Many resources are needed to download a project. Please understand that we have to compensate our server costs. Thank you in advance. Project price only 1 $
You can buy this project and download/modify it how often you want.
package edu.arizona.sista.discourse.rstparser
import edu.arizona.sista.processors.Document
/**
* Stores the RST discourse tree for one document
* User: mihais
* Date: 4/4/14
*/
@SerialVersionUID(1L)
class DiscourseTree (
/** Label of this tree, if non-terminal */
var relationLabel:String,
/** Direction of the relation, if non-terminal */
var relationDirection:RelationDirection.Value,
/** Children of this non-terminal node */
var children:Array[DiscourseTree],
/** Nucleus or Satellite; used only during reading */
val kind:TreeKind.Value,
/** Raw text attached to this node */
val rawText:String,
/** Character offsets for the rawText; used only during reading */
var charOffsets:(Int, Int),
/** Position of the first token in the annotation for this discourse tree */
var firstToken: TokenOffset = null,
/** Position of the last token in the annotation for this discourse tree; this is inclusive! */
var lastToken: TokenOffset = null,
/** Index of the first EDU in this tree in the array of EDUs for the first sentence */
var firstEDU: Int = -1,
/** Index of the last EDU in this tree in the array of EDUs for the last sentence (inclusive) */
var lastEDU: Int = -1) extends Serializable {
/** Creates a terminal DiscourseTree */
def this(sent:Int, ft:Int, lt:Int, doc:Document, e:Int) =
this("", RelationDirection.None, null, TreeKind.Nucleus,
doc.sentences(sent).words.slice(ft, lt + 1).mkString(" "), (-1, -1),
new TokenOffset(sent, ft), new TokenOffset(sent, lt),
e, e)
/** Creates a simple non-terminal tree */
def this(l:String, d:RelationDirection.Value, c:Array[DiscourseTree]) =
this(l, d, c, TreeKind.Nucleus, null, (-1, -1),
c.head.firstToken, c.last.lastToken,
c.head.firstEDU, c.last.lastEDU)
def isTerminal:Boolean = children == null
def firstSentence:Int = firstToken.sentence
def lastSentence:Int = lastToken.sentence
def tokenCount:Int = {
if(isTerminal) {
assert(firstToken.sentence == lastToken.sentence)
lastToken.token - firstToken.token + 1
} else {
var sum = 0
for(c <- children) {
sum += c.tokenCount
}
sum
}
}
def eduCount:Int = {
if(isTerminal) 1
else {
var sum = 0
for(c <- children) {
sum += c.eduCount
}
sum
}
}
override def toString:String = {
val os = new StringBuilder
print(os, 0, printChildren = true, printKind = false)
os.toString()
}
def toString(printChildren:Boolean, printKind:Boolean):String = {
val os = new StringBuilder
print(os, 0, printChildren, printKind)
os.toString()
}
def print(os:StringBuilder,
offset:Int,
printChildren:Boolean,
printKind:Boolean) {
var i = 0
while(i < offset) {
os.append(" ")
i += 1
}
var printedText = false
if(printKind) {
os.append(kind)
printedText = true
}
if(relationLabel.length > 0) {
if(printKind) {
os.append(":")
}
os.append(relationLabel)
if(relationDirection != RelationDirection.None) {
os.append(" (")
os.append(relationDirection)
os.append(")")
}
printedText = true
}
if(rawText != null) {
if(printedText) os.append(" ")
os.append("TEXT")
if(charOffsets._1 != -1) {
os.append("(")
os.append(charOffsets._1)
os.append(", ")
os.append(charOffsets._2)
os.append(")")
}
os.append(":")
os.append(rawText)
}
if(printChildren) {
os.append("\n")
if(! isTerminal) {
for(c <- children) c.print(os, offset + 2, printChildren, printKind)
}
}
}
}
case class TokenOffset (sentence:Int, token:Int) {
override def toString:String = s"($sentence, $token)"
}
object TreeKind extends Enumeration {
type TreeKind = Value
val Nucleus, Satellite, Root = Value
}
object RelationDirection extends Enumeration {
type RelationDirection = Value
val LeftToRight, RightToLeft, None = Value
}
