The [NodePath] built-in [Variant] type represents a path to a node or property in a hierarchy of nodes. It is designed to be efficiently passed into many built-in methods (such as [method Node.get_node], [method Object.set_indexed], [method Tween.tween_property], etc.) without a hard dependence on the node or property they point to.
A node path is represented as a [String] composed of slash-separated ([code]/[/code]) node names and colon-separated ([code]:[/code]) property names (also called "subnames"). Similar to a filesystem path, [code]".."[/code] and [code]"."[/code] are special node names. They refer to the parent node and the current node, respectively.
The following examples are paths relative to the current node:
In some situations, it's possible to omit the leading [code]:[/code] when pointing to an object's property. As an example, this is the case with [method Object.set_indexed] and [method Tween.tween_property], as those methods call [method NodePath.get_as_property_path] under the hood. However, it's generally recommended to keep the [code]:[/code] prefix.
Node paths cannot check whether they are valid and may point to nodes or properties that do not exist. Their meaning depends entirely on the context in which they're used.
You usually do not have to worry about the [NodePath] type, as strings are automatically converted to the type when necessary. There are still times when defining node paths is useful. For example, exported [NodePath] properties allow you to easily select any node within the currently edited scene. They are also automatically updated when moving, renaming or deleting nodes in the scene tree editor. See also [annotation @GDScript.@export_node_path].
[b]Note:[/b] In a boolean context, a [NodePath] will evaluate to [code]false[/code] if it is empty ([code]NodePath("")[/code]). Otherwise, a [NodePath] will always evaluate to [code]true[/code].
[b]Note:[/b] In GDScript, it's also possible to convert a constant string into a node path by prefixing it with [code]^[/code]. [code]^"path/to/node"[/code] is equivalent to [code]NodePath("path/to/node")[/code].
Returns a copy of this node path with a colon character ([code]:[/code]) prefixed, transforming it to a pure property path with no node names (relative to the current node).
Returns the node name indicated by [param idx], starting from 0. If [param idx] is out of bounds, an error is generated. See also [method get_subname_count] and [method get_name_count].
Returns the property name indicated by [param idx], starting from 0. If [param idx] is out of bounds, an error is generated. See also [method get_subname_count].
Returns the 32-bit hash value representing the node path's contents.
[b]Note:[/b] Node paths with equal hash values are [i]not[/i] guaranteed to be the same, as a result of hash collisions. Node paths with different hash values are guaranteed to be different.
Returns [code]true[/code] if the node path is absolute. Unlike a relative path, an absolute path is represented by a leading slash character ([code]/[/code]) and always begins from the [SceneTree]. It can be used to reliably access nodes from the root node (e.g. [code]"/root/Global"[/code] if an autoload named "Global" exists).
Returns the slice of the [NodePath], from [param begin] (inclusive) to [param end] (exclusive), as a new [NodePath].
The absolute value of [param begin] and [param end] will be clamped to the sum of [method get_name_count] and [method get_subname_count], so the default value for [param end] makes it slice to the end of the [NodePath] by default (i.e. [code]path.slice(1)[/code] is a shorthand for [code]path.slice(1, path.get_name_count() + path.get_subname_count())[/code]).
If either [param begin] or [param end] are negative, they will be relative to the end of the [NodePath] (i.e. [code]path.slice(0, -2)[/code] is a shorthand for [code]path.slice(0, path.get_name_count() + path.get_subname_count() - 2)[/code]).
