AT&T Home | AT&T Labs | Research
AT&T Labs, Inc. - Research

The Yoix® Scripting Language

Home | What's New | Grammar | Documentation | Download | License | YChart | YDAT | YWAIT | Byzgraf | FAQs
Node typedict
 
A Node is a type of Element. It becomes a component of a Graph by assigning that Graph value to the node's parent field.

Constants referred to here are contained in the yoix.graph dictionary.

The complete set of fields in a Node are:
attribute([int flags], [Object name[, Object value]]) A Builtin that allows one to add, change, delete or retrieve any type of attribute. The flags argument, when present, specifies the desired action, action scope and the target attribute type. The various values described below can be combined by bitwise-OR to specify the flags argument. The default action is either retrieval or replacement, depending on the context (see below). The following actions can be indicated specifically: CREATE, DELETE, or REPLACE. CREATE is for creating new attributes, but otherwise generates a badvalue error when an existing attribute with the same name is accessible. Using REPLACE avoids this check. A DELETE action is ignored when the requested attribute does not already exist. The default action scope is local, but can be specified as SCOPED, which means that any default attributes that are inherited will be accessible along with the locally defined attributes. Otherwise, only the locally defined attributes will be accessed. By default, the target attributes type are the element's attributes, but specific default attributes to access can be specified using NODEDFLT, EDGEDFLT, or GRAPHDFLT. When working with the local scope, requesting any action (retrieval, creation, deletion or replacement) and targeting element attributes, the attributes defined as part of the current element are accessed (i.e., defaults do not come into play at all). When working with the local scope, requesting any action (retrieval, creation, deletion or replacement) and targeting a default attribute type (graph, node or edge), the default attributes accessed depends on the current element. For a node element, the default attributes owned by its parent are accessed. When scoping and retrieval are in effect and the element is targeted, then first this element's own attributes are accessed, failing that the default node attributes owned by the parent graph are accessed and so on for its parent graph and up to the root graph, and failing all that the global node attributes are accessed, if any. When a specific default type is specified, the search is the same as above except that this element's own attributes are skipped and that the attribute type searched is determined by the flags supplied instead of by the element's type. When scoping and creation, deletion or replacement are in effect and the element attributes are targeted, then the default node attributes owned by the parent graph are accessed. Finally, when a specific default type is specified, the behavior is the same as above except that the attribute type searched is determined by the flags supplied instead of by the element's type.

When no value is supplied, only retrieval or deletion actions are allowed, with retrieval being the default. If a name is also absent, then the deletion action clears all targeted attributes. A supplied name can be a String, Array or Dictionary. When it is a Dictionary, it is treated as supplying names and values and specifically supplying a value argument in that case will cause an error. When a value is supplied, either explicitly or through a name argument Dictionary, then the default action is replacement. Return values are either the requested retrievals, the replaced values or NULL, as appropriate. Multiple values are returned in an Array.

attributes After initialization, this field is read-only. It provides a Dictionary of the local attributes of this element and is equivalent to using the above built-in with no arguments.
bfs([int depth], [Function func], [Object args, ...]) A Builtin for traversing graphs in a breadth-first search manner. Only other node elements are visited starting from the current node element with the connective relationship being determined by the edges between the nodes. If the graph is a directed graph, then to reach node A from node B, there must be an edge directed from A to B existing. The traversal can be restricted to a specified depth or distance out from the current node. If func is supplied, it is called each time an element is traversed and in the context of that element. If one or more args are supplied, they are used as arguments to func. The function need not return a value, but if it does return a non-zero value, it will cause the traversal to stop at that element. The defaults are to have no limit on depth and no function provided. An array of visited elements is returned.
dfs([int depth], [Function func], [Object args, ...]) This Builtin is the same as the bfs built-in, described above, except that the traversal of the graph is in the manner of a depth-first search.
flags An int representing the bitwise-ORing of the basic characteristics or preferences of this node element. Currently, no values are available for nodes.
name A String that is the current name of this element. Assigning to this field changes the name of this element. If name is currently assigned to another element within the overall graph containing this element, a nameclash error results.
parent The Graph which is the parent graph of this element. Assigning to this value adds or removes elements from a graph.
root A read-only field that is the root Graph for this element.
text([int format[, int mode[, int depth[, String prefix]]]]) A Builtin that returns a String representation of the current graph. Current format values are either XML, which results in XML output (see http://www.w3c.org/), or DOT, which results in dot output (see http://www.research.att.com/sw/tools/graphviz/). The default is XML output. Setting mode determines how the graph is traversed and possible values are: BFS, for breadth-first search mode, DFS, for depth-first search mode, and WALK, for a natural walk of the graph. The default is a natural walk. See the walk built-in below for a description of a natural walk. The depth argument limits traversal to the specified number of layers. A negative value, which is the default, indicates no limit. A prefix may be applied to each line of output to allow special marking or indentation of all the text lines. The default is to have no prefix.
walk([Function func], [Object args, ...]) This Builtin traverses the graph in the manner of a natural walk starting at the current element. For a node element, however, that simply means visiting the node itself. If func is supplied, it is called each time an element is traversed and in the context of that element. If one or more args are supplied, they are used as arguments to func. The function need not return a value, but if it does return a non-zero value, it will cause the traversal to stop at that element. An array of visited elements is returned.
Several permanent fields have not been documented and should not be used in Yoix applications.
 
 Example:   The following code sample builds a graph and dumps its XML format and a breadth-first search anchored at node_1 to the standard output:
import yoix.graph.*;
import yoix.stdio.*;

String graph_str = "\
<graph name=g directed=1 strict=1 color=white label=Example>\n\
    <node_attributes color=blue />\n\
    <node name=node_0 color=red />\n\
    <node name=node_1 />\n\
    <node name=node_2 color=green />\n\
    <node name=node_3 />\n\
    <node name=node_4 color=orange />\n\
    <edge name=edge_3_0 tail=node_3 head=node_0 color=yellow />\n\
    <edge name=edge_0_2 tail=node_0 head=node_2 color=green />\n\
    <edge name=edge_2_0 tail=node_2 head=node_0 color=red />\n\
    <edge name=edge_1_2 tail=node_1 head=node_2 color=blue />\n\
    <edge name=edge_4_0 tail=node_4 head=node_0 color=red />\n\
    <edge name=edge_1_4 tail=node_1 head=node_4 color=blue />\n\
</graph>\n\
";

Graph g = xmlGraph(graph_str);

stdout.nextline = g.text();

prnt_name() {
    stdout.nextline = name;
}

Node n = g.element("node_1");
n.bfs(prnt_name);
The result on standard output looks like:
<graph name=g directed=1 strict=1 color=white label=example>
    <node_attributes color=blue />
    <node name=node_0 color=red />
    <node name=node_1 />
    <node name=node_2 color=green />
    <node name=node_3 />
    <node name=node_4 color=orange />
    <edge name=edge_3_0 tail=node_3 head=node_0 color=yellow />
    <edge name=edge_0_2 tail=node_0 head=node_2 color=green />
    <edge name=edge_2_0 tail=node_2 head=node_0 color=red />
    <edge name=edge_1_2 tail=node_1 head=node_2 color=blue />
    <edge name=edge_4_0 tail=node_4 head=node_0 color=red />
    <edge name=edge_1_4 tail=node_1 head=node_4 color=blue />
</graph>

node_1
node_2
node_4
node_0
 
 See Also:   countElements, Edge, Element, Graph, listElements, xmlGraph

 

Yoix is a registered trademark of AT&T Inc.