Introduction

This document is a tutorial introduction to the basics of the Go programming language, intended for programmers familiar with C or C++. It is not a comprehensive guide to the language; at the moment the document closest to that is the language specification. After you've read this tutorial, you should look at Effective Go, which digs deeper into how the language is used and talks about the style and idioms of programming in Go. An interactive introduction to Go is available, called A Tour of Go.

The presentation here proceeds through a series of modest programs to illustrate key features of the language. All the programs work (at time of writing) and are checked into the repository in the directory /doc/progs/.

Hello, World

Let's start in the usual way:

Every Go source file declares, using a package statement, which package it's part of. It may also import other packages to use their facilities. This program imports the package fmt to gain access to our old, now capitalized and package-qualified, friend, fmt.Printf.

Functions are introduced with the func keyword. The main package's main function is where the program starts running (after any initialization).

String constants can contain Unicode characters, encoded in UTF-8. (In fact, Go source files are defined to be encoded in UTF-8.)

The comment convention is the same as in C++:

/* ... */
// ...

Later we'll have much more to say about printing.

Semicolons

You might have noticed that our program has no semicolons. In Go code, the only place you typically see semicolons is separating the clauses of for loops and the like; they are not necessary after every statement.

In fact, what happens is that the formal language uses semicolons, much as in C or Java, but they are inserted automatically at the end of every line that looks like the end of a statement. You don't need to type them yourself.

For details about how this is done you can see the language specification, but in practice all you need to know is that you never need to put a semicolon at the end of a line. (You can put them in if you want to write multiple statements per line.) As an extra help, you can also leave out a semicolon immediately before a closing brace.

This approach makes for clean-looking, semicolon-free code. The one surprise is that it's important to put the opening brace of a construct such as an if statement on the same line as the if; if you don't, there are situations that may not compile or may give the wrong result. The language forces the brace style to some extent.

Compiling

Go is a compiled language. At the moment there are two compilers. Gccgo is a Go compiler that uses the GCC back end. There is also a suite of compilers with different (and odd) names for each architecture: 6g for the 64-bit x86, 8g for the 32-bit x86, and more. These compilers run significantly faster but generate less efficient code than gccgo. At the time of writing (late 2009), they also have a more robust run-time system although gccgo is catching up.

Here's how to compile and run our program. With 6g, say,

$ 6g helloworld.go  # compile; object goes into helloworld.6
$ 6l helloworld.6   # link; output goes into 6.out
$ ./6.out
Hello, world; or Καλημέρα κόσμε; or こんにちは 世界
$

With gccgo it looks a little more traditional.

$ gccgo helloworld.go
$ ./a.out
Hello, world; or Καλημέρα κόσμε; or こんにちは 世界
$

Echo

Next up, here's a version of the Unix utility echo(1):

This program is small but it's doing a number of new things. In the last example, we saw func introduce a function. The keywords var, const, and type (not used yet) also introduce declarations, as does import. Notice that we can group declarations of the same sort into parenthesized lists, one item per line, as in the import and const clauses here. But it's not necessary to do so; we could have said

const Space = " "
const Newline = "\n"

This program imports the "os" package to access its Stdout variable, of type *os.File. The import statement is actually a declaration: in its general form, as used in our ``hello world'' program, it names the identifier (fmt) that will be used to access members of the package imported from the file ("fmt"), found in the current directory or in a standard location. In this program, though, we've dropped the explicit name from the imports; by default, packages are imported using the name defined by the imported package, which by convention is of course the file name itself. Our ``hello world'' program could have said just import "fmt".

You can specify your own import names if you want but it's only necessary if you need to resolve a naming conflict.

Given os.Stdout we can use its WriteString method to print the string.

After importing the flag package, we use a var declaration to create and initialize a global variable, called omitNewline, to hold the value of echo's -n flag. The variable has type *bool, pointer to bool.

In main.main, we parse the arguments (the call to flag.Parse) and then create a local string variable with which to build the output.

The declaration statement has the form

var s string = ""

This is the var keyword, followed by the name of the variable, followed by its type, followed by an equals sign and an initial value for the variable.

Go tries to be terse, and this declaration could be shortened. Since the string constant is of type string, we don't have to tell the compiler that. We could write

var s = ""

or we could go even shorter and write the idiom

s := ""

The := operator is used a lot in Go to represent an initializing declaration. There's one in the for clause on the next line:

The flag package has parsed the arguments and left the non-flag arguments in a list that can be iterated over in the obvious way.

The Go for statement differs from that of C in a number of ways. First, it's the only looping construct; there is no while or do. Second, there are no parentheses on the clause, but the braces on the body are mandatory. The same applies to the if and switch statements. Later examples will show some other ways for can be written.

The body of the loop builds up the string s by appending (using +=) the arguments and separating spaces. After the loop, if the -n flag is not set, the program appends a newline. Finally, it writes the result.

Notice that main.main is a niladic function with no return type. It's defined that way. Falling off the end of main.main means ''success''; if you want to signal an erroneous return, call

os.Exit(1)

The os package contains other essentials for getting started; for instance, os.Args is a slice used by the flag package to access the command-line arguments.

An Interlude about Types

Go has some familiar types such as int and uint (unsigned int), which represent values of the ''appropriate'' size for the machine. It also defines explicitly-sized types such as int8, float64, and so on, plus unsigned integer types such as uint, uint32, etc. These are distinct types; even if int and int32 are both 32 bits in size, they are not the same type. There is also a byte synonym for uint8, which is the element type for strings.

Floating-point types are always sized: float32 and float64, plus complex64 (two float32s) and complex128 (two float64s). Complex numbers are outside the scope of this tutorial.

Speaking of string, that's a built-in type as well. Strings are immutable values—they are not just arrays of byte values. Once you've built a string value, you can't change it, although of course you can change a string variable simply by reassigning it. This snippet from strings.go is legal code:

However the following statements are illegal because they would modify a string value:

s[0] = 'x'
(*p)[1] = 'y'

In C++ terms, Go strings are a bit like const strings, while pointers to strings are analogous to const string references.

Yes, there are pointers. However, Go simplifies their use a little; read on.

Arrays are declared like this:

var arrayOfInt [10]int

Arrays, like strings, are values, but they are mutable. This differs from C, in which arrayOfInt would be usable as a pointer to int. In Go, since arrays are values, it's meaningful (and useful) to talk about pointers to arrays.

The size of the array is part of its type; however, one can declare a slice variable to hold a reference to any array, of any size, with the same element type. A slice expression has the form a[low : high], representing the internal array indexed from low through high-1; the resulting slice is indexed from 0 through high-low-1. In short, slices look a lot like arrays but with no explicit size ([] vs. [10]) and they reference a segment of an underlying, usually anonymous, regular array. Multiple slices can share data if they represent pieces of the same array; multiple arrays can never share data.

Slices are much more common in Go programs than regular arrays; they're more flexible, have reference semantics, and are efficient. What they lack is the precise control of storage layout of a regular array; if you want to have a hundred elements of an array stored within your structure, you should use a regular array. To create one, use a compound value constructor—an expression formed from a type followed by a brace-bounded expression like this:

[3]int{1,2,3}

In this case the constructor builds an array of 3 ints.

When passing an array to a function, you almost always want to declare the formal parameter to be a slice. When you call the function, slice the array to create (efficiently) a slice reference and pass that. By default, the lower and upper bounds of a slice match the ends of the existing object, so the concise notation [:] will slice the whole array.

Using slices one can write this function (from sum.go):

{{code "progs/sum.go" `/sum/` `/^}/`}}

Note how the return type (int) is defined for sum by stating it after the parameter list.

To call the function, we slice the array. This code (we'll show a simpler way in a moment) constructs an array and slices it:

x := [3]int{1,2,3}
s := sum(x[:])

If you are creating a regular array but want the compiler to count the elements for you, use ... as the array size:

x := [...]int{1,2,3}
s := sum(x[:])

That's fussier than necessary, though. In practice, unless you're meticulous about storage layout within a data structure, a slice itself—using empty brackets with no size—is all you need:

s := sum([]int{1,2,3})

There are also maps, which you can initialize like this:

m := map[string]int{"one":1 , "two":2}

The built-in function len, which returns number of elements, makes its first appearance in sum. It works on strings, arrays, slices, maps, and channels.

By the way, another thing that works on strings, arrays, slices, maps and channels is the range clause on for loops. Instead of writing

for i := 0; i < len(a); i++ { ... }

to loop over the elements of a slice (or map or ...) , we could write

for i, v := range a { ... }

This assigns i to the index and v to the value of the successive elements of the target of the range. See Effective Go for more examples of its use.

An Interlude about Allocation

Most types in Go are values. If you have an int or a struct or an array, assignment copies the contents of the object. To allocate a new variable, use the built-in function new, which returns a pointer to the allocated storage.

type T struct { a, b int }
var t *T = new(T)

or the more idiomatic

t := new(T)

Some types—maps, slices, and channels (see below)—have reference semantics. If you're holding a slice or a map and you modify its contents, other variables referencing the same underlying data will see the modification. For these three types you want to use the built-in function make:

m := make(map[string]int)

This statement initializes a new map ready to store entries. If you just declare the map, as in

var m map[string]int

it creates a nil reference that cannot hold anything. To use the map, you must first initialize the reference using make or by assignment from an existing map.

Note that new(T) returns type *T while make(T) returns type T. If you (mistakenly) allocate a reference object with new rather than make, you receive a pointer to a nil reference, equivalent to declaring an uninitialized variable and taking its address.

An Interlude about Constants

Although integers come in lots of sizes in Go, integer constants do not. There are no constants like 0LL or 0x0UL. Instead, integer constants are evaluated as large-precision values that can overflow only when they are assigned to an integer variable with too little precision to represent the value.

const hardEight = (1 << 100) >> 97  // legal

There are nuances that deserve redirection to the legalese of the language specification but here are some illustrative examples:

var a uint64 = 0  // a has type uint64, value 0
a := uint64(0)    // equivalent; uses a "conversion"
i := 0x1234       // i gets default type: int
var j int = 1e6   // legal - 1000000 is representable in an int
x := 1.5          // a float64, the default type for floating constants
i3div2 := 3/2     // integer division - result is 1
f3div2 := 3./2.   // floating-point division - result is 1.5

Conversions only work for simple cases such as converting ints of one sign or size to another and between integers and floating-point numbers, plus a couple of other instances outside the scope of a tutorial. There are no automatic numeric conversions of any kind in Go, other than that of making constants have concrete size and type when assigned to a variable.

An I/O Package

Next we'll look at a simple package for doing Unix file I/O with an open/close/read/write interface. Here's the start of file.go:

{{code "progs/file.go" `/package/` `/^}/`}}

The first few lines declare the name of the package—file—and then import two packages. The os package hides the differences between various operating systems to give a consistent view of files and so on; here we're going to use its error handling utilities and reproduce the rudiments of its file I/O.

The other item is the low-level, external syscall package, which provides a primitive interface to the underlying operating system's calls. The syscall package is very system-dependent, and the way it's used here works only on Unix-like systems, but the general ideas explored here apply broadly. (A Windows version is available in file_windows.go.)

Next is a type definition: the type keyword introduces a type declaration, in this case a data structure called File. To make things a little more interesting, our File includes the name of the file that the file descriptor refers to.

Because File starts with a capital letter, the type is available outside the package, that is, by users of the package. In Go the rule about visibility of information is simple: if a name (of a top-level type, function, method, constant or variable, or of a structure field or method) is capitalized, users of the package may see it. Otherwise, the name and hence the thing being named is visible only inside the package in which it is declared. This is more than a convention; the rule is enforced by the compiler. In Go, the term for publicly visible names is ''exported''.

In the case of File, all its fields are lower case and so invisible to users, but we will soon give it some exported, upper-case methods.

First, though, here is a factory to create a File:

{{code "progs/file.go" `/newFile/` `/^}/`}}

This returns a pointer to a new File structure with the file descriptor and name filled in. This code uses Go's notion of a ''composite literal'', analogous to the ones used to build maps and arrays, to construct a new heap-allocated object. We could write

n := new(File)
n.fd = fd
n.name = name
return n

but for simple structures like File it's easier to return the address of a composite literal, as is done here in the return statement from newFile.

We can use the factory to construct some familiar, exported variables of type *File:

The newFile function was not exported because it's internal. The proper, exported factory to use is OpenFile (we'll explain that name in a moment):

{{code "progs/file.go" `/func.OpenFile/` `/^}/`}}

There are a number of new things in these few lines. First, OpenFile returns multiple values, a File and an error (more about errors in a moment). We declare the multi-value return as a parenthesized list of declarations; syntactically they look just like a second parameter list. The function syscall.Open also has a multi-value return, which we can grab with the multi-variable declaration on the first line; it declares r and e to hold the two values, both of type int (although you'd have to look at the syscall package to see that). Finally, OpenFile returns two values: a pointer to the new File and the error. If syscall.Open fails, the file descriptor r will be negative and newFile will return nil.

About those errors: The Go language includes a general notion of an error: a pre-defined type error with properties (described below) that make it a good basis for representing and handling errors. It's a good idea to use its facility in your own interfaces, as we do here, for consistent error handling throughout Go code. In Open we use a conversion to translate Unix's integer errno value into the integer type os.Errno, which is an implementation of error

Why OpenFile and not Open? To mimic Go's os package, which our exercise is emulating. The os package takes the opportunity to make the two commonest cases - open for read and create for write - the simplest, just Open and Create. OpenFile is the general case, analogous to the Unix system call Open. Here is the implementation of our Open and Create; they're trivial wrappers that eliminate common errors by capturing the tricky standard arguments to open and, especially, to create a file:

{{code "progs/file.go" `/^const/` `/^}/`}}

{{code "progs/file.go" `/func.Create/` `/^}/`}}

Back to our main story. Now that we can build Files, we can write methods for them. To declare a method of a type, we define a function to have an explicit receiver of that type, placed in parentheses before the function name. Here are some methods for *File, each of which declares a receiver variable file.

There is no implicit this and the receiver variable must be used to access members of the structure. Methods are not declared within the struct declaration itself. The struct declaration defines only data members. In fact, methods can be created for almost any type you name, such as an integer or array, not just for structs. We'll see an example with arrays later.

The String method is so called because of a printing convention we'll describe later.

The methods use the public variable os.ErrInvalid to return the (error version of the) Unix error code EINVAL. The os library defines a standard set of such error values.

We can now use our new package:

The ''./'' in the import of ''./file'' tells the compiler to use our own package rather than something from the directory of installed packages. (Also, ''file.go'' must be compiled before we can import the package.)

Now we can compile and run the program. On Unix, this would be the result:

$ 6g file.go                       # compile file package
$ 6g helloworld3.go                # compile main package
$ 6l -o helloworld3 helloworld3.6  # link - no need to mention "file"
$ ./helloworld3
hello, world
can't open file; err=No such file or directory
$

Rotting cats

Building on the file package, here's a simple version of the Unix utility cat(1), progs/cat.go:

By now this should be easy to follow, but the switch statement introduces some new features. Like a for loop, an if or switch can include an initialization statement. The switch statement in cat uses one to create variables nr and er to hold the return values from the call to f.Read. (The if a few lines later has the same idea.) The switch statement is general: it evaluates the cases from top to bottom looking for the first case that matches the value; the case expressions don't need to be constants or even integers, as long as they all have the same type.

Since the switch value is just true, we could leave it off—as is also the situation in a for statement, a missing value means true. In fact, such a switch is a form of if-else chain. While we're here, it should be mentioned that in switch statements each case has an implicit break.

The argument to file.Stdout.Write is created by slicing the array buf. Slices provide the standard Go way to handle I/O buffers.

Now let's make a variant of cat that optionally does rot13 on its input. It's easy to do by just processing the bytes, but instead we will exploit Go's notion of an interface.

The cat subroutine uses only two methods of f: Read and String, so let's start by defining an interface that has exactly those two methods. Here is code from progs/cat_rot13.go:

{{code "progs/cat_rot13.go" `/type.reader/` `/^}/`}}

Any type that has the two methods of reader—regardless of whatever other methods the type may also have—is said to implement the interface. Since file.File implements these methods, it implements the reader interface. We could tweak the cat subroutine to accept a reader instead of a *file.File and it would work just fine, but let's embellish a little first by writing a second type that implements reader, one that wraps an existing reader and does rot13 on the data. To do this, we just define the type and implement the methods and with no other bookkeeping, we have a second implementation of the reader interface.

{{code "progs/cat_rot13.go" `/type.rotate13/` `/end.of.rotate13/`}}

(The rot13 function called in Read is trivial and not worth reproducing here.)

To use the new feature, we define a flag:

and use it from within a mostly unchanged cat function:

{{code "progs/cat_rot13.go" `/func.cat/` `/^}/`}}

(We could also do the wrapping in main and leave cat mostly alone, except for changing the type of the argument; consider that an exercise.) The if at the top of cat sets it all up: If the rot13 flag is true, wrap the reader we received into a rotate13 and proceed. Note that the interface variables are values, not pointers: the argument is of type reader, not *reader, even though under the covers it holds a pointer to a struct.

Here it is in action:

$ echo abcdefghijklmnopqrstuvwxyz | ./cat
abcdefghijklmnopqrstuvwxyz
$ echo abcdefghijklmnopqrstuvwxyz | ./cat --rot13
nopqrstuvwxyzabcdefghijklm
$

Fans of dependency injection may take cheer from how easily interfaces allow us to substitute the implementation of a file descriptor.

Interfaces are a distinctive feature of Go. An interface is implemented by a type if the type implements all the methods declared in the interface. This means that a type may implement an arbitrary number of different interfaces. There is no type hierarchy; things can be much more ad hoc, as we saw with rot13. The type file.File implements reader; it could also implement a writer, or any other interface built from its methods that fits the current situation. Consider the empty interface

type Empty interface {}

Every type implements the empty interface, which makes it useful for things like containers.

Sorting

Interfaces provide a simple form of polymorphism. They completely separate the definition of what an object does from how it does it, allowing distinct implementations to be represented at different times by the same interface variable.

As an example, consider this simple sort algorithm taken from progs/sort.go:

{{code "progs/sort.go" `/func.Sort/` `/^}/`}}

The code needs only three methods, which we wrap into sort's Interface:

{{code "progs/sort.go" `/interface/` `/^}/`}}

We can apply Sort to any type that implements Len, Less, and Swap. The sort package includes the necessary methods to allow sorting of arrays of integers, strings, etc.; here's the code for arrays of int

Here we see methods defined for non-struct types. You can define methods for any type you define and name in your package.

And now a routine to test it out, from progs/sortmain.go. This uses a function in the sort package, omitted here for brevity, to test that the result is sorted.

{{code "progs/sortmain.go" `/func.ints/` `/^}/`}}

If we have a new type we want to be able to sort, all we need to do is to implement the three methods for that type, like this:

Printing

The examples of formatted printing so far have been modest. In this section we'll talk about how formatted I/O can be done well in Go.

We've seen simple uses of the package fmt, which implements Printf, Fprintf, and so on. Within the fmt package, Printf is declared with this signature:

Printf(format string, v ...interface{}) (n int, errno error)

The token ... introduces a variable-length argument list that in C would be handled using the stdarg.h macros. In Go, variadic functions are passed a slice of the arguments of the specified type. In Printf's case, the declaration says ...interface{} so the actual type is a slice of empty interface values, []interface{}. Printf can examine the arguments by iterating over the slice and, for each element, using a type switch or the reflection library to interpret the value. It's off topic here but such run-time type analysis helps explain some of the nice properties of Go's Printf, due to the ability of Printf to discover the type of its arguments dynamically.

For example, in C each format must correspond to the type of its argument. It's easier in many cases in Go. Instead of %llud you can just say %d; Printf knows the size and signedness of the integer and can do the right thing for you. The snippet

prints

18446744073709551615 -1

In fact, if you're lazy the format %v will print, in a simple appropriate style, any value, even an array or structure. The output of

18446744073709551615 {77 Sunset Strip} [1 2 3 4]

You can drop the formatting altogether if you use Print or Println instead of Printf. Those routines do fully automatic formatting. The Print function just prints its elements out using the equivalent of %v while Println inserts spaces between arguments and adds a newline. The output of each of these two lines is identical to that of the Printf call above.

If you have your own type you'd like Printf or Print to format, just give it a String method that returns a string. The print routines will examine the value to inquire whether it implements the method and if so, use it rather than some other formatting. Here's a simple example.

Since *testType has a String method, the default formatter for that type will use it and produce the output

77 Sunset Strip

Observe that the String method calls Sprint (the obvious Go variant that returns a string) to do its formatting; special formatters can use the fmt library recursively.

Another feature of Printf is that the format %T will print a string representation of the type of a value, which can be handy when debugging polymorphic code.

It's possible to write full custom print formats with flags and precisions and such, but that's getting a little off the main thread so we'll leave it as an exploration exercise.

You might ask, though, how Printf can tell whether a type implements the String method. Actually what it does is ask if the value can be converted to an interface variable that implements the method. Schematically, given a value v, it does this:

type Stringer interface {
    String() string
}

s, ok := v.(Stringer)  // Test whether v implements "String()"
if ok {
    result = s.String()
} else {
    result = defaultOutput(v)
}

The code uses a ``type assertion'' (v.(Stringer)) to test if the value stored in v satisfies the Stringer interface; if it does, s will become an interface variable implementing the method and ok will be true. We then use the interface variable to call the method. (The ''comma, ok'' pattern is a Go idiom used to test the success of operations such as type conversion, map update, communications, and so on, although this is the only appearance in this tutorial.) If the value does not satisfy the interface, ok will be false.

In this snippet the name Stringer follows the convention that we add ''[e]r'' to interfaces describing simple method sets like this.

A related interface is that defined by the error builtin type, which is just

type error interface {
    Error() string
}

Other than the method name (Error vs. String), this looks like a Stringer; the different name guarantees that types that implement Stringer don't accidentally satisfy the error interface. Naturally, Printf and its relatives recognize the error interface, just as they do Stringer, so it's trivial to print an error as a string.

One last wrinkle. To complete the suite, besides Printf etc. and Sprintf etc., there are also Fprintf etc. Unlike in C, Fprintf's first argument is not a file. Instead, it is a variable of type io.Writer, which is an interface type defined in the io library:

type Writer interface {
    Write(p []byte) (n int, err error)
}

(This interface is another conventional name, this time for Write; there are also io.Reader, io.ReadWriter, and so on.) Thus you can call Fprintf on any type that implements a standard Write method, not just files but also network channels, buffers, whatever you want.

Prime numbers

Now we come to processes and communication—concurrent programming. It's a big subject so to be brief we assume some familiarity with the topic.

A classic program in the style is a prime sieve. (The sieve of Eratosthenes is computationally more efficient than the algorithm presented here, but we are more interested in concurrency than algorithmics at the moment.) It works by taking a stream of all the natural numbers and introducing a sequence of filters, one for each prime, to winnow the multiples of that prime. At each step we have a sequence of filters of the primes so far, and the next number to pop out is the next prime, which triggers the creation of the next filter in the chain.

Here's a flow diagram; each box represents a filter element whose creation is triggered by the first number that flowed from the elements before it.

To create a stream of integers, we use a Go channel, which, borrowing from CSP's descendants, represents a communications channel that can connect two concurrent computations. In Go, channel variables are references to a run-time object that coordinates the communication; as with maps and slices, use make to create a new channel.

Here is the first function in progs/sieve.go:

{{code "progs/sieve.go" `/Send/` `/^}/`}}

The generate function sends the sequence 2, 3, 4, 5, ... to its argument channel, ch, using the binary communications operator <-. Channel operations block, so if there's no recipient for the value on ch, the send operation will wait until one becomes available.

The filter function has three arguments: an input channel, an output channel, and a prime number. It copies values from the input to the output, discarding anything divisible by the prime. The unary communications operator <- (receive) retrieves the next value on the channel.

{{code "progs/sieve.go" `/Copy.the/` `/^}/`}}

The generator and filters execute concurrently. Go has its own model of process/threads/light-weight processes/coroutines, so to avoid notational confusion we call concurrently executing computations in Go goroutines. To start a goroutine, invoke the function, prefixing the call with the keyword go; this starts the function running in parallel with the current computation but in the same address space:

go sum(hugeArray) // calculate sum in the background

If you want to know when the calculation is done, pass a channel on which it can report back:

ch := make(chan int)
go sum(hugeArray, ch)
// ... do something else for a while
result := <-ch  // wait for, and retrieve, result

Back to our prime sieve. Here's how the sieve pipeline is stitched together:

{{code "progs/sieve.go" `/func.main/` `/^}/`}}

The first line of main creates the initial channel to pass to generate, which it then starts up. As each prime pops out of the channel, a new filter is added to the pipeline and its output becomes the new value of ch.

The sieve program can be tweaked to use a pattern common in this style of programming. Here is a variant version of generate, from progs/sieve1.go:

{{code "progs/sieve1.go" `/func.generate/` `/^}/`}}

This version does all the setup internally. It creates the output channel, launches a goroutine running a function literal, and returns the channel to the caller. It is a factory for concurrent execution, starting the goroutine and returning its connection.

The function literal notation used in the go statement allows us to construct an anonymous function and invoke it on the spot. Notice that the local variable ch is available to the function literal and lives on even after generate returns.

The same change can be made to filter:

{{code "progs/sieve1.go" `/func.filter/` `/^}/`}}

The sieve function's main loop becomes simpler and clearer as a result, and while we're at it let's turn it into a factory too:

{{code "progs/sieve1.go" `/func.sieve/` `/^}/`}}

Now main's interface to the prime sieve is a channel of primes:

{{code "progs/sieve1.go" `/func.main/` `/^}/`}}

Multiplexing

With channels, it's possible to serve multiple independent client goroutines without writing an explicit multiplexer. The trick is to send the server a channel in the message, which it will then use to reply to the original sender. A realistic client-server program is a lot of code, so here is a very simple substitute to illustrate the idea. It starts by defining a request type, which embeds a channel that will be used for the reply.

{{code "progs/server.go" `/type.request/` `/^}/`}}

The server will be trivial: it will do simple binary operations on integers. Here's the code that invokes the operation and responds to the request:

{{code "progs/server.go" `/type.binOp/` `/^}/`}}

The type declaration makes binOp represent a function taking two integers and returning a third.

The server routine loops forever, receiving requests and, to avoid blocking due to a long-running operation, starting a goroutine to do the actual work.

{{code "progs/server.go" `/func.server/` `/^}/`}}

There's a new feature in the signature of server: the type of the service channel specifies the direction of communication. A channel of plain chan type can be used both for sending and receiving. However, the type used when declaring a channel can be decorated with an arrow to indicate that the channel can be used only to send (chan<-) or to receive (<-chan) data. The arrow points towards or away from the chan to indicate whether data flows into or out of the channel. In the server function, service <-chan *request is a "receive only" channel that the function can use only to read new requests.

We instantiate a server in a familiar way, starting it and returning a channel connected to it:

{{code "progs/server.go" `/func.startServer/` `/^}/`}}

The returned channel is send only, even though the channel was created bidirectionally. The read end is passed to server, while the send end is returned to the caller of startServer, so the two halves of the channel are distinguished, just as we did in startServer.

Bidirectional channels can be assigned to unidirectional channels but not the other way around, so if you annotate your channel directions when you declare them, such as in function signatures, the type system can help you set up and use channels correctly. Note that it's pointless to make unidirectional channels, since you can't use them to communicate. Their purpose is served by variables assigned from bidirectional channels to distinguish the input and output halves.

Here's a simple test. It starts a server with an addition operator and sends out N requests without waiting for the replies. Only after all the requests are sent does it check the results.

{{code "progs/server.go" `/func.main/` `/^}/`}}

One annoyance with this program is that it doesn't shut down the server cleanly; when main returns there are a number of lingering goroutines blocked on communication. To solve this, we can provide a second, quit channel to the server:

{{code "progs/server1.go" `/func.startServer/` `/^}/`}}

It passes the quit channel to the server function, which uses it like this:

{{code "progs/server1.go" `/func.server/` `/^}/`}}

Inside server, the select statement chooses which of the multiple communications listed by its cases can proceed. If all are blocked, it waits until one can proceed; if multiple can proceed, it chooses one at random. In this instance, the select allows the server to honor requests until it receives a quit message, at which point it returns, terminating its execution.

All that's left is to strobe the quit channel at the end of main:

{{code "progs/server1.go" `/adder,.quit/`}} ... {{code "progs/server1.go" `/quit....true/`}}

There's a lot more to Go programming and concurrent programming in general but this quick tour should give you some of the basics.