The GO programming language

Marco Sabatini
Senior Software Engineer
sabatini.m@gmail.com
https://it.linkedin.com/in/sabatinimarco
Contacts
1

Agenda
Agenda 1° Day
• Overview
• Variable types overview and scope
• Environment setup
• Packages
• Exercises
Agenda 3° Day
• Concurrent programming
• File Handling
• Linking C shared libs
• Exercises
Agenda 2° Day
• Pointers
• Collections
• Procedural programming
• Struct
• Interfaces
• Exercises
3
Agenda 4°Day
• Gorilla web toolkit
• Gorm
• Exercises

Agenda 1° Day
Agenda
• Overview
• Packages
• Exercises
4

Overview (authors)
• Go, also commonly referred to as golang, is a programming
language initially developed at Google in 2007 by Robert
Griesemer, Rob Pike, and Ken Thompson.
• It is a statically-typed language with syntax loosely derived
from that of C, adding garbage collection, type safety, some
dynamic-typing capabilities, additional built-in types such as
variable-length arrays and key-value maps, and a large
standard library.
•Go programs are “compiled” rather than interpreted so as to
have the best possible performance.
5

Overview (Google has big problems)
Go is a programming language designed by Google to help
solve Google’s problems.
What problems?
• C++ for servers, plus lots of Java and Python
• thousand of engineers
• gazillions of lines of code
• distributed build system
• gazillions of machines, which are treated as a modest
number of compute clusters
6

Overview (The reason of go)
Goals:
• eliminate slowness
• eliminate clumsiness
• improve effectiveness
• maintain (even improve) scale
Go was designed by and for people who write – read – debug
– maintain, large software systems.
Go’s purpose is not research programming language
design.
Go’s purpose is to make its designers’ programming lives better.
7

Overview (project status from Google)
•Go became a public open source project on November 10, 2009. After a
couple of years of very active design and development, stability was called
for and Go 1 was released on March 28, 2012. Go 1, which includes a
language specification, standard libraries, and custom tools, provides a
stable foundation for creating reliable products, projects, and publications.
•With that stability established, we are using Go to develop programs,
products, and tools rather than actively changing the language and
libraries. In fact, the purpose of Go 1 is to provide long-term stability.
Backwards-incompatible changes will not be made to any Go 1 point
release. We want to use what we have to learn how a future version of Go
might look, rather than to play with the language underfoot
•Of course, development will continue on Go itself, but the focus will be on
performance, reliability, portability and the addition of new functionality
such as improved support for internationalization.
9

Overview (important stuff)
• Object oriented without inheritance
• Syntax sugar declaration
• Strong and Static types
• Interfaces, but no declaration (duck typing)
• Functions and Methods (with receivers)
• Visibility with the first letter of symbol
• No exceptions (error interface)
• Unused imports and variables occurs compile error
• Excellent and complete standard library
10

Overview (cross compilation)
12

Overview (let’s start coding…)
// hello.go
package main
import ( ➊ "fmt"
"os"
"strings"
)
func main() {
who := "World!" ➋
if len(os.Args) > 1 { /* os.Args[0] is "hello" or "hello.exe" */ ➌
who = strings.Join(os.Args[1:], " ") ➍ }
fmt.Println("Hello", who) ➎ }
Go uses C++-style comments: // for single-line
comments that finish at the end of the line and /* ...
*/ for comments that can span multiple lines.
Every piece of Go code exists inside a package, and
every Go program must have a main package with a
main() function which serves as the program’s entry
point, that is, the function that is executed first.
The import statement (➊) imports three packages
from the standard library.
The first statement in the main() function (➋; using the :=
operator) is called a short variable declaration in Go
terminology.
The os.Args variable is a slice of strings (➌).
If the user has entered one or more command line
arguments the if condition is satisfied and we set the who
string to contain all the arguments joined up as a single
string (➍).
Finally, in the last statement (➎), we print Hello, a space,
the string held in the who variable, and a newline.
13

Syntax overview 1/2
• Basically C-like with reversed types and declarations,
plus keywords to introduce each type of declaration.
• Basic control structures are familiar:
• Note: no parentheses, but braces required.
14

Syntax overview 2/2
Semicolons terminate statements but:
• lexer inserts them automatically at end of line if the previous token could
end a statement.
• Note: much cleaner, simpler than JavaScript rule!
Thus no semis needed in this program:
In practice, Go code almost never has semicolons outside for and if clauses
15

Declaration
Declarations are introduced by a keyword (var, const,
type, func) and are reversed compared to C:
Why are they reversed? Earlier example:
• Both p and q have type *int.
• Also functions read better and are consistent with other
declarations. 16

Var Identifier
• Variable declarations are introduced by “var” identifier.
• They may have a type or an initialization expression;
• One or both must be present.
• Initializers must match variables (and types!).
17

Distributing Var
Annoying to type var all the time. Group with parens:
Applies to const, type, var but not func!!!
18

The := “short declaration”
Within functions (only), declarations of the form:
can be shortened to:
(Another reason for the name/type reversal.)
The type is that of the value (for ideal numbers, get int or
float64 or complex128, accordingly.)
These are used a lot and are available in places such
as for loop initializers.
19

Const
• Constant declarations are introduced by const.
• They must have a "constant expression", evaluated at compile
time, as initialiser and may have an optional type specifier.
20

Iota
Constant declarations can use the counter iota, which
starts at 0 in each const block and increments at each
implicit semicolon (end of line).
Shorthand: Previous type and expressions repeat.
21

Type
Type declarations are introduced by type. We'll learn more
about types later but here are some examples:
22

New
The built-in function new allocates memory. Syntax is like
a function call, with type as argument, similar to C++.
Returns a pointer to the allocated object (pointers )
Later we'll see how to build slices and such.
There is no delete or free; Go has garbage collection.
23

Assignment
Assignment is easy and familiar:
But multiple assignment works too:
Functions can return multiple values (details later):
24

If
Basic form is familiar, but no dangling else problem:
Initialization statement allowed; requires semicolon.
Useful with multivariate functions:
Missing condition means true, which is not too useful in this
context but handy in for, switch.
25

For
Basic form is familiar:
Missing condition means true:
But you can leave out the semis too:
Don't forget multivariate assigments:
26

Switch 1/2
Switches are somewhat similar to C's.
But there are important syntactic and semantic differences: 
- expressions need not be constant or even int.
- no automatic fall through 
- instead, lexically last statement can be fall-through
- multiple cases can be comma-separated
27

Switch 2/2
Go's switch is more powerful than C's. Familiar form:
The expressions can be any type and a missing switch
expression means true. Result: if-else chain:
or
28

Break and continue …
The break and continue statements work as in C. They
may specify a label to affect an outer structure:
29

Agenda 1° Day
Agenda
• Overview
• packages
• Exercises
30

Go Typing
• Go is a statically typed programming language. This
means that variables always have a specific type and
that type cannot change.
• Static typing may seem cumbersome at first. You'll spend
a large amount of your time just trying to fix your
program so that it finally compiles. But types help us
reason about what our program is doing and catch a
wide variety of common mistakes.
• Go comes with several built-in data types which we will
now look at in more detail.
31

Numeric 1/3
A. Generally we split numbers into two different kinds:
integers and floating-point numbers.
• Go's integer types are: uint8, uint16, uint32, uint64,
int8, int16, int32 and int64. 8, 16, 32 and 64 tell us
how many bits each of the types use.
• In addition there two alias types: byte which is the
same as uint8 and rune which is the same as int32.
• There are also 3 machine dependent integer types: uint, int
and uintptr. They are machine dependent because their
size depends on the type of architecture you are using.
32

Numeric 2/3
• Floating point numbers are numbers that contain a decimal component
(real numbers). (1.234, 123.4, 0.00001234, 12340000)
• Like integers floating point numbers have a certain size (32 bit or 64 bit).
Using a larger sized floating point number increases it's precision. (how
many digits it can represent)
• In addition to numbers there are several other values which can be
represented: “not a number” (NaN, for things like 0/0) and positive and
negative infinity. (+∞ and −∞)
• Go has two floating point types: float32 and float64
• Additional types for representing complex numbers (numbers with
imaginary parts): complex64 and complex128
33

String 1/2
• Go strings are made up of individual bytes, usually one
for each character. (Characters from other languages
like Chinese are represented by more than one byte)
• String literals can be created using double quotes
"Hello World" or back ticks `Hello World`.
• Common operations:
• len(“Hello World!!!”)
• “Hello World”[1] (access individual char)
• “Hello ”+”World!!!” (string concatenation)
• Examples!!!
35

String 2/2
String memory organization
36

Boolean
A boolean value (named after George Boole) is a special
1 bit integer type used to represent true and false (or
on and off). Three logical operators are used with
boolean values:
37

Agenda 1° Day
Agenda
• Overview
• packages
• Exercises
38

Golang Setting programming environment 1/2
Compiler&Runtime
• https://golang.org/doc/install (package download web
pages)
• export GOPATH=$HOME/go (environment variable to
store GO base path)
• export PATH=$PATH:$GOPATH/bin (add “GO”
runtime binary to path)
39

Golang Setting programming environment 2/2
IDE
• GoClipse is an Eclipse IDE for the Go programming language. The purpose of
GoClipse is to create an environment in which development of Go is easy for a
beginner to intermediate user.
• https://goclipse.github.io ( documentation and download link )
• Code completion ( setting GOPATH )
• Source navigation
• Compilation
• Execution (Run As)
• Debugging (Debug As)
40

Agenda 1° Day
Agenda
• Overview
• Packages
• Exercises
41

Golang source code project structure
Go code must be kept inside a workspace.
A workspace is a directory hierarchy with three directories at its root:
• root_project
• src
• bin
• pkg
then you have to export this env variables:
export GOROOT=/usr/local/go
export GOPATH=~/workspace/me/go
export PATH=$PATH:$GOROOT/bin:$GOPATH/bin
build: “go install”
42

Golang packages&docs
Packages only really make sense in the context of a
separate program which uses them.
• Execute “go install”
creates average.a
linkable object file.
• Can reuse your code
Example of documentation:
godoc average_kg
godoc -http:”6060”
Golang has a large number of precompiled packages we can use in our
software!!!
43

Testing code
• simple_test_example: go test simple_test_example
• bench_test_example: go test bench_test_example
44

Agenda 1° Day
Agenda
• Overview
• packages
• Exercises
45

Exercises
1. reverse a string ascii and utf8
2. packaging reverse string ascii&utf8 func
3. pattern match (find email in a text)
4. packaging pattern match email func
5. pass values from CLI to previous exercises (args)
6. pass values from CLI to previous exercises (flags
pkg)
7. fmt.Scanf (read doc and use it)
8. remove duplicates char in a string
46

Agenda 2° Day
Agenda
• Pointers
• Collections
• Struct
• Interfaces
• Exercises
47

Pointers Overview 1/2
When we call a function that takes an argument, that
argument is copied to the function:
In this program the zero function will not modify the original x
variable in the main function. But what if we wanted to? One
way to do this is to use a special data type known as a pointer
48

Pointers Overview 2/2
Pointers reference a location in memory where a value is
stored rather than the value itself. (They point to
something else) By using a pointer (*int) the zero
function is able to modify the original variable.
49

‘*’ and ‘&’ operators 1/2
1. In Go a pointer is represented using the * (asterisk)
character followed by the type of the stored value.
2. ‘*’ is also used to “dereference” pointer variables.
Dereferencing a pointer gives us access to the value the
pointer points to.
3. *xPtr = 0 we are saying “store the int 0 in the memory
location xPtr refers to”.
4. xPtr = 0 instead we will get a compiler error because xPtr is
not an int it's a *int, which can only be given another *int.
50

‘*’ and ‘&’ operators 1/2
We use the ‘&’ operator to find the
address of a variable.
&i returns a *int (pointer to an int)
because i is an int.
This is what allows us to modify the
original variable. &i and ptr are the
same memory location.
51

The New function
Another way to get a pointer is to use the built-in new
function:
In some programming languages there is a significant difference between using new
and &, with great care being needed to eventually delete anything created with new.
Go is not like this, it's a garbage collected programming language which means
memory is cleaned up automatically when nothing refers to it anymore.
new takes a type as an
argument, allocates
enough memory to fit a
value of that type and
returns a pointer to it.
52

Garbage Collector
Go 1 > 1.4 has used a parallel stop-the-world (STW) collector. While STW collection
has many downsides, it does at least have predictable and controllable heap growth
behavior. The sole tuning knob for the STW collector was “GOGC”, the relative heap
growth between collections. The default setting, 100%, triggered garbage collection
every time the heap size doubled over the live heap size as of the previous collection
Go 1.5 will introduce a concurrent collector.
53

Agenda 2° Day
Agenda
• Pointers
• Collections
• Struct
• Interfaces
• Exercises
54

Arrays
An array is a numbered sequence of elements of a
single type with a fixed length.
In Go they look like this:
declares “ar” to be an array of 3 integers, initially all
set to zero.
Size is part of the type. Built-in function len reports size:
55

Arrays literals
All the composite types have similar syntax for creating
values. For arrays it look like this:
Array of 3 integers:
Array of 10 integers, first three not zero:
Don't want to initialize them all? Use key:value pairs:
56

Arrays are values
Arrays are values, not implicit pointers as in C.
You can take an array's address, yielding a pointer to
the array (to pass it efficiently to a function):
Output
57

Pointers to arrays literals
You can take the address of an array literal to get a
pointer to a newly created instance:
Output
58

Slices
• A slice is a reference to a section of an array.
• Slices are used much more often than plain arrays.
• A slice is very cheap. (More about this soon.)
• A slice type looks like an array type without a size:
Built-in len(a) returns the number of elements.
Create a slice by "slicing" an array or slice:
Valid indexes of a will then be 0 and 1; len(a)==2.
59

Slice shorthands
When slicing, first index defaults to 0:
ar[:n] means the same as ar[0:n].
Second index defaults to len(array/slice):
ar[n:] means the same as ar[n:len(ar)].
Thus to create a slice from an array:
ar[:] means the same as ar[0:len(ar)].
60

A slices references an array
Conceptually
61

Making Slices
Slice literals look like array literals without a size:
What this does is create an array of length 5 and then create a slice
to refer to it.
We can also allocate a slice (and underlying array) with the built-in
function make:
• Why make not new? Because we need to make a slice, not
just allocate the memory.
• Note make([]int, 10) returns []int while new([]int) returns *[]int.
• Use make to create slices, maps, and channels.
62

Slices capacity
A slice refers to an underlying array, so there may be
elements off the end of the slice that are present in the
array.
The built-in function cap (capacity) reports how long the
slice could possibly grow after.
len(a) is 2 and cap(a) is 5. We can "reslice":
len(a) is now 4 but cap(a) is still 5.
63

Resizing Slices
Slices can be used like growable arrays. Allocate one using
make with two numbers - length and capacity - and reslice as it
grows:
Thus sl's length is always the number of elements, but it
grows as needed.
This style is cheap and idiomatic in Go!!!
64

Slices are cheap
• Feel free to resize slices as required. They are cheap to pass around;
no need to allocate.
• Remember they are references, so underlying storage can be modified.
• For instance, I/O uses slices, not counts:
Split a buffer:
Strings can be sliced too, with similar efficiency. 65

Slice functions append()
Go includes two built-in functions to assist with slices:
append and copy. Here is an example of append:
After running this program slice1 has [1,2,3] and slice2
has [1,2,3,4,5]. append creates a new slice by taking
an existing slice (the first argument) and appending
all the following arguments to it.
66

Here is an example of copy:
After running this program slice1 has [1,2,3] and slice2
has [1,2]. The contents of slice1 are copied into slice2,
but since slice2 has room for only two elements only
the first two elements of slice1 are copied.
Slice functions copy()
67

Maps
Maps are another reference type.
They are declared like this:
This declares a map indexed with key type string and
value type float64. It is analogous to the C++ type
*map<string,float64> (note the *).
Given a map m, len(m) returns the number of keys.
68

Map creation
As with a slice, a map variable refers to nothing; you
must put something in it before it can be used.
Three ways:
1) Literal: list of colon-separated key:value pairs
2) Creation
3) Assignment
69

Indexing Map
Next few examples all use:
Access an element as a value; if not present, get zero
value for the map's value type:
Set an element (setting twice updates value for key)
70

Testing existence
To test if a key is present in the map, we can use a multi-
value assignment, the "comma ok" form:
or idiomatically:
If x is present in the map, sets the boolean to true and the value to the
entry for the key. If not, sets the boolean to false and the value to the
zero for its type.
71

Deleting
Deleting an entry in the map is a multi-variate assignment
to the map entry:
If keep is true, assigns v to the map; if keep is false,
deletes the entry for key x. So to delete an entry:
delete(m, x )
72

For and range (Arrays, Slices, Maps)
The for loop has a special syntax for iterating over arrays,
slices, maps (and more, as we'll see tomorrow).
With only one variable in the range, get the key:
Variables can be assigned or declared using ‘:=‘ .
For arrays and slices, get index and value.
73

Range Over String
A for using range on a string loops over Unicode code points, not
bytes. (Use []byte for bytes, or use a standard for). The string
is assumed to contain UTF-8.
If erroneous UTF-8 is encountered, the character is set
to U+FFFD and the index advances by one byte.
74

Agenda 2° Day
Agenda
• Pointers
• Collections
• Struct
• Interfaces
• Exercises
75

Functions
A function is an independent section of code that maps zero or more
input parameters to zero or more output parameters. Functions
(also known as procedures or subroutines) are often represented
as a black box: (the black box represents the function)
We will now begin writing programs that use more than
one function.
76

Your second function
This program computes the average of a series of numbers.
Finding the average like this is a very general problem, so
its an ideal candidate for definition as a function.
The average function will need to take in a slice of float64s
and return one float64. Insert this before the main function:
77

Functions (variable scope)
Functions don't have access to anything in the calling
function.
ok
okWrong
78

Functions (call stack)
Call Stack
Each time we call a function we push it onto the call stack and
each time we return from a function we pop the last function
off of the stack.
79

Returning multiple values
Go is also capable of returning multiple values from a function
Three changes are necessary:
A. change the return type to contain multiple types
separated by ,
B. change the expression after the return so that it
contains multiple expressions separated by ,
C. change the assignment statement so that
multiple values are on the left side of the := or
=.
Multiple values are often used to return an error value along with the
result (x, err := f()), or a boolean to indicate success (x, ok := f()).
80

Variadic functions
• By using “…" before the type
name of the last parameter you
can indicate that it takes zero
or more of those parameters.
• In this case we take zero or
more ints.
• We invoke the function like any
other function except we can
pass as many ints as we want.
This is precisely how the fmt.Println function is implemented:
81

Closure 1/2
It is possible to create functions inside of functions:
add is a local variable that has the
type func(int, int) int (a function that
takes two ints and returns an int).
When you create a local function like
this it also has access to other local
variables.
A function like this together with the
non-local variables it references is
known as a closure. In this case
increment and the variable x form
the closure.
82

Closure 2/2
It is possible to create functions inside of functions:
• One way to use closure is
by writing a function which
returns another function.
• makeEvenGenerator returns
a function which generates
even numbers. Each time
it's called it adds 2 to the
local i variable which.
83

Recursion
Finally a function is able to call itself
No Factorial example please!!!
Some hacking on Go source code (path.go) !!!
84

Defer
Go has a special statement called defer which schedules
a function call to be run after the function completes
defer is often used when resources need to be freed in
some way. For example when we open a file we need
to make sure to close it later.
This has 3 advantages:
(1) it keeps our Close call near our Open call so its easier
to understand,
(2) if our function had multiple return statements (perhaps
one in an if and one in an else) Close will happen before
both of them and
(3) deferred functions are run even if a run-time panic
occurs.
85

Panic & Recover
We can handle a run-time panic with the built-in recover function.
Recover stops the panic and returns the value that was passed to the call to
panic.
A panic generally indicates a programmer error (for example attempting to access
an index of an array that's out of bounds, forgetting to initialize a map, etc.)
Error Ok
86

Memoization
If a pure function is expensive to compute and frequently used with the same
arguments, we can use memoization to reduce the processing overhead
The memoization technique is where we store the results of a computation
so that when the same computation is next requested we are able to return
the stored results rather than perform the computation all over again
87

Higher order function
A higher order function is a function that takes one or more other
functions as arguments and uses them in its own body.
Output?
88

Agenda 2° Day
Agenda
• Pointers
• Collections
• Structs
• Interfaces
• Exercises
89

Structs
Structs should feel very familiar: simple declarations of data
fields.
More usual:
Structs allow the programmer to define the layout of memory.
90

Structs are values
Structs are values and new(StructType) returns a pointer to a
zero value (memory all zeros).
There is no -> notation for structure pointers. Go provides the
indirection for you.
91

Making structs
Structs are values so you can make a zeroed one just
by declaring it.
You can also allocate one with new.
Struct literals have the expected syntax
As with arrays, taking the address of a struct literal gives the address
of a newly created value.These examples are constructors.
92

Exporting types and fields
The fields (and methods, coming up soon) of a struct must start
with an Uppercase letter to be visible outside the package.
Private type and fields:
Exported type and fields:
Exported type with mix of fields:
You may even have a private type with exported
fields. Exercise: when is that useful? 93

Anonymous fields
• Inside a struct, you can declare fields, such as another
struct, without giving a name for the field.
• These are called anonymous fields and they act as if
the inner struct is simply inserted or "embedded" into
the outer.
• This simple mechanism provides a way to derive some
or all of your implementation from another type or
types.
An example follows!!!
94

An anonymous struct field
B acts as if it has four fields, ax,
ay, bx, and by. It's almost as if
B is {ax, ay int; bx, by float64}.
However, literals for B must be
filled out in detail:
Prints: 1, 2, 3, 4
95

Anonymous fields have type as name
But it's richer than simple interpolation of the fields: B
also has a field A. The anonymous field looks like a
field whose name is its type.
Prints {1 2}.
If A came from another package,the field is called A:
96

Anonymous fields of any type
Any named type, or pointer to one, may be used in an
anonymous field and it may appear at any location in
the struct.
Prints 3.5 7 hello
97

Conflicts and hiding
If there are two fields with the same name (possibly a
type-derived name), these rules apply:
1) An outer name hides an inner name. 
This provides a way to override a field/method.
2) If the same name appears twice at the same level, it
is an error if the name is used by the program. (If it's
not used, it doesn't matter.)
No rules to resolve the ambiguity; it must be fixed.
98

Conflict examples
Using c.a is an error: is it c.A.a or c.B.a?
Using d.b is OK: it's the float64, not d.B.b Can get at
the inner b by D.B.b.
99

Methods on struct pointers
• Go has no classes, but you can attach methods to any type.
• The methods are declared, separate from the type declaration,
as functions with an explicit receiver (like java this).
• The obvious struct case:
Note: explicit receiver (no automatic this), in this case of
type *Point, used within the method.
100

Methods on struct values
A method does not require a pointer as a receiver.
This is a bit expensive, because the Point3 will always be
passed to the method by value, but it is valid Go.
101

Invoking a methods
Just as you expect.
A non-struct example:
102

Ground rules for methods
• Methods are attached to a named type, say Foo, and
are statically bound.
• The type of a receiver in a method can be either *Foo
or Foo. You can have some Foo methods and some
*Foo methods.
• Foo itself cannot be a pointer type, although the
methods can have receiver type *Foo.
• The type Foo must be defined in the same package
as all its methods.
103

Pointers & Values
Go automatically indirects/dereferences values for you when
invoking methods.
For instance, even though a method has receiver type *Point you
can invoke it on an addressable value of type Point.
Similarly, if methods are on Point3 you can use a value
of type *Point3:
104

Methods on anonymous fields
Naturally, when an anonymous field is embedded in a
struct, the methods of that type are embedded as
well - in effect, it inherits the methods.
This mechanism offers a simple way to emulate some
of the effects of subclassing and inheritance.
105

Overriding a method
Overriding works just as with fields.
Of course you can have
multiple anonymous fields
with various types - a
simple version of multiple
inheritance. The conflict
resolution rules keep things
simple, though.
107

Another example
A more compelling use of an anonymous field.
Note that Lock's receiver is (the address of) the Mutex field,
not the surrounding structure. (Contrast to subclassing)
108

Other types 1/3
Methods are not just for structs. They can be defined for any
(non-pointer) type.
The type must be defined in your package though. You can't
write a method for int but you can declare a new int type and
give it methods.
109

Other types 2/3
Now we have an enumeration-like type that knows how
to print itself.
110

Other types 3/3
By techniques to be divulged soon, fmt.Print and friends
can identify values that implement the method.
String as we defined for type Day. Such values are
automatically formatted by invoking the method.
Thus:
prints 0 Monday 1 Tuesday.
Println can tell a plain 0 from a 0 of type Day.
So define a String method for your types and they will print
nicely with no more work.
111

Fileds and methods visibility
Go is very different from C++ in the area of visibility. Go's rules:
1) Go has package scope (C++ has file scope).
2) ︎Spelling determines exported/local (pub/priv).
3)︎Structs in the same package have full access to one another's
fields and methods.
4) Local type can export its fields and methods. 
5) No true subclassing, so no notion of "protected".
These simple rules seem to work well in practice.
112

Agenda 2° Day
Agenda
• Pointers
• Collections
• Structs
• Interfaces
• Exercises
113

Interfaces Overview
• So far, all the types we have examined have been
concrete: they implement something.
• There is one more type to consider: the interface type.
• It is completely abstract; it implements nothing.
Instead, it specifies a set of properties an
implementation must provide.
• Interface as a concept is very close to that of Java,
and Java has an interface type, but the "interface
value" concept of Go is novel.
114

Interfaces definition
The word "interface" is a bit overloaded in Go:
1. There is the concept of an interface
2. There is an interface type
3. There are values of that type.
Definition:
An interface is a set of methods.
To turn it around, the methods implemented by a concrete
type such as a struct form the interface of that type.
115

Example
We saw this simple type before:
The interface of type Point has the method:
It's not
because the interface abstracts away the receiver.
We embedded Point in a new type, NamedPoint;
NamedPoint has the same interface.
116

The interface type
An interface type is a specification of an interface, a set
of methods implemented by some other types. Here's
a simple one, with only one method:
This is a definition of the interface implemented by Point,
or in our terminology,
Point implements Abs Interface
Also, NamedPoint and Point3 implement Abs Interface
Methods are written inside the interface declaration.
117

Example
MyFloat implements AbsInterface even though
float64 does not.
(Aside: MyFloat is not a "boxing" of float64; its
representation is identical to float64.)
118

Interface embedding
Interfaces can embed other interfaces and the effect is al-
most the same as if we had written the embedded
interface’s method signatures in the interface that embeds it
119

Interface value
Once a variable is declared with interface type, it may
store any value that implements that interface.
120

In memory
ai is not a pointer! It's a multiword data structure.
At different times it has different value and type:
121

Summary
1) Interfaces define sets of methods. They are pure and
abstract: no implementation, no data fields. Go has a
clear separation between interface and implementation.
2) Interface values are just that: values. They contain any
concrete value that implements all the methods defined in
the interface. That concrete value may or may not be a
pointer.
3) Types implement interfaces just by having methods.
They do not have to declare that they do so. For instance,
every type implements the empty interface, interface{}.
122

Example io.Writer
Here is the actual signature of fmt.Fprintf:
It doesn't write to a file, it writes to something of type io.Writer,
that is, Writer defined in the io package:
Fprintf can therefore be used to write to any type that
has a canonical Write method, including files, pipes,
network connections, and...
123

Example Buffered I/O
... a write buffer. This is from the bufio package:
bufio.Writer implements the canonical Write method.
It also has a factory: give it an io.Writer, it will return a
buffered io.Writer in the form of a bufio.Writer:
And of course, os.File implements Writer too.
124

Putting it together
Buffering works for anything that Writes.
Feels almost like Unix pipes, doesn't it? The
composability is powerful; see crypto packages.
125

Other public interfaces in io
The io package has:
Reader
Writer
ReadWriter
ReadWriteCloser
These are stylized interfaces but obvious: they capture the
functionality of anything implementing the functions listed in
their names.
This is why we can have a buffered I/O package with an
implementation separate from the I/O itself: it both accepts and
provides interface values.
126

Comparison
In C++ terms, an interface type is like a pure abstract
class, specifying the methods but implementing none
of them.
In Java terms, an interface type is much like a Java
interface.
However, in Go there is a major difference: 
A type does not need to declare the interfaces it
implements, nor does it need to inherit from an
interface type. If it has the methods, it implements the
interface.
127

Anonymous fields work too
LockedBufferedWriter implements io.Writer, but also,
through the anonymous Mutex,
128

HTTP service example
This is the interface defined by the HTTP server package. To
serve HTTP, define a type that implements this interface and
connect it to the server (details omitted).
129

A function that serves HTTP
Now we define a type to implement ServeHTTP:
Convert function to attach method, implement the interface:
130

Containers & the empty interface
Sketch of the implementation of vectors. (In practice,
tend to use raw slices instead, but this is informative):
Vectors can contain anything because any type implements the
empty interface. (In fact every element could be of different
type.)
131

Type assertions
Once you put something into a Vector, it's stored as an interface
value. Need to "unbox" it to get the original back: use a "type
assertion". Syntax:
Will fail if type is wrong - but see next slide.
Type assertions always execute at run time. Compiler
rejects assertions guaranteed to fail.
132

Interface to interface conversion
So far we've only moved regular values into and out of
interface values, but interface values that contain the
appropriate methods can also be converted.
In effect, it's the same as unboxing the interface value
to extract the underlying concrete value, then boxing
it again for the new interface type.
The conversion's success depends on the underlying
value, not the original interface type.
133

Interface to interface example
Given:
These are all OK:
134

Testing with type assertions
Can use "comma ok" type assertions to
test a value for type.
135

Testing with a type switch
Special syntax:
136

Does v implement m()?
Going one step further, can test whether a
value implements a method.
This is how Print etc. check if type can print itself.
137

Reflection and …
There is a reflection package (reflect) that builds on these ideas to let you
examine values to discover their
type. Too intricate to describe here but Printf etc. use it to analyze the its
arguments.
Inside Printf, the args variable becomes a slice of the
specified type, i.e. []interface{}, and Printf uses the
reflection package to unpack each element to analyze its
type.
138

Reflection and Print
As a result, Printf and its friends know the actual types of their arguments.
Because they know if the
argument is unsigned or long, there is no %u or %ld, only %d.
This is also how Print and Println can print the arguments nicely without a format
string.
There is also a %v ("value") format that gives default nice output from Printf for
values of any type.
Prints -1 hello [1 2 3] 456. 
In fact, %v is identical to the formatting done by Print
and Println
139

Agenda 2° Day
Agenda
• Pointers
• Collections
• Structs
• Interfaces
• Exercises
140

Exercise
1. swap var
2. sort int array
3. sort string array
4. make append function for slice
5. circular linked list
6. stack
141

Agenda 3° Day
Agenda
• File Handling
• Exercises
142

Goroutines
Terminology:
There are many terms for "things that run concurrently"
- process, thread, POSIX thread, NPTL thread,
lightweight process, ..., but these all mean slightly
different things. None means exactly how Go does
concurrency.
So we introduce a new term: goroutine.
143

Definition
A goroutine is a Go function or method executing
concurrently in the same address space as other
goroutines. A running program consists of one or
more goroutines.
It's not the same as a thread, process, etc. It's a
goroutine.
Note: Concurrency and parallelism are different
concepts. Look them up if you don't understand the
difference.
144

Starting a goroutine
Invoke a function or method and say go:
Prints???
145

Some simple facts
Goroutines are cheap.
Goroutines exit by returning from their top-level
function, or just falling off the end.
Goroutines can run concurrently on different
processors, sharing memory.
You don't have to worry about stack size.
146

Stacks
In gccgo, at least for now, goroutines are pthreads. In 6g
they're multiplexed onto threads, so they are much
cheaper.
In both worlds, stacks are small (a few kB) and grow as
needed.
Thus goroutines use little memory, you can have lots of
them, and they can dynamically have huge stacks.
The programmer shouldn't have to think about the stack
size issue, and in Go, it doesn't even come up.
147

Scheduling
Goroutines are multiplexed as needed onto system threads.
When a goroutine executes a blocking system call, no
other goroutine is blocked.
Plan to do the same for CPU-bound goroutines at some
point, but for now, if you want user - level parallelism in 6g*
you must set shell var:
GOMAXPROCS or call runtime.GOMAXPROCS(n).
GOMAXPROCS tells the runtime scheduler how many user-
space-executing goroutines to run at once, ideally on
different CPU cores.
*gccgo always uses one thread per goroutine.
148

Thread communication (Channels)
Unless two goroutines can communicate, they can't
coordinate.
Go has a type called a channel that provides
communication and synchronization capabilities.
It also has special control structures that build on
channels to make concurrent programming easy.
151

Channel type
In its simplest form the type looks like this:
With a value of this type, you can send and receive items of
elementType.
Channels are a reference type, which means if you assign one
chan variable to another, both variables access the same
channel. It also means you use make to allocate one:
152

The communication operator: <-
The arrow points in the direction of data flow.
As a binary operator, <- sends the value on the right to
the channel on the left:
As a prefix unary operator, <- receives from a channel:
153

Semantics
By default, communication is synchronous. (We'll talk about
asynchronous communication later.) This means:
1) A send operation on a channel blocks until a receiver is
available for the same channel.
2) A receive operation for a channel blocks until a sender is
available for the same channel.
Communication is therefore a form of synchronization: two
goroutines exchanging data through a channel
synchronized at the moment of communication.
154

Let’s pump some data
Now we start a
looping receiver.
You can still sneak in
and grab a value:
155

Functions returning channels
In the previous example, pump was like a generator
spewing out values. But there was a lot of fuss allocating
channels etc. Let's package it up into a function returning the
channel of values.
156

Range & Channels
The range clause on for loops accepts a channel as an operand,
in which case the for loops over the values received from the
channel. We rewrote pump; here's the rewrite for suck, making
it launch the goroutine as well:
157

Closing channel
How does range know when the channel is done
delivering data? The sender calls the built-in
function close: close(ch)
Receiver tests if the sender has closed the channel
using "comma ok":
val, ok := <-ch
 
Result is (value, true) while values are available;
once channel is closed and drained, result is (zero,
false).
158

Range on a channel, by hand
A for range on a channel such as:
is equivalent to:
159

Close
Only the sender should call close. 
Only the receiver can ask if channel has been closed.
Can only ask while getting a value (avoids races).
Call close only when it's necessary to signal to the
receiver that no more values will arrive.
Most of the time, close isn't needed; it's not analogous
to closing a file.
Channels are garbage-collected regardless.
160

Channel directionality 1/2
In its simplest form a channel variable is an unbuffered
(synchronous) value that can be used to send and receive.
A channel type may be annotated to specify that it may only
send or only receive:
161

Channel directionality 2/2
All channels are created bidirectional, but we can assign them to
directional channel variables. Useful for instance in functions,
for (type) safety:
162

Synchronous channel
Output:
sending 10 (happens immediately)
sent 10 (60s later, these 2 lines appear)
received 10
163

Asynchronous channels
A buffered, asynchronous channel is created by telling
make the number of elements in the buffer.
Output:
sending 10 (happens
immediately)
sent 10 (now)
received 10 (60s later)
164

Buffer is not part of the type
Note that the buffer's size, or even its existence, is not
part of the channel's type, only of the value. This code is
therefore legal, although dangerous:
Buffering is a property of the value, not of the type.
165

Select
Select is a control structure in Go analogous to a
communications switch statement. Each case must
be a communication, either send or receive.
Select executes one runnable case at random. If no case is runnable, it blocks
until one is. A
default clause is always runnable.
168

Random bit generator
Prints 0 1 1 0 0 1 1 1 0 1 ...
169

Testing for communicability
Can a communication proceed without blocking? Select
with a default clause can tell us:
The default clause executes if no other case can
proceed, so this is the idiom for non-blocking
receive; non-blocking send is the obvious variant.
170

Timeout
Can a communication succeed in a given interval? Package
time contains the After function:
It delivers a value (the current time) on the returned channel
after the specified interval.
Use it in select to implement timeout:
171

Multiplexing
Channels are first-class values, which means they can
be sent over channels. This property makes it easy to
write a service multiplexer since the client can supply,
along with its request, the channel on which to reply,
Or more typically
172

Starting the server
Use the "function returning channel" idiom to create a
channel to a new server:
174

The client
A similar example is worked in more detail in the tutorial,
but here's a variant:
Requests are ready; send them:
Can retrieve results in any order; r.replyc demuxes:
175

Teardown
In the multiplex example, the server runs forever. To
tear it down cleanly, signal with a channel. This server
has the same functionality but with a quit channel:
176

Starting the server
The rest of the code is mostly the same, with an extra channel:
177

Teardown: the client
The client is unaffected until it's ready to shut down the server:
All done, signal server to exit:
178

Concurrency issues
Many issues, of course, but Go tries to take care of
them. Channel send and receive are atomic. The
select statement is very carefully defined and
implemented, etc.
But goroutines run in shared memory, communication
networks can deadlock, multithreaded debuggers
suck, and so on.
What to do?
179

Go primitives
Don't program the way you would in C or C++ or even
Java.
Channels give you synchronization and communication
both, and that makes them powerful but also easy to
reason about if you use them well.
The rule is:
Do not communicate by sharing memory. Instead,
share memory by communicating.
The very act of communication guarantees
synchronization!
180

Agenda 3° Day
Agenda
• File Handling
• Exercises
181

Go: I/O
• The two main interfaces are Reader and Writer.
• Many functions in Go take Readers or Writers as arguments.
• For example the io package has a Copy function which
copies data from a Reader to a Writer
To read or write to a []byte or a string you can use the Buffer
struct found in the bytes package:
182

Files: basic operations: read 1/2
Open & read file
183

Files: basic operations: read 2/2
There's a shorter way to read a file
184

Files: basic operations: create
Create a file “test.txt” and
write test in content
185

Files: basic operations: read a directory 1/2
Read directory “.” and
get file infos
186

Read directory
recursively using
path/filepath package

Agenda 3° Day
Agenda
• File Handling
• Exercises
188

cgo 1/2
• Cgo enables the creation of Go packages that call C
code.
• To use cgo write normal Go code that imports a
pseudo-package "C".
189

cgo 1/2
package main
/*
#cgo CFLAGS: -I./
#cgo LDFLAGS: -L. -lkeyboard
#include <keyboard.h>
*/
import "C"
import (
"fmt"
)
func main() {
C.InitKeyboard()
fmt.Printf("nEnter: ")
for {
r := C.GetCharacter()
fmt.Printf("%c", r)
if r == 'q' {
break
}
}
C.CloseKeyboard()
}
1. CFLAGS:parameters to the
compiler.
2. ./ folder: compilercan find our
header files.
3. LDFLAGS: parameters to the
linker.
4. -L: (minus capital L) which tells
the linker where it can find
our dynamic library
5. -l: (minus lowercase L) which
tells the linker the name of
our library.
190

Agenda 3° Day
Agenda
• File Handling
• Exercises
191

Exercises
• Read dir
• Analize file extension
• apply regular expression file
content
• multithread mode
192

Agenda 4° Day
Agenda
• Gorilla web toolkit (Go Web Application Toolkit)
• Gorm (Go ORM)
• Exercises
193

Overview
Web Toolkit build with go
• gorilla/context: stores global request variables.
• gorilla/mux: is a powerful URL router and dispatcher.
• gorilla/reverse: produces reversible regular expressions for regexp-based muxes.
• gorilla/rpc: implements RPC over HTTP with codec for JSON-RPC.
• gorilla/schema: converts form values to a struct.
• gorilla/securecookie encodes and decodes authenticated and optionally encrypted
cookie values.
• gorilla/sessions: saves cookie and filesystem sessions and allows custom session
backends.
• gorilla/websocket: implements the WebSocket protocol defined in RFC 6455.
194

Simple Web Application
GorillaToolkitAppTest
196

Agenda 4° Day
Agenda
• Gorilla web toolkit (Go Web Application Toolkit)
• Gorm (Go ORM)
• Exercises
197

Overview
Gorm is ORM (Object Relational Mapper) library for Golang
• Full-Featured ORM (almost)
• Chainable API
• Auto Migrations
• Relations (Has One, Has Many, Belongs To, Many To Many, Polymorphism)
• Callbacks (Before/After Create/Save/Update/Delete/Find)
• Preloading (eager loading)
• Transactions
• Embed Anonymous Struct
• Soft Deletes
• Customizable Logger
• Iteration Support via Rows
• Every feature comes with tests
• Developer Friendly
198

Installation
Gorm toolkit: go get -u github.com/jinzhu/gorm
Postgresql driver: go get github.com/lib/pq
Gorilla web toolkit: go get github.com/gorilla/mux
199

Simple example
GormHelloWorld
200

http://www.golangbootcamp.com/book/collection_types#uid56
202

The GO programming language

Recommandé

Recommandé

Contenu connexe

Tendances

Tendances (19)

Similaire à The GO programming language

Similaire à The GO programming language (20)

Dernier

Dernier (20)

The GO programming language