# FUNCTION

# Public

# ADOPT-CHILDREN (NEW-PARENT OLD-PARENT)

HELPER FUINCTION: removes the children of old-parent and
the new-parent takes all of them.

# CONNECT (PARENT CHILD)

connect two nodes as a parent and a child.

# DISCONNECT (PARENT CHILD)

disconnect a parent and its child.
signals CHILD-NOT-FOUND < SIMPLE-CONDITION.

# MAP-RRT-TREE-CONTENT-RECURSIVELY (NODE FN)

Map over the contents of RRT-TREE-NODEs of the tree and
return each result in a nested tree
with the same structure as the original random-tree.

# MAP-RRT-TREE-NODE-RECURSIVELY (NODE FN)

Map over the RRT-TREE-NODEs of the tree and
return the results in a nested cons tree
with the same structure as that of the original random-tree.

# MAPC-RRT-TREE-CONTENT-RECURSIVELY (NODE FN)

Mapc over the contents of RRT-TREE-NODEs ofthe tree and returns nil.
Only for the side effect.

# MAPC-RRT-TREE-NODE-RECURSIVELY (NODE FN)

Mapc over the RRT-TREE-NODEs of the tree and returns nil.
Only for the side effect.

# NEGLECT (PARENT)

HELPER FUNCTION: disconnect all children from the
specified parent

# NNEXT-BRANCH (TREE)

TREE -> TREE
Destructively modifies and return an RRT-TREE. If the
`tree' has a finish node, it finds a path from the root to
the end and then replace the root with the next node in that path.
Otherwise it choose one child of the root at random and replace the
root with it. In both cases the new root is orphanized.

# ORPHANIZE (CHILD)

HELPER FUNCTION: ensure a node doesn't have a parent

# RESULT-PATH (TREE)

TREE -> (list (node V))
Returns a list of C-space points of the computed paths
from the root to the end. Returns nil if the path was not found. The
list contains the root of the tree.

# RESULT-PATH-NODES (TREE)

TREE -> (list V)
Returns the nodes of the computed path in a list, from
the root to the end. Returns nil if the path was not found. The list
contains the root of the tree.

# RRT-NODE (CONTENT)

Identical to =(make-instance 'rrt-tree-node :content content)=

# RRT-SEARCH (RANDOM-GENERATOR NEW-V-GENERATOR EDGE-PROHIBITED-P FINISH-P &KEY START-V TREE (TREE-CLASS 'RRT-TREE-TREE) (MAX-NODES 15) (MAX-ITERATION 30) RUN-ON-NODE)

RRT-search function.
let V as a type variable.
+ V :: a vector class which represents the point in C-space.
(configuration-space-distance V V) should return a number.
+ (node V) :: an rrt-tree-node instance whose `content' slot is V.
non-holonomic parameters like velocity and acceralation
should be stored within (node V), not in V.
`rrt-search' returns the result tree as its primary value. The
secondaly value is the total number of the nodes, and third value is
the number of iteration done in the search.
the arguments should be of types as listed in the following :
+ start-v : V
+ random-generator: (no args) -> V
+ new-v-generator: V, V -> V ; nearest, random -> actual
+ edge-prohibited-p: V, V -> bool ; nearest, new -> result
+ finish-p: V -> bool ; new -> result
+ run-on-node: V, V -> t ; nearest, new ->

# Private

# Undocumented

# %NEAREST-NODE-LIST (TARGET-CONTENT LST)

# GENERIC-FUNCTION

# Public

# CONFIGURATION-SPACE-DISTANCE (POINT1 POINT2)

V, V -> NUMBER
This generic function should provide a method
to measure the distance between two points in C-space
(configuration space).

# COUNT-NODES (TREE)

TREE -> FIXNUM
This generic function should provide a method
to count the number of leafs.

# INSERT (CONTENT TREE)

V, TREE -> TREE
This generic function is allowed provide
an additional procedure during the insertion of a `content' to the `tree'.
The code in this method does something other than the linking of the
parent and child nodes, and optimizes nearest-search. For example,
if you want to use B-tree for the nearest search,
here you can add codes for inserting a content into a B-tree.
Always returns `tree' as a result of :around method of rrt-tree-mixin.

# LEAFS (TREE)

TREE -> (list (node V))
This generic function should provide a method
to accumulate all leafs into a list.
A leaf is a node with no children.

# NEAREST-NODE (TARGET TREE)

V, tree -> (node V), NUMBER, V
This generic function should implement a method
which finds the nearest node in a `tree' to the `target'.
It returns multiple-values.
1. The first return value should be the nearest node.
2. The second value should be the distance between the target
and the nearest node.
3. The third value should be the content of the node.

# SLOT-ACCESSOR

# Public

# FINISH-NODE (OBJECT)

# SETFFINISH-NODE (NEW-VALUE OBJECT)

# Undocumented

# CHILDREN (OBJECT)

# SETFCHILDREN (NEW-VALUE OBJECT)

# CONTENT (OBJECT)

# SETFCONTENT (NEW-VALUE OBJECT)

# NODES (OBJECT)

# SETFNODES (NEW-VALUE OBJECT)

# PARENT (OBJECT)

# SETFPARENT (NEW-VALUE OBJECT)

# ROOT (OBJECT)

# SETFROOT (NEW-VALUE OBJECT)

# CLASS

# Public

# RRT-TREE-LIST

an rrt-tree implementation which uses
a simple linear search method for nearest-search.

# RRT-TREE-MIXIN

The abstract interface mixin class of rrt-tree. User do not create
an instance of it, but rather inherit it. It has three slots and accessors with the same names:
+ root : the root node of the tree. of type rrt-tree-node.
+ finish : a slot which contains the last node of the computed path.
It is bound to nil, it means the previous search has failed to find
a path which satisfies the goal condition.

# RRT-TREE-NODE

Node class of Random Tree.

# RRT-TREE-TREE

an rrt-tree implementation which
seaches from the root in nearest-search.