net.sf.jlinkgrammar

Class Linkage

java.lang.Object

net.sf.jlinkgrammar.Linkage

public class Linkage
extends java.lang.Object

The linkage routines are primarily used to walk linkages and retrieve information about the sentence. You must fist call dictionary_create, sentence,_create, sentence_parse, in that order. Then you can process links.

See Also:

Linkage(int, Sentence, ParseOptions)

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
(package private) static class	Linkage.LinkageAndList

Field Summary

Fields

Modifier and Type	Field and Description
static Linkage.LinkageAndList[]	andlist
int	current Allows user to select particular sublinkage
LinkageInfo	info index and cost information
int	num_sublinkages One for thin linkages, bigger for fat
ParseOptions	opts
static char[][]	picture
Sentence	sent
Sublinkage[]	sublinkage A parse with conjunctions will have several
boolean	unionized if true, union of links has been computed
java.util.ArrayList<java.lang.String>	word array of word spellings
static int[]	wordtype
static char[][]	xpicture

Constructor Summary

Constructors

Constructor and Description
Linkage(int k, Sentence sent, ParseOptions opts)

Method Summary

Methods

Modifier and Type	Method and Description
void	adjust_subordinate_clauses(int numcon_total, int numcon_subl)
java.lang.String	build_linkage_postscript_string() builds a postscript printable string of the current sublinkage
(package private) void	compute_chosen_words(Sentence sent)
void	extract_fat_linkage(Sentence sent, ParseOptions opts) The old C comment said -- This procedure mimics analyze_fat_linkage in order to extract the sublinkages and copy them to the Linkage data structure passed in.
(package private) static void	extract_links(int index, int cost, ParseInfo pi) Generate the list of all links of the indexth parsing of the sentence.
(package private) void	extract_thin_linkage(Sentence sent, ParseOptions opts) extract thin links from a sentence
java.lang.String	header(int mode) Supplies the postscript header used when generating a postscript output file
(package private) boolean	link_already_appears(Link link, int a)
int	linkage_and_cost()
int	linkage_compute_union() computes the union of the current sublinkages.
CNode	linkage_constituent_tree() CNode linkage_constituent_tree(); The other useful method for dealing with a cosntituent tree is: String linkage_print_constituent_tree(int mode);
int	linkage_disjunct_cost()
java.lang.String[]	linkage_get_link_domain_names(int index)
java.lang.String	linkage_get_link_label(int index)
java.lang.String	linkage_get_link_llabel(int index)
int	linkage_get_link_lword(int index)
int	linkage_get_link_num_domains(int index)
java.lang.String	linkage_get_link_rlabel(int index)
int	linkage_get_link_rword(int index)
int	linkage_get_num_links() This is need to set up the for loop to walk down a set of links for printing or other processing returns the size of the linkage array as the number of links
int	linkage_get_num_sublinkages() This is a toplevel routine used to setup a for loop to process each sublinkage (usually an AND disjunct).
int	linkage_get_num_words() Gives the number or word spellings or individual word spellings for the current sublinkage
Sentence	linkage_get_sentence() getter for the sentence from an instantiated linkage
java.lang.String	linkage_get_violation_name()
java.lang.String	linkage_get_word(int w)
boolean	linkage_has_inconsistent_domains()
boolean	linkage_is_canonical()
boolean	linkage_is_improper()
int	linkage_link_cost()
void	linkage_post_process(Postprocessor postprocessor) post process the sublinkages according to the named postprocessor
java.lang.String	linkage_print_constituent_tree(int mode) String linkage_print_constituent_tree(int mode)
java.lang.String	linkage_print_diagram()
java.lang.String	linkage_print_links_and_domains()
java.lang.String	linkage_print_postscript(int mode) Print out a postscript version of the constituent tree
boolean	linkage_set_current_sublinkage(int index) setter function to set the current sublinkage to indicated index TODO - does this function have to insure a compact array? If so should it return the next available index?
int	linkage_unused_word_cost()
(package private) boolean	links_are_equal(Link l, Link m)
CNode	parse_string(CNode n, java.util.StringTokenizer tok)
void	print_a_link(java.lang.StringBuffer s, int link) Prints the information contained in a link into a supplied string buffer.
void	process_linkage(ParseOptions opts)
void	set_centers(boolean print_word_0)
void	setParseOptions(ParseOptions opts) setter for object's copy of the ParseOptions
void	setSentence(Sentence sent) setter for object's copy of the sentence
(package private) int	token_type(java.lang.String token)
java.lang.String	trailer(int mode)
(package private) Sublinkage	unionize_linkage()
static boolean	uppercompare(java.lang.String s, java.lang.String t)

Methods inherited from class java.lang.Object

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

Field Detail

word

public java.util.ArrayList<java.lang.String> word

array of word spellings

info

public LinkageInfo info

index and cost information

num_sublinkages

public int num_sublinkages

One for thin linkages, bigger for fat

current

public int current

Allows user to select particular sublinkage

sublinkage

public Sublinkage[] sublinkage

A parse with conjunctions will have several

unionized

public boolean unionized

if true, union of links has been computed

sent

public Sentence sent

opts

public ParseOptions opts

andlist

public static Linkage.LinkageAndList[] andlist

wordtype

public static int[] wordtype

picture

public static char[][] picture

xpicture

public static char[][] xpicture

Constructor Detail

Linkage

public Linkage(int k,
       Sentence sent,
       ParseOptions opts)

Method Detail

setSentence

public void setSentence(Sentence sent)

setter for object's copy of the sentence

Parameters:

sent -

setParseOptions

public void setParseOptions(ParseOptions opts)

setter for object's copy of the ParseOptions

Parameters:

opts -

compute_chosen_words

void compute_chosen_words(Sentence sent)

extract_links

static void extract_links(int index,
                 int cost,
                 ParseInfo pi)

Generate the list of all links of the indexth parsing of the sentence. For this to work, you must have already called parse, and already built the whole_set.

extract_thin_linkage

void extract_thin_linkage(Sentence sent,
                        ParseOptions opts)

extract thin links from a sentence

Parameters:

sent - sentence to extract links from. TODO - does not set the current Linkage objects this.sent to sent. This may be a bug!

opts - options to use when processing TODO - does not set the current Linkage objects this.opts to opts. This may be a bug!

extract_fat_linkage

public void extract_fat_linkage(Sentence sent,
                       ParseOptions opts)

The old C comment said -- This procedure mimics analyze_fat_linkage in order to extract the sublinkages and copy them to the Linkage data structure passed in. BUT no linkage structure is passed in the Linkage structure is the instance variable of this Linkage.

Parameters:

sent - sentence to extract links from. TODO - does not set the current Linkage objects this.sent to sent. This may be a bug!

opts - options to use when processing TODO - does not set the current Linkage objects this.opts to opts. This may be a bug!

linkage_unused_word_cost

public int linkage_unused_word_cost()

Returns:

an integer value of the unused word cost

linkage_disjunct_cost

public int linkage_disjunct_cost()

Returns:

an integer value of the disjunct cost

linkage_and_cost

public int linkage_and_cost()

Returns:

the cos of And's

linkage_link_cost

public int linkage_link_cost()

Returns:

the link cost

linkage_get_num_sublinkages

public int linkage_get_num_sublinkages()

This is a toplevel routine used to setup a for loop to process each sublinkage (usually an AND disjunct).

Returns:

the number of sublinkages

linkage_get_num_words

public int linkage_get_num_words()

Gives the number or word spellings or individual word spellings for the current sublinkage

Returns:

integer number of spellings

linkage_get_num_links

public int linkage_get_num_links()

This is need to set up the for loop to walk down a set of links for printing or other processing returns the size of the linkage array as the number of links

Returns:

the size of the linkage array as the number of links

linkage_set_current_sublinkage

public boolean linkage_set_current_sublinkage(int index)

setter function to set the current sublinkage to indicated index TODO - does this function have to insure a compact array? If so should it return the next available index?

Returns:

false if the likage index is already in use, true otherwise

linkage_get_sentence

public Sentence linkage_get_sentence()

getter for the sentence from an instantiated linkage

Returns:

the Sentence instance object held in this Linkage instance object

links_are_equal

boolean links_are_equal(Link l,
                      Link m)

link_already_appears

boolean link_already_appears(Link link,
                           int a)

unionize_linkage

Sublinkage unionize_linkage()

linkage_compute_union

public int linkage_compute_union()

computes the union of the current sublinkages. Arcs may cross.

Returns:

the integer number of sublinkages

process_linkage

public void process_linkage(ParseOptions opts)

linkage_get_word

public java.lang.String linkage_get_word(int w)

linkage_get_link_label

public java.lang.String linkage_get_link_label(int index)

linkage_get_link_llabel

public java.lang.String linkage_get_link_llabel(int index)

linkage_get_link_rlabel

public java.lang.String linkage_get_link_rlabel(int index)

linkage_get_link_rword

public int linkage_get_link_rword(int index)

linkage_get_link_lword

public int linkage_get_link_lword(int index)

linkage_get_link_num_domains

public int linkage_get_link_num_domains(int index)

linkage_get_link_domain_names

public java.lang.String[] linkage_get_link_domain_names(int index)

linkage_get_violation_name

public java.lang.String linkage_get_violation_name()

linkage_is_canonical

public boolean linkage_is_canonical()

linkage_is_improper

public boolean linkage_is_improper()

linkage_has_inconsistent_domains

public boolean linkage_has_inconsistent_domains()

linkage_print_links_and_domains

public java.lang.String linkage_print_links_and_domains()

print_a_link

public void print_a_link(java.lang.StringBuffer s,
                int link)

Prints the information contained in a link into a supplied string buffer.

Parameters:

s - Empty stringbuffer supplied by the caller shich the method will use

link - the index of the link

build_linkage_postscript_string

public java.lang.String build_linkage_postscript_string()

builds a postscript printable string of the current sublinkage

set_centers

public void set_centers(boolean print_word_0)

linkage_print_diagram

public java.lang.String linkage_print_diagram()

linkage_print_constituent_tree

public java.lang.String linkage_print_constituent_tree(int mode)

String linkage_print_constituent_tree(int mode)

Parameters:

mode - mode 1: treebank-style constituent tree mode 2: flat, bracketed tree [A like [B this B] A] mode 3: flat, treebank-style tree (A like (B this) )

Returns:

Returns the constituent tree in one of three modes, treebank, flat bracketed, or flat treebank

token_type

int token_type(java.lang.String token)

parse_string

public CNode parse_string(CNode n,
                 java.util.StringTokenizer tok)

linkage_constituent_tree

public CNode linkage_constituent_tree()

CNode linkage_constituent_tree(); The other useful method for dealing with a cosntituent tree is: String linkage_print_constituent_tree(int mode);

Returns:

A CNode is a standard tree data structure. The children of a node are stored as a linked list, with the end of the list indicated by next==NULL. The start and end fields of a node indicate the span of the constituent, with the first word indexed by 0. Leaves are defined by the condition child==NULL.

linkage_print_postscript

public java.lang.String linkage_print_postscript(int mode)

Print out a postscript version of the constituent tree

Parameters:

mode - mode 1: treebank-style constituent tree mode 2: flat, bracketed tree [A like [B this B] A] mode 3: flat, treebank-style tree (A like (B this) )

Returns:

A Java String version of a postscript file that can be sent to a print device.

adjust_subordinate_clauses

public void adjust_subordinate_clauses(int numcon_total,
                              int numcon_subl)

uppercompare

public static boolean uppercompare(java.lang.String s,
                   java.lang.String t)

linkage_post_process

public void linkage_post_process(Postprocessor postprocessor)

post process the sublinkages according to the named postprocessor

trailer

public java.lang.String trailer(int mode)

header

public java.lang.String header(int mode)

Supplies the postscript header used when generating a postscript output file

Parameters:

mode - Print out the constituent tree. mode 1: treebank-style constituent tree mode 2: flat, bracketed tree [A like [B this B] A] mode 3: flat, treebank-style tree (A like (B this) )

Returns:

Retunrs the String of the necessary postscript header.