Softparsmap - Manual
<<< Previous		Next >>>

Tree Parsers

The interface TreeParser is used to parse in and out gene trees. The Newick and Schreiber format is implemented. It is a redundant interface when it is using streams as well as strings, but since both are used often it is worth it. If you want to create your own parser, extend AbstractTreeParser and the abstract tag found in the file def.xml. For more information see section Create a Tag Instance.

The Newick Format

TreeParserNewick is a parser that supports the Newick format. There are three different versions defined, but you can easily define new ones by extending any of them and change their attributes. Following definitions are taken from the def.xml file. Take a look at the Quick Start for Jooc to find out how attributes and heritage work.

    <tree_parser did="abstract" eid="instance::linked, interface"
                  class="softparsmap.AbstractTreeParser"
                  abstract="yes"
                  encoding="@@string:=iso-8859-1"
                  />

    <tree_parser did="newick" eid="abstract"
                  class="softparsmap.TreeParserNewick"
                  instance_holder="singleton"
                  edge_type="@@instance||edge_type::interface:=unknown"
                  markers_begin="@@string:={"
                  markers_end="@@string:=}"
                  recursive_begin="@@string:=("
                  recursive_end="@@string:=)"
                  recursive_child_separator="@@string:=,"
                  marker_edge_value="@@string:={value}"
                  marker_leaf_label="@@string:={label}"
                  template_node_data="@@string:="
                  template_leaf="@@string:={label}"
                  before_body="@@string:="
                  after_body="@@string:=;"
                  />

    <tree_parser did="newick_edge_value" eid="newick"
                  template_node_data="{value}"
                  template_leaf="{label}"
                  />

    <tree_parser did="newick_edge_value_leaves_too" eid="newick"
                  template_node_data="{value}"
                  template_leaf="{label}:{value}"
                  />

The parser named newick parse trees without any edge data "((1, 2), (3, 4))", the newick_edge_value parser parse trees with internal edge data "((1, 2):100.0, (3, 4):95.0)", and the newick_edge_value_leaves_too parser parse trees with internal edge values as well as leaf edge data "((1:100.0, 2:100:0):100.0, (3:100.0, 4:100.0):95.0)". Here is the list of important attributes and what they control.

encoding - provides the parser with an encoding to use when parsing in and out gene tree using streams
edge_type - defines which edge type tag instance that will be used
before_body and after_body - give the possibility to end and begin the output with constant strings
default_node_edge_value and default_leaf_edge_value - define the default edge values in case they are missing
recursive_begin and recursive_end - define the start and the end strings of the recursive part of the Newick format. The strings must be unique in the tree string
recursive_child_separator - defines the child separator of the Newick format. The string must be unique in the tree string
marker_edge_value - defines the marker for the edge value
marker_leaf_label - defines the marker for the label
template_node_data - defines the template for edge data for internal nodes
template_leaf - defines the template for labels and data for leaves

The Schreiber Format

This parser is divided up into a tree structure parser and a node converter. This is done in order to make it more flexible by allowing combinations of these two.

As of now there are three structures supported. The most common one is that of type [[1, 2], [3, [1]]] which parse the tree structure and call the node converter on every internal node and leaf.

    <tree_structure_parser did="schreiber" eid="instance::linked, interface"
                           class="softparsmap.TreeStructureParserSchreiber"
                           instance_holder="singleton"
                           before_core="@@string:=["
                           after_core="@@string:=]"
                           left_node_marker="@@string:=["
                           right_node_marker="@@string:=]"
                           node_divider="@@string:=,"
                           child_node_divider="@@string:=,"
                           />

Combining this structure parser with the node converter

    <node_converter did="schreiber_gene_node_label" 
                    eid="string_template_gene_tree"
                    class="softparsmap.StringGeneNodeConverterSchreiberLabel"
                    edge_type="@@instance||edge_type::interface:=unknown"
                    left_linked_node="@@string:=["
                    right_linked_node="@@string:=]"
                    />

creates the standard Schreiber tree parser

    <tree_parser did="schreiber_gene_label" eid="dual"
                 tree_structure_parser="schreiber"
                 node_converter="schreiber_gene_node_label"
                 />

There are two more structure parsers and six more node converters, creating almost 7*3=21 combinations. Almost, because some combinations is not valid. For more information see the def.xml.

<<< Previous	Home	Next >>>
Gene and Species Trees		In-paralogous