Added support for separate state sets for each character. Not yet handling polymorphism/uncertaintly correctly.

Author: balhoff

src/mesquite/nexml/InterpretNEXML/NexmlWriters/NexmlCharactersBlockWriter.java

@@ -31,14 +31,14 @@
 import org.nexml.model.OTUs;
 
 public class NexmlCharactersBlockWriter extends NexmlBlockWriter {
-	
+
 	@SuppressWarnings("serial")
 	private static final Map<String , String> xmlMolecularDataTypeFor = new HashMap<String, String>() {{
 		put(DNAData.DATATYPENAME, MolecularMatrix.DNA);
 		put(RNAData.DATATYPENAME, MolecularMatrix.RNA);
 		put(ProteinData.DATATYPENAME, MolecularMatrix.Protein);
 	}};	
-	
+
 	/**
 	 * 
 	 * @param employerEmployee
@@ -57,60 +57,79 @@ protected Annotatable writeBlock(Document xmlProject, FileElement mesBlock) {
 		Taxa mesTaxa = mesData.getTaxa();
 		OTUs xmlTaxa = findEquivalentTaxa(mesTaxa,xmlProject);			
 		org.nexml.model.Matrix<?> xmlMatrix = null;		
-		CharacterStateSet xmlCharacterStateSet = null;
 		String mesDataType = mesData.getDataTypeName();
 		if ( xmlMolecularDataTypeFor.containsKey(mesDataType) ) {
 			xmlMatrix = xmlProject.createMolecularMatrix(xmlTaxa,xmlMolecularDataTypeFor.get(mesDataType));
-			xmlCharacterStateSet = ((MolecularMatrix)xmlMatrix).getCharacterStateSet();
 		}
 		else if ( mesDataType.equalsIgnoreCase(CategoricalData.DATATYPENAME) ) {
 			xmlMatrix = xmlProject.createCategoricalMatrix(xmlTaxa);   
-			xmlCharacterStateSet = ((CategoricalMatrix)xmlMatrix).createCharacterStateSet();
 		}
 		else if ( mesDataType.equalsIgnoreCase(ContinuousData.DATATYPENAME) ) {
 			xmlMatrix = xmlProject.createContinuousMatrix(xmlTaxa);      			
 		}			
 		else {
 			MesquiteMessage.warnProgrammer("Can't write data type "+mesDataType);
-		}   		
-		writeCharacterStates(mesData, xmlMatrix, xmlCharacterStateSet);
+		} 
+		writeCharacterStates(mesData, xmlMatrix);
 		return xmlMatrix;
 	}
-	
+
 	/**
 	 * 
 	 * @param mesData
 	 * @param xmlMatrix
 	 * @param xmlCharacterStateSet
 	 */
 	@SuppressWarnings("unchecked")
-	private void writeCharacterStates(CharacterData mesData, org.nexml.model.Matrix<?> xmlMatrix, CharacterStateSet xmlCharacterStateSet) {
+	private void writeCharacterStates(CharacterData mesData, org.nexml.model.Matrix<?> xmlMatrix) {
 		String mesDataType = mesData.getDataTypeName();
+		
 		int mesNchar = mesData.getNumChars();
 		List<Character> xmlCharacters = new ArrayList<Character>(mesNchar);
 		for ( int j = 0; j < mesNchar; j++ ) {
+			CharacterStateSet xmlCharacterStateSet = null;
+			if ( xmlMolecularDataTypeFor.containsKey(mesDataType) ) {
+				xmlCharacterStateSet = ((MolecularMatrix)xmlMatrix).getCharacterStateSet();
+			}
+			else if ( mesDataType.equalsIgnoreCase(CategoricalData.DATATYPENAME) ) {
+				xmlCharacterStateSet = ((CategoricalMatrix)xmlMatrix).createCharacterStateSet();
+			}
 			Character xmlChar = xmlMatrix.createCharacter(xmlCharacterStateSet);
 			String mesCharacterName = mesData.getCharacterName(j);
 			if ( null != mesCharacterName && ! mesCharacterName.equals("") ) {
 				xmlChar.setLabel(mesCharacterName);
 			}
+			if ( mesDataType.equalsIgnoreCase(CategoricalData.DATATYPENAME) ) {
+				CategoricalData data = ((CategoricalData)mesData);
+				int maxStateIndex = data.maxStateWithName(j);
+				for (int stateIndex = 0; stateIndex <= maxStateIndex; stateIndex++) {
+					String symbol = String.valueOf(data.getSymbol(stateIndex));
+					org.nexml.model.CharacterState state = xmlChar.getCharacterStateSet().createCharacterState(symbol);
+					state.setSymbol(symbol);
+					if (data.hasStateName(j, stateIndex)) {
+						String stateLabel = data.getStateName(j, stateIndex);
+						state.setLabel(stateLabel);
+					}
+				}
+			}
 			xmlCharacters.add(xmlChar);
 		}
 		for ( int j = 0; j < mesData.getNumTaxa(); j++ ) {
 			CharacterState[] mesChars = mesData.getCharacterStateArray(j, 0, mesNchar);
+			String unassignedSymbol = String.valueOf(mesData.getUnassignedSymbol());
 			Taxon mesTaxon = mesData.getTaxa().getTaxon(j);
 			OTU xmlTaxon = findEquivalentTaxon(mesTaxon,xmlMatrix.getOTUs());    			
 			for ( int k = 0; k < mesNchar; k++ ) {
 				Character xmlChar = xmlCharacters.get(k);
 				String mesCharString = mesChars[k].toDisplayString();
-				if ( mesCharString != null && !mesCharString.equals("-") ) {    					
+				if ( mesCharString != null && !mesCharString.equals("-") && !mesCharString.equals(unassignedSymbol)) {    					
 					if ( mesDataType.equalsIgnoreCase(ContinuousData.DATATYPENAME) ) {
 						MatrixCell<Double> xmlCell = (MatrixCell<Double>) xmlMatrix.getCell(xmlTaxon,xmlChar);
-						xmlCell.setValue((Double)xmlMatrix.parseSymbol(mesCharString));
+						xmlCell.setValue((Double)xmlMatrix.parseSymbol(mesCharString, xmlChar));
 					}
 					else if ( mesDataType.equalsIgnoreCase(CategoricalData.DATATYPENAME) ) {    						
 						MatrixCell<org.nexml.model.CharacterState> xmlCell = (MatrixCell<org.nexml.model.CharacterState>) xmlMatrix.getCell(xmlTaxon,xmlChar);
-						xmlCell.setValue((org.nexml.model.CharacterState)xmlMatrix.parseSymbol(mesCharString));
+						xmlCell.setValue((org.nexml.model.CharacterState)xmlMatrix.parseSymbol(mesCharString, xmlChar));
 					}
 					else if ( xmlMolecularDataTypeFor.containsKey(mesDataType) ) {
 						MatrixCell<org.nexml.model.CharacterState> xmlCell = (MatrixCell<org.nexml.model.CharacterState>) xmlMatrix.getCell(xmlTaxon,xmlChar);

src/org/nexml/model/Matrix.java

@@ -69,7 +69,7 @@ public interface Matrix<T> extends OTUsLinkable, Annotatable, Segmented<Characte
 	 * @param symbol
 	 * @return
 	 */
-	T parseSymbol(String symbol);
+	T parseSymbol(String symbol, Character character);
 
 	/**
 	 * Creates a row element for OTU otu, and populates

src/org/nexml/model/impl/CategoricalMatrixImpl.java

@@ -14,53 +14,53 @@
 import org.w3c.dom.Element;
 
 class CategoricalMatrixImpl extends
-		MatrixImpl<CharacterState> implements CategoricalMatrix {
-	
+MatrixImpl<CharacterState> implements CategoricalMatrix {
+
 	private Set<CharacterStateSet> mCharacterStateSets = new HashSet<CharacterStateSet>();
 	private MolecularCharacterStateSetImpl mMolecularCharacterStates = null;	
-	
-    /**
-     * Protected constructors that take a DOM document object but not
-     * an element object are used for generating new element nodes in
-     * a NeXML document. On calling such constructors, a new element
-     * is created, which can be retrieved using getElement(). After this
-     * step, the Impl class that called this constructor would still 
-     * need to attach the element in the proper location (typically
-     * as a child element of the class that called the constructor). 
-     * @param document a DOM document object
-     * @author rvosa
-     */	
+
+	/**
+	 * Protected constructors that take a DOM document object but not
+	 * an element object are used for generating new element nodes in
+	 * a NeXML document. On calling such constructors, a new element
+	 * is created, which can be retrieved using getElement(). After this
+	 * step, the Impl class that called this constructor would still 
+	 * need to attach the element in the proper location (typically
+	 * as a child element of the class that called the constructor). 
+	 * @param document a DOM document object
+	 * @author rvosa
+	 */	
 	protected CategoricalMatrixImpl(Document document) {
 		super(document,"Standard");
 	}
-	
-    /**
-     * Protected constructors are intended for recursive parsing, i.e.
-     * starting from the root element (which maps onto DocumentImpl) we
-     * traverse the element tree such that for every child element that maps
-     * onto an Impl class the containing class calls that child's protected
-     * constructor, passes in the element of the child. From there the 
-     * child takes over, populates itself and calls the protected 
-     * constructors of its children. These should probably be protected
-     * because there is all sorts of opportunity for outsiders to call
-     * these in the wrong context, passing in the wrong elements etc.
-     * @param document the containing DOM document object. Every Impl 
-     * class needs a reference to this so that it can create DOM element
-     * objects
-     * @param element the equivalent NeXML element (e.g. for OTUsImpl, it's
-     * the <otus/> element)
-     * @author rvosa
-     */
+
+	/**
+	 * Protected constructors are intended for recursive parsing, i.e.
+	 * starting from the root element (which maps onto DocumentImpl) we
+	 * traverse the element tree such that for every child element that maps
+	 * onto an Impl class the containing class calls that child's protected
+	 * constructor, passes in the element of the child. From there the 
+	 * child takes over, populates itself and calls the protected 
+	 * constructors of its children. These should probably be protected
+	 * because there is all sorts of opportunity for outsiders to call
+	 * these in the wrong context, passing in the wrong elements etc.
+	 * @param document the containing DOM document object. Every Impl 
+	 * class needs a reference to this so that it can create DOM element
+	 * objects
+	 * @param element the equivalent NeXML element (e.g. for OTUsImpl, it's
+	 * the <otus/> element)
+	 * @author rvosa
+	 */
 	protected CategoricalMatrixImpl(Document document, Element element, OTUsImpl otus) {
 		super(document, element);
 		for ( Element stateSetElement : getChildrenByTagName( getFormatElement(), CharacterStateSetImpl.getTagNameClass() ) ) {
 			createCharacterStateSet(stateSetElement);
 		}
-		
+
 		for ( Element characterElement : getChildrenByTagName( getFormatElement(), CharacterImpl.getTagNameClass() ) ) {
 			createCharacter(characterElement);
 		}
-		
+
 		for ( Element row : getChildrenByTagName( getMatrixElement(), "row") ) {
 			OTU otu = otus.getThingById(row.getAttribute("otu"));
 			MatrixRowImpl<CharacterState> matrixRow = new MatrixRowImpl<CharacterState>(getDocument(),row, this, false);
@@ -87,7 +87,7 @@ protected CharacterStateSet createCharacterStateSet(Element statesElement) {
 		mCharacterStateSets.add(charStateSet); // XXX Make this into a setter?
 		return charStateSet;
 	}
-	
+
 	/**
 	 * This is equivalent to creating a <states> element, i.e.
 	 * a container for state elements, polymorphic_state_set elements
@@ -96,6 +96,7 @@ protected CharacterStateSet createCharacterStateSet(Element statesElement) {
 	 * If the format element object doesn't exist yet it's created here
 	 * @author rvosa
 	 */
+	@Override
 	public CharacterStateSet createCharacterStateSet() {
 		CharacterStateSetImpl characterStateSet = new CharacterStateSetImpl(getDocument());
 		List<Element> currentCharElements = getChildrenByTagName(getFormatElement(), "char");
@@ -119,10 +120,11 @@ else if ( ! currentSetElements.isEmpty() ) {
 	 * (non-Javadoc)
 	 * @see org.nexml.model.CategoricalMatrix#getCharacterStateSets()
 	 */
+	@Override
 	public Set<CharacterStateSet> getCharacterStateSets() {
 		return Collections.unmodifiableSet(mCharacterStateSets);
 	}
-	
+
 	/**
 	 * This method creates a char element, i.e. a column definition.
 	 * Because NeXML requires for categorical matrices that these
@@ -131,14 +133,15 @@ public Set<CharacterStateSet> getCharacterStateSets() {
 	 * in here, from which the attribute's value is set. 
 	 * @author rvosa
 	 */
+	@Override
 	public Character createCharacter(CharacterStateSet characterStateSet) {
 		CharacterImpl character = new CharacterImpl(getDocument());
 		addThing(character);
 		character.setCharacterStateSet(characterStateSet);
 		attachFundamentalDataElement(getFormatElement(), character.getElement());
 		return character;
 	}	
-	
+
 	protected Character createCharacter(Element element) {
 		CharacterImpl character = new CharacterImpl(getDocument(),element);
 		addThing(character);
@@ -147,7 +150,7 @@ protected Character createCharacter(Element element) {
 		character.setCharacterStateSet(stateSet);
 		return character;
 	}
-	
+
 	protected CharacterStateSet lookupCharacterStateSetById(String stateSetId) {
 		if ( null == stateSetId ) {
 			return null;
@@ -161,68 +164,67 @@ protected CharacterStateSet lookupCharacterStateSetById(String stateSetId) {
 	}
 
 	public CharacterStateSet getDNACharacterStateSet() {
-	    if (mMolecularCharacterStates == null){
-	        mMolecularCharacterStates = new MolecularCharacterStateSetImpl(getDocument());
-	    }
-	    CharacterStateSet result = mMolecularCharacterStates.getDNAStateSet();
-	    CharacterStateSetImpl characterStateSet = (CharacterStateSetImpl)result;
-	    if (mCharacterStateSets.add(characterStateSet)){
-	        if ( null == getFormatElement() ) {
-	            setFormatElement( getDocument().createElementNS(DEFAULT_NAMESPACE,"format") );
-	            getElement().insertBefore( getFormatElement(), getElement().getFirstChild() );
-	        }
-	        getFormatElement().insertBefore( characterStateSet.getElement(), getFormatElement().getFirstChild() );
-	    }
-	    return result;
+		if (mMolecularCharacterStates == null){
+			mMolecularCharacterStates = new MolecularCharacterStateSetImpl(getDocument());
+		}
+		CharacterStateSet result = mMolecularCharacterStates.getDNAStateSet();
+		CharacterStateSetImpl characterStateSet = (CharacterStateSetImpl)result;
+		if (mCharacterStateSets.add(characterStateSet)){
+			if ( null == getFormatElement() ) {
+				setFormatElement( getDocument().createElementNS(DEFAULT_NAMESPACE,"format") );
+				getElement().insertBefore( getFormatElement(), getElement().getFirstChild() );
+			}
+			getFormatElement().insertBefore( characterStateSet.getElement(), getFormatElement().getFirstChild() );
+		}
+		return result;
 	}
 
 	public CharacterStateSet getRNACharacterStateSet() {
-        if (mMolecularCharacterStates == null){
-            mMolecularCharacterStates = new MolecularCharacterStateSetImpl(getDocument());
-        }
-        CharacterStateSet result = mMolecularCharacterStates.getRNAStateSet();
-        CharacterStateSetImpl characterStateSet = (CharacterStateSetImpl)result;
-        if (mCharacterStateSets.add(characterStateSet)){
-            if ( null == getFormatElement() ) {
-                setFormatElement( getDocument().createElementNS(DEFAULT_NAMESPACE,"format") );
-                getElement().insertBefore( getFormatElement(), getElement().getFirstChild() );
-            }
-            getFormatElement().insertBefore( characterStateSet.getElement(), getFormatElement().getFirstChild() );
-        }
-        return result;
-     }
-    
-    public CharacterStateSet getProteinCharacterStateSet(){
-        if (mMolecularCharacterStates == null){
-            mMolecularCharacterStates = new MolecularCharacterStateSetImpl(getDocument());
-        }
-        CharacterStateSet result = mMolecularCharacterStates.getProteinStateSet();
-        CharacterStateSetImpl characterStateSet = (CharacterStateSetImpl)result;
-        if (mCharacterStateSets.add(characterStateSet)){
-            if ( null == getFormatElement() ) {
-                setFormatElement( getDocument().createElementNS(DEFAULT_NAMESPACE,"format") );
-                getElement().insertBefore( getFormatElement(), getElement().getFirstChild() );
-            }
-            getFormatElement().insertBefore( characterStateSet.getElement(), getFormatElement().getFirstChild() );
-        }
-        return result;
-    }
-
-	public CharacterState parseSymbol(String symbol) {
-		CharacterStateSet lastSet = null;
-		for ( CharacterStateSet stateSet : getCharacterStateSets() ) {
-			CharacterState state = stateSet.lookupCharacterStateBySymbol(symbol);
-			if ( null != state ) {
-				return state;
+		if (mMolecularCharacterStates == null){
+			mMolecularCharacterStates = new MolecularCharacterStateSetImpl(getDocument());
+		}
+		CharacterStateSet result = mMolecularCharacterStates.getRNAStateSet();
+		CharacterStateSetImpl characterStateSet = (CharacterStateSetImpl)result;
+		if (mCharacterStateSets.add(characterStateSet)){
+			if ( null == getFormatElement() ) {
+				setFormatElement( getDocument().createElementNS(DEFAULT_NAMESPACE,"format") );
+				getElement().insertBefore( getFormatElement(), getElement().getFirstChild() );
 			}
-			lastSet = stateSet;
+			getFormatElement().insertBefore( characterStateSet.getElement(), getFormatElement().getFirstChild() );
+		}
+		return result;
+	}
+
+	public CharacterStateSet getProteinCharacterStateSet(){
+		if (mMolecularCharacterStates == null){
+			mMolecularCharacterStates = new MolecularCharacterStateSetImpl(getDocument());
+		}
+		CharacterStateSet result = mMolecularCharacterStates.getProteinStateSet();
+		CharacterStateSetImpl characterStateSet = (CharacterStateSetImpl)result;
+		if (mCharacterStateSets.add(characterStateSet)){
+			if ( null == getFormatElement() ) {
+				setFormatElement( getDocument().createElementNS(DEFAULT_NAMESPACE,"format") );
+				getElement().insertBefore( getFormatElement(), getElement().getFirstChild() );
+			}
+			getFormatElement().insertBefore( characterStateSet.getElement(), getFormatElement().getFirstChild() );
+		}
+		return result;
+	}
+
+	@Override
+	public CharacterState parseSymbol(String symbol, Character character) {
+		CharacterStateSet stateSet = character.getCharacterStateSet();
+		CharacterState state = stateSet.lookupCharacterStateBySymbol(symbol);
+		if ( null != state ) {
+			return state;
+		} else {
+			return stateSet.createCharacterState(symbol);	
 		}
-		return lastSet.createCharacterState(symbol);
 	}
 
 	@Override
 	String getSplitString() {
 		return "\\s+";
 	}
 
- }
+}

src/org/nexml/model/impl/ContinuousMatrixImpl.java

@@ -99,7 +99,8 @@ public Character createCharacter(CharacterStateSet stateSet) {
 		return createCharacter();
 	}
 
-	public Double parseSymbol(String symbol) {
+	@Override
+	public Double parseSymbol(String symbol, Character character) {
 		return Double.parseDouble(symbol);
 	}