Closing Go Channels

So far, we have seen basic usages of channels—this lesson presents the definition and the usage of nil channels, signal channels, and buffered channels.

Closing a channel

It helps to remember that the zero value of the channel type is nil, and that if we send a message to a closed channel, the program panics. However, if we try to read from a closed channel, we get the zero value of the type of that channel. So, after closing a channel, we can no longer write to it, but we can still read from it. To be able to close a channel, the channel must not be receive-only.

Additionally, a nil channel always blocks, which means that both reading and writing from nil channels block execution. This property of channels can be very useful when we want to disable a branch of a select statement by assigning the nil value to a channel variable. Finally, if we try to close a nil channel, our program is going to panic.

Coding example

This is best illustrated in the closeNil.go program: