What's the Difference Between Nil, Non-nil, and Empty Slices in Go?

nil vs. Non-nil vs. Empty Slices in Go

In Go, slices are defined by a pointer to an array, length, and capacity. However, there are various types of slices: nil slices, non-nil slices with zero length and capacity, and empty slices.

Nil Slices

Nil slices have no underlying array to point to and thus have a length and capacity of zero. They are essentially non-existent slices that behave as though they have no elements.

Non-nil Slices with Zero Length and Capacity

These slices are initialized with zero length and zero capacity. Unlike nil slices, they have an underlying array to point to, but it's empty. This distinction is important because non-nil slices can potentially grow in size, while nil slices cannot.

Empty Slices

Empty slices are essentially the same as non-nil slices with zero length and capacity. However, the term "empty slice" is often used to emphasize that a specific slice has no elements, regardless of whether it's nil or non-nil.

Observable Behavior

While nil and non-nil slices with zero length and capacity may not share the same internal structure, they exhibit almost identical observable behavior:

• They can be passed to len() and cap() functions for obtaining length and capacity information.
• They can be iterated over with a for range loop, but will have no iterations.
• They can be sliced without violating slice expression restrictions.
• Their content cannot be modified because they have zero length.

Distinguishing Non-nil Empty Slices from Nil Slices

The only reliable way to distinguish between nil and non-nil empty slices is by comparing the slice value to the nil identifier. However, note that some packages (e.g., encoding/json and fmt) may act differently based on whether a slice is nil or not.

Testing if a Slice is Empty

To determine if a slice is empty, compare its length to zero: len(s) == 0. This will return true for both nil and non-nil empty slices.

Understanding the Pointer in Non-nil Empty Slices

In the internal representation of a slice, a pointer points to an underlying array. In the case of non-nil empty slices, this pointer may not be nil. However, it will point to a zero-sized underlying array. The Go specification allows for different types of zero-sized values to have the same memory address.

In summary, while nil slices and non-nil slices with zero length and capacity may appear similar in terms of observable behavior, their underlying structures distinguish them. Non-nil slices have an allocated but empty underlying array, while nil slices have no allocated array.

The above is the detailed content of What's the Difference Between Nil, Non-nil, and Empty Slices in Go?. For more information, please follow other related articles on the PHP Chinese website!