Inheritance diagram for joshua.decoder.ff.tm.AbstractGrammar:

Collaboration diagram for joshua.decoder.ff.tm.AbstractGrammar:

Public Member Functions
	AbstractGrammar ()
int	getOOVRuleID ()
void	sortGrammar (List< FeatureFunction > models)
boolean	isSorted ()
void	writeGrammarOnDisk (String file)
void	changeGrammarCosts (Map< String, Double > weightTbl, HashMap< String, Integer > featureMap, double[] scores, String prefix, int column, boolean negate)
void	obtainRulesIDTable (Map< String, Integer > rulesIDTable)
Static Public Attributes
static final int	OOV_RULE_ID = 0
Protected Member Functions
void	setSorted (boolean sorted)
Protected Attributes
boolean	sorted
Private Member Functions
void	sort (Trie node, List< FeatureFunction > models)
void	changeGrammarCosts (Trie trie, HashMap< String, Integer > featureMap, double[] scores, String prefix, int column, boolean negate)
void	obtainRulesIDTable (Trie trie, Map< String, Integer > rulesIDTable)
Static Private Attributes
static final Logger	logger = Logger.getLogger(AbstractGrammar.class.getName())

Detailed Description

Partial implementation of the Grammar interface that provides logic for sorting a grammar.

Note: New classes implementing the Grammar interface should probably inherit from this class, unless a specific sorting technique different from that implemented by this class is required.

Author:: Zhifei Li; Lane Schwartz

Constructor & Destructor Documentation

joshua.decoder.ff.tm.AbstractGrammar.AbstractGrammar ( )

Constructs an empty, unsorted grammar.

See also:: Grammar::isSorted()

Member Function Documentation

void joshua.decoder.ff.tm.AbstractGrammar.changeGrammarCosts	(	Map< String, Double >	weightTbl,
		HashMap< String, Integer >	featureMap,
		double[]	scores,
		String	prefix,
		int	column,
		boolean	negate
	)

Implements joshua.decoder.ff.tm.Grammar.

Here is the call graph for this function:

Here is the caller graph for this function:

void joshua.decoder.ff.tm.AbstractGrammar.changeGrammarCosts	(	Trie	trie,
		HashMap< String, Integer >	featureMap,
		double[]	scores,
		String	prefix,
		int	column,
		boolean	negate
	)		`[private]`

Here is the call graph for this function:

int joshua.decoder.ff.tm.AbstractGrammar.getOOVRuleID ( )

Gets the integer identifier of this grammar's out-of-vocabulary (OOV) rule.

Returns:: the integer identifier of this grammar's out-of-vocabulary (OOV) rule

Implements joshua.decoder.ff.tm.Grammar.

Here is the caller graph for this function:

boolean joshua.decoder.ff.tm.AbstractGrammar.isSorted ( )

Determines whether the rules in this grammar have been sorted based on the latest feature function values.

This method is needed for the cube-pruning algorithm.

Returns:: true if the rules in this grammar have been sorted based on the latest feature function values, false otherwise

Implements joshua.decoder.ff.tm.Grammar.

void joshua.decoder.ff.tm.AbstractGrammar.obtainRulesIDTable ( Map< String, Integer > rulesIDTable )

Implements joshua.decoder.ff.tm.Grammar.

Here is the call graph for this function:

Here is the caller graph for this function:

void joshua.decoder.ff.tm.AbstractGrammar.obtainRulesIDTable	(	Trie	trie,
		Map< String, Integer >	rulesIDTable
	)		`[private]`

Here is the call graph for this function:

void joshua.decoder.ff.tm.AbstractGrammar.setSorted ( boolean sorted ) [protected]

Sets the flag indicating whether this grammar is sorted.

This method is called by sortGrammar(ArrayList) to indicate that the grammar has been sorted.

Its scope is protected so that child classes that override sortGrammar will also be able to call this method to indicate that the grammar has been sorted.

Parameters:

sorted

Here is the caller graph for this function:

void joshua.decoder.ff.tm.AbstractGrammar.sort	(	Trie	node,
		List< FeatureFunction >	models
	)		`[private]`

Recursively sorts the grammar using the provided feature functions.

This method first sorts the rules stored at the provided node, then recursively calls itself on the child nodes of the provided node.

Parameters:

node	Grammar node in the `Trie` whose rules should be sorted.
models	Feature function models to use during sorting.

Here is the call graph for this function:

Here is the caller graph for this function:

void joshua.decoder.ff.tm.AbstractGrammar.sortGrammar ( List< FeatureFunction > models )

Cube-pruning requires that the grammar be sorted based on the latest feature functions. To avoid synchronization, this method should be called before multiple threads are initialized for parallel decoding

Implements joshua.decoder.ff.tm.Grammar.

Here is the call graph for this function:

void joshua.decoder.ff.tm.AbstractGrammar.writeGrammarOnDisk ( String file )

Implements joshua.decoder.ff.tm.Grammar.