# HG changeset patch
# User František Kučera <franta-hg@frantovo.cz>
# Date 1373402552 -7200
# Node ID b188eae2c092c95d1343f0c240d9f8ede7570513
# Parent  939fa8d8663ef5ebad2235a17abeeae944373898
generator: sorted index and synonyms

diff -r 939fa8d8663e -r b188eae2c092 java/dictionary-generator/src/cz/frantovo/telco/dictionary/Generator.java
--- a/java/dictionary-generator/src/cz/frantovo/telco/dictionary/Generator.java	Tue Jul 09 22:41:49 2013 +0200
+++ b/java/dictionary-generator/src/cz/frantovo/telco/dictionary/Generator.java	Tue Jul 09 22:42:32 2013 +0200
@@ -26,10 +26,10 @@
 import java.io.FileOutputStream;
 import java.io.FileWriter;
 import java.io.IOException;
-import java.nio.ByteBuffer;
-import java.nio.charset.Charset;
 import java.util.ArrayList;
 import java.util.List;
+import java.util.SortedSet;
+import java.util.TreeSet;
 import java.util.logging.Level;
 import java.util.logging.Logger;
 import javax.xml.parsers.DocumentBuilder;
@@ -66,18 +66,15 @@
 public class Generator {
 
 	private static final Logger log = Logger.getLogger(Generator.class.getName());
-	private static final String EMAIL_TOKEN = "ixumhht68";
+	private static final String EML_TO_KEN = "ixumhht68";
 	private final DocumentBuilderFactory documentBuilderFactory;
 	private final DocumentBuilder documentBuilder;
 	private final XPathFactory xpathFactory;
 	private final XPath xpath;
 	private final TransformerFactory xslFactory;
 	private final Transformer xsl;
-	private final Charset utf8;
 
 	public Generator() throws ParserConfigurationException, TransformerConfigurationException {
-		utf8 = Charset.forName("UTF-8");
-
 		documentBuilderFactory = DocumentBuilderFactory.newInstance();
 		documentBuilderFactory.setNamespaceAware(true);
 		documentBuilder = documentBuilderFactory.newDocumentBuilder();
@@ -101,6 +98,9 @@
 		DataOutputStream indexOutputStream = null;
 		BufferedWriter infoWriter = null;
 
+		SortedSet<IndexEntry> indexEntries = new TreeSet<>();
+		SortedSet<SynonymsEntry> synonymsEntries = new TreeSet<>();
+
 		try {
 			dictOutputStream = new FileOutputStream(dictFile);
 			synonymOutputStream = new DataOutputStream(new FileOutputStream(synonymFile));
@@ -112,12 +112,8 @@
 			// TODO: tags - labels/descriptions
 			xsl.setParameter("tags", sourceDocument.getElementsByTagNameNS(DICTIONARY, "tags").item(0));
 
-			/**
-			 * TODO: sorting
-			 */
 			long offset = 0;
 			long conceptIndex = 0;
-			long synonymCount = 0;
 			for (Node conceptNode : nodeIterable(sourceDocument.getElementsByTagNameNS(DICTIONARY, "concept"))) {
 				ByteArrayOutputStream conceptXhtml = new ByteArrayOutputStream();
 				xsl.transform(new DOMSource(conceptNode), new StreamResult(conceptXhtml));
@@ -134,15 +130,23 @@
 					}
 				}
 
-				synonymCount = +writeSynonyms(synonymOutputStream, names, conceptIndex);
-				writeIndex(indexOutputStream, names.get(0), offset, length);
+				IndexEntry indexEntry = new IndexEntry(names.get(0), offset, length);
+				indexEntries.add(indexEntry);
+
+				for (int i = 1; i < names.size(); i++) {
+					String name = names.get(i);
+					synonymsEntries.add(new SynonymsEntry(indexEntry, name));
+				}
 
 				offset = offset + length;
 				conceptIndex++;
 			}
 
+			writeIndex(indexOutputStream, indexEntries);
+			writeSynonyms(synonymOutputStream, synonymsEntries);
+
 			indexOutputStream.flush();
-			writeInfo(infoWriter, sourceDocument, conceptIndex + 1, synonymCount, indexFile.length());
+			writeInfo(infoWriter, sourceDocument, conceptIndex + 1, synonymsEntries.size(), indexFile.length());
 		} catch (SAXException | IOException | TransformerException | XPathExpressionException e) {
 			log.log(Level.SEVERE, "unable to generate", e);
 		} finally {
@@ -153,24 +157,17 @@
 		}
 	}
 
-	private void writeIndex(DataOutputStream indexOutputStream, String name, long offset, long length) throws IOException {
-		indexOutputStream.write(name.getBytes(utf8));
-		indexOutputStream.write(0);
-		indexOutputStream.writeInt((int) offset); // unsigned int 32
-		indexOutputStream.writeInt((int) length); // unsigned int 32
+	private void writeIndex(DataOutputStream indexOutputStream, SortedSet<IndexEntry> indexEntries) throws IOException {
+		long ordinal = 0;
+		for (IndexEntry e : indexEntries) {
+			e.serialize(indexOutputStream);
+			e.setOrdinal(ordinal++);
+		}
 	}
 
-	private int writeSynonyms(DataOutputStream synonymOutputStream, List<String> names, long baseIndex) throws IOException {
-		if (names.size() > 1) {
-			for (int i = 1; i < names.size(); i++) {
-				String name = names.get(i);
-				synonymOutputStream.write(name.getBytes(utf8));
-				synonymOutputStream.write(0);
-				synonymOutputStream.writeInt((int) baseIndex); // unsigned int 32
-			}
-			return names.size() - 1;
-		} else {
-			return 0;
+	private void writeSynonyms(DataOutputStream synonymOutputStream, SortedSet<SynonymsEntry> synonymsEntries) throws IOException {
+		for (SynonymsEntry s : synonymsEntries) {
+			s.serialize(synonymOutputStream);
 		}
 	}
 
@@ -184,7 +181,7 @@
 		infoWriter.write("idxfilesize=" + idxfilesize + "\n");
 		infoWriter.write("idxoffsetbits=32\n");
 		infoWriter.write("author=František Kučera\n");
-		infoWriter.write("email=telco-dictionary." + EMAIL_TOKEN + "@" + "frantovo.cz\n");
+		infoWriter.write("em" + "ail=telco" + "-dictionary." + EML_TO_KEN + "@" + "fran" + "tovo.cz\n");
 		infoWriter.write("website=https://telco.frantovo.cz\n");
 		infoWriter.write("description=A dictionary for telecommunications licensed under GNU FDL\n");
 		infoWriter.write("date=2013.07.09\n");
diff -r 939fa8d8663e -r b188eae2c092 java/dictionary-generator/src/cz/frantovo/telco/dictionary/IndexEntry.java
--- a/java/dictionary-generator/src/cz/frantovo/telco/dictionary/IndexEntry.java	Tue Jul 09 22:41:49 2013 +0200
+++ b/java/dictionary-generator/src/cz/frantovo/telco/dictionary/IndexEntry.java	Tue Jul 09 22:42:32 2013 +0200
@@ -17,14 +17,69 @@
  */
 package cz.frantovo.telco.dictionary;
 
+import java.io.DataOutputStream;
+import java.io.IOException;
+import java.nio.charset.StandardCharsets;
+import java.util.Objects;
+
 /**
  *
  * @author Ing. František Kučera (frantovo.cz)
  */
 public class IndexEntry implements Comparable<IndexEntry> {
 
+	private String name;
+	private long offset;
+	private long length;
+	private long ordinal;
+
+	public IndexEntry(String name, long offset, long length) {
+		this.name = name;
+		this.offset = offset;
+		this.length = length;
+	}
+
+	public void serialize(DataOutputStream indexOutputStream) throws IOException {
+		indexOutputStream.write(name.getBytes(StandardCharsets.UTF_8));
+		indexOutputStream.write(0);
+		indexOutputStream.writeInt((int) offset); // unsigned int 32
+		indexOutputStream.writeInt((int) length); // unsigned int 32
+	}
+
+	public void setOrdinal(long ordinal) {
+		this.ordinal = ordinal;
+	}
+
+	public long getOrdinal() {
+		return ordinal;
+	}
+
 	@Override
 	public int compareTo(IndexEntry o) {
-		throw new UnsupportedOperationException("Not supported yet."); //To change body of generated methods, choose Tools | Templates.
+		int nameDiff = name.compareTo(o.name);
+		if (nameDiff == 0) {
+			int offsetDiff = ((Long) offset).compareTo(o.offset);
+			if (offsetDiff == 0) {
+				return ((Long) length).compareTo(o.length);
+			} else {
+				return offsetDiff;
+			}
+		} else {
+			return nameDiff;
+		}
+	}
+
+	@Override
+	public boolean equals(Object o) {
+		return o instanceof IndexEntry && compareTo((IndexEntry) o) == 0;
+	}
+
+	@Override
+	public int hashCode() {
+		int hash = 5;
+		hash = 53 * hash + Objects.hashCode(this.name);
+		hash = 53 * hash + (int) (this.offset ^ (this.offset >>> 32));
+		hash = 53 * hash + (int) (this.length ^ (this.length >>> 32));
+		return hash;
 	}
 }
diff -r 939fa8d8663e -r b188eae2c092 java/dictionary-generator/src/cz/frantovo/telco/dictionary/SynonymsEntry.java
--- a/java/dictionary-generator/src/cz/frantovo/telco/dictionary/SynonymsEntry.java	Tue Jul 09 22:41:49 2013 +0200
+++ b/java/dictionary-generator/src/cz/frantovo/telco/dictionary/SynonymsEntry.java	Tue Jul 09 22:42:32 2013 +0200
@@ -17,14 +17,51 @@
  */
 package cz.frantovo.telco.dictionary;
 
+import java.io.DataOutputStream;
+import java.io.IOException;
+import java.nio.charset.StandardCharsets;
+import java.util.Objects;
+
 /**
  *
  * @author Ing. František Kučera (frantovo.cz)
  */
 public class SynonymsEntry implements Comparable<SynonymsEntry> {
 
+	private IndexEntry base;
+	private String name;
+
+	public SynonymsEntry(IndexEntry base, String name) {
+		this.base = base;
+		this.name = name;
+	}
+
+	public void serialize(DataOutputStream synonymOutputStream) throws IOException {
+		synonymOutputStream.write(name.getBytes(StandardCharsets.UTF_8));
+		synonymOutputStream.write(0);
+		synonymOutputStream.writeInt((int) base.getOrdinal()); // unsigned int 32
+	}
+
 	@Override
 	public int compareTo(SynonymsEntry o) {
-		throw new UnsupportedOperationException("Not supported yet."); //To change body of generated methods, choose Tools | Templates.
+		int nameDiff = name.compareTo(o.name);
+		if (nameDiff == 0) {
+			return base.compareTo(o.base);
+		} else {
+			return nameDiff;
+		}
+	}
+
+	@Override
+	public boolean equals(Object o) {
+		return o instanceof IndexEntry && compareTo((SynonymsEntry) o) == 0;
+	}
+
+	@Override
+	public int hashCode() {
+		int hash = 3;
+		hash = 47 * hash + Objects.hashCode(this.base);
+		hash = 47 * hash + Objects.hashCode(this.name);
+		return hash;
 	}
 }