workbench.cfgParser

Class Parser

java.lang.Object

workbench.cfgParser.Parser

All Implemented Interfaces:

Runnable

Direct Known Subclasses:

AriesParser, AriesTelemetryParser, PhoenixParser, RemusParser, SeahorseParser

public abstract class Parser
extends Object
implements Runnable

Parses an instance of a context free grammar (CFG) contained in a file. Grammars to be parsed should be defined in Chomsky Normal Form (CNF) with a set of ParseFules of the form (A -> BC) where A, B, and C are all variables or A -> b where A is a variable and b is a terminal. This class builds a parse tree using the CK algorithm.

Author:

dtdavis

Field Summary

Fields

Modifier and Type	Field and Description
protected CfgInstance	cfgInstance The CFG object that is to be parsed
protected Dictionary	dictionary Dictionary with the Hashtable matching words to parts of speech
protected int	numRules The number of rules used to define the context free grammar
protected ParseNode	parseTree Tree representing a successful parse of a phrase
protected static ParseRule[]	ruleSet Fixed set of rules that will be used to parse phrases

Constructor Summary

Constructors

Constructor and Description
Parser() Creates a new instance of PhraseParser but does not generate the actual instance
Parser(File inFile) Creates a new instance of Parser
Parser(InputStream inStream) Creates a new instance of Parser
Parser(String fileName) Creates a new instance of Parser

Method Summary

Modifier and Type	Method and Description
protected ArrayList<Integer>	applyRules(int b, int c) Applies all of the rules to a portion of the subphrase
protected abstract void	generateRuleSet() Generates the array of production rules that define the CFG
protected String	getToken(ParseNode node, int n) Returns the starting token of the portion of the phrase covered by the nth generation right child node of this one.
protected ParseNode	parse() Utilizes the CYK algorithm to Generate a parse tree for the phrase
void	run() Implements Runnable Recursively parses all of the potential parse trees in the agenda using a branch and bound search

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

numRules

protected int numRules

The number of rules used to define the context free grammar

cfgInstance

protected CfgInstance cfgInstance

The CFG object that is to be parsed

parseTree

protected ParseNode parseTree

Tree representing a successful parse of a phrase

dictionary

protected Dictionary dictionary

Dictionary with the Hashtable matching words to parts of speech

ruleSet

protected static ParseRule[] ruleSet

Fixed set of rules that will be used to parse phrases

Constructor Detail

Parser

public Parser()

Creates a new instance of PhraseParser but does not generate the actual instance

Parser

public Parser(String fileName)
       throws FileNotFoundException

Creates a new instance of Parser

Parameters:

fileName - path to the file to be parsed

Throws:

FileNotFoundException

Parser

public Parser(File inFile)
       throws FileNotFoundException

Creates a new instance of Parser

Parameters:

inFile - File object from which context free grammar instance is to be read

Throws:

FileNotFoundException

Parser

public Parser(InputStream inStream)

Creates a new instance of Parser

Parameters:

inStream - InputStream from which context free grammar instance is to be read

Method Detail

generateRuleSet

protected abstract void generateRuleSet()

Generates the array of production rules that define the CFG

parse

protected ParseNode parse()

Utilizes the CYK algorithm to Generate a parse tree for the phrase

Returns:

ParseNode

getToken

protected String getToken(ParseNode node,
                          int n)

Returns the starting token of the portion of the phrase covered by the nth generation right child node of this one. Useful in automatically generated CFGs representing space delimited number series whose parse trees are set up so that the left child always refers to a production of the form A -> a where a is a terminal

Parameters:

node - whose successors are being traversed

n - generation of the right child whose left child terminal is required

Returns:

token string

run

public void run()

Implements Runnable Recursively parses all of the potential parse trees in the agenda using a branch and bound search

Specified by:

run in interface Runnable

applyRules

protected ArrayList<Integer> applyRules(int b,
                                        int c)

Applies all of the rules to a portion of the subphrase

Parameters:

b - variable B of a production of the form A -> BC

c - variable C of a production of the form A -> BC

Returns:

the enumerated phrase type that matches the argument array