a program by breaking it into pieces that can be executed independently.
• Communication is the means to coordinate the independent executions.
• Go supports concurrency. Go provides:
  • concurrent execution (goroutines).
  • synchronization and messaging (channels).
  • multi-way concurrent control (select).
• IMPORTANT POINTS:
  • Concurrency is powerful.
  • Concurrency is not parallelism.
  • Concurrency enables parallelism.
  • Concurrency makes parallelism (and scaling and everything else) easy.

+ Parallelism

• Parallelism is the simultaneous execution of computations.
• Programming as the simultaneous execution of (possibly related) computations.
• It’s all about doing lots of things at once.

+ Concurrency vs. parallelism

• “Concurrency is about dealing with lots of things at once. Parallelism is about doing lots of things at once.” — Rob Pike
  • Concurrency is a property of a program where two or more tasks can be in progress simultaneously. Parallelism is a run-time property where two or more tasks are being executed simultaneously. Through concurrency you want to define a proper structure to your program. Concurrency can use parallelism for getting its job done but remember parallelism is not the ultimate goal of concurrency.
• Concurrency is about dealing with lots of things at once.
• Parallelism is about doing lots of things at once.
• Not the same, but related.
• Concurrency is about structure, parallelism is about execution.
• Concurrency provides a way to structure a solution to solve a problem that may (but not necessarily) be parallelizable.
• An analogy
  • Concurrent: Mouse, keyboard, display, and disk drivers.
  • Parallel: Vector dot product.

Goroutines

• In Go, concurrency is achieved by using Goroutines.
• Goroutines are functions or methods which can run concurrently with others methods and functions.
• Goroutines are lightweight threads that are managed by the Go runtime.
• They are very much similar like threads in Java but light weight and cost of creating them is very low.
• When we run a function as a Goroutine, we are running the function concurrently.
• Place the keyword go before a function call to execute it as a Goroutine.
• To run a method or function concurrently prefix it with keyword go.

• For e.g.

    package main
    import (
        "fmt"
        "time"
    )
    func print() {
        fmt.Println("Printing from goroutine")
    }
    func main() {
        go print()
        time.Sleep(1 * time.Second)
        fmt.Println("Printing from main")
    }

• Program is terminated when main() function execution is completed. When the program terminates, all Goroutines are terminated regardless of the fact if all the Goroutines has completed execution or not.
• We can also run Anonymous Functions as Goroutines as follows:

    // Executing anonymous function as Goroutine
    go func() {
        //
    }()

+ Advantages of Goroutines over Threads

• Goroutines have a faster startup time than threads.
• Goroutines come with built-in primitives to communicate safely between themselves called as channels.
• Goroutines are extremely cheap when compared to threads. They are only a few kb in stack size and the stack can grow and shrink according to needs of the application whereas in the case of threads the stack size has to be specified and is fixed.

Channels

• Channels are conduits (pipes) that we can use to pass values of a particular type from one Goroutine to another.
• Channels are a mechanism for communication.
• Channels allows Goroutines to share memory by communicating
• We can use Channel Operators: <-, -> to send and receive values.
  • NOTE: The data flows in the direction of the arrow.
• We can create a Channel using built-in make() function as below:

    ch := make(chan type) // type: Data Type

• Normal Channels are Synchronous. i.e. Both the sending side and the receiving side of the channel wait until the other side is ready.

+ Buffered Channels

• Buffered Channels are Asynchronous. i.e. Sending and Receiving messages through Buffered Channels will not block unless the Channel is full.
• We can create a Buffered Channel same way as we create the Normal Channels using the built-in make() function. The only difference is, we can pass the second parameter to make() function which indicates the Buffered Channel’s Buffering Capacity.
• For e.g.

    ch := make(chan type, capacity)

• NOTE: If we pass the Buffering Capacity as 1, we are creating a Normal Channel. To create a Buffered Channel, we have to pass Buffering Capacity as greater than 1

Mutexes And Wait Groups From GoSync Package

• Since Goroutines run in a same address space, they have access to shared memory and this access must be synchronised. Go’s motto is to share memory by communicating (Goroutines and Channels makes this possible).
• Sometimes, some problems are better suited to using the traditional forms of synchronisation. Go allows us to make use of these Synchonisation Primitives by using the Sync package.

+ Mutexes

• A Race Condition happens when two or more threads can access shared data and try to change that shared data at the same time. We can use Mutex to solve this problem.
• A Mutex is a Mutual Exclusion Lock. It’s a Synchronisation Primitive.
• It is used to protect shared data which is simultaneously accessed by multiple treads.
• Making changes to shared data and reading a shared data (e.g. for printing purposes) are still considered accessing the same data simultaneously.

+ Wait Groups

• Wait Groups are another Synchronisation Primitive.
• A Wait Group basically waits for collection of Goroutines to finish execution.

Go Vet

• go vet command helps us catch errors which are not generally caught by Go Compiler.
• For e.g.
  • Go to directory: 62-Go-Vet-To-Catch-Errors
  • Execute file with below command:

      go vet main.go

  • It will catch the error where int is supplied to fmt.Printf() whereas string value should’ve been supplied. This error isn’t caught by Go Compiler since the program is still syntactically correct.

Go Documentation Server On Local Machine

• To start a Go Documentation Server On Local Machine, execute:

    # Install godoc with following command first:
    # go get golang.org/x/tools/cmd/godoc
    # Then execute:
    godoc -http=:6060

• In browser, visit:

    http://localhost:6060