Changeset 23906 for gs3-extensions/atlas-src/trunk/src/org/greenstone/atlas/server/GateScanner.java

Timestamp:

2011-04-15T13:38:25+12:00 (13 years ago)

Author:

sjm84

Message:

Committing most recent version of ATLAS

Location:

gs3-extensions/atlas-src/trunk/src/org/greenstone/atlas

Files:

• . (added)
• server (moved) (moved from gs3-extensions/atlas-src/trunk/src/org/greenstone/server )
• server/GateScanner.java (modified) (7 diffs)

gs3-extensions/atlas-src/trunk/src/org/greenstone/atlas/server/GateScanner.java

@@ -4 +4 @@
 import gate.AnnotationSet;
 import gate.Corpus;
-import gate.CorpusController;
 import gate.Document;
 import gate.Factory;
@@ -10 +9 @@
 import gate.Gate;
 import gate.GateConstants;
-import gate.ProcessingResource;
-import gate.corpora.RepositioningInfo;
 import gate.creole.ANNIEConstants;
-import gate.creole.ResourceInstantiationException;
 import gate.creole.SerialAnalyserController;
 import gate.util.GateException;
@@ -22 +18 @@
 import java.net.URL;
 import java.util.ArrayList;
-import java.util.HashSet;
-import java.util.Iterator;
-import java.util.Set;
+import java.util.HashMap;
 
 import org.apache.log4j.BasicConfigurator;
@@ -37 +31 @@
         {
             BasicConfigurator.configure();
-            Gate.setGateHome(new File("/home/sam/Desktop/Research/Gate5"));
-            Gate.setPluginsHome(new File("/home/sam/Desktop/Research/Gate5/plugins"));
-            Gate.setUserConfigFile(new File("/home/sam/Desktop/Research/Gate5/gate.xml"));
+            Gate.setGateHome(new File("C:\\Users\\sjm84\\Desktop\\stuff\\Gate5.2"));
+            Gate.setPluginsHome(new File("C:\\Users\\sjm84\\Desktop\\stuff\\Gate5.2\\plugins"));
+            Gate.setUserConfigFile(new File("C:\\Users\\sjm84\\Desktop\\stuff\\Gate5.2\\gate.xml"));
             Gate.init();
-            Gate.getCreoleRegister().registerDirectories(new URL("file:///home/sam/Desktop/Research/Gate5/plugins/ANNIE")); 
+            Gate.getCreoleRegister().registerDirectories(new File("C:\\Users\\sjm84\\Desktop\\stuff\\Gate5.2\\plugins\\ANNIE").toURI().toURL()); 
             Gate.getUserConfig().put(GateConstants.DOCUMENT_ADD_SPACE_ON_UNPACK_FEATURE_NAME, new Boolean(false));
             
@@ -70 +64 @@
     } // execute()
     
-    public ArrayList<Word> classifyText(String originalText)
+    public HashMap<String, Word> classifyText(String originalText)
     {
-        ArrayList<Word> words = new ArrayList<Word>();
+        HashMap<String, Word> words = new HashMap<String, Word>();
         try
         {
-            // create a GATE corpus and add a document for each command-line
-            // argument
             Corpus corpus = (Corpus) Factory.createResource("gate.corpora.CorpusImpl");
             FeatureMap params = Factory.newFeatureMap();
@@ -91 +83 @@
             AnnotationSet annotations = doc.getAnnotations();
             
-            int index = 0;
             for(int j = 0; j < annotations.size(); j++)
             {
@@ -102 +93 @@
                 if(currentAnnotation.getType().equals("Token"))
                 {
-                    Word w = new Word();
-                    w.setIndex(index++);
-                    w.setClassification((String)currentAnnotation.getFeatures().get("category"));
-                    w.setValue((String)currentAnnotation.getFeatures().get("string"));
-                    
-                    words.add(w);
+                    Word w = null;
+                    if(words.containsKey((String)currentAnnotation.getFeatures().get("string")))
+                    {
+                        w = words.get((String)currentAnnotation.getFeatures().get("string"));
+                        w.addClassification((String)currentAnnotation.getFeatures().get("category"));
+                        w.setValue((String)currentAnnotation.getFeatures().get("string"));
+                    }
+                    else
+                    {
+                        w = new Word();
+                        w.addClassification((String)currentAnnotation.getFeatures().get("category"));
+                        w.setValue((String)currentAnnotation.getFeatures().get("string"));
+                        words.put(w.getValue(), w);
+                    }
                 }
             }