Replace json-simple with GSON

.gitignore

@@ -1,3 +1,4 @@
 data
+data-test
 target
 output

pom.xml

@@ -30,9 +30,9 @@
         <version>8.6.3</version>
     </dependency>
     <dependency>
-        <groupId>com.googlecode.json-simple</groupId>
+      <groupId>com.google.code.gson</groupId>
-        <artifactId>json-simple</artifactId>
+      <artifactId>gson</artifactId>
-        <version>1.1.1</version>
+      <version>2.8.6</version>
     </dependency>
   </dependencies>
 

src/main/java/org/RI/P2/Indexer.java

@@ -19,16 +19,16 @@ import org.apache.lucene.analysis.Analyzer;
 import org.apache.lucene.analysis.core.WhitespaceAnalyzer;
 import org.apache.lucene.analysis.en.EnglishAnalyzer;
 import org.apache.lucene.analysis.miscellaneous.PerFieldAnalyzerWrapper;
-import org.apache.lucene.analysis.standard.StandardAnalyzer;
 import org.apache.lucene.document.Document;
+import org.apache.lucene.document.Field;
+import org.apache.lucene.document.StringField;
+import org.apache.lucene.document.TextField;
 import org.apache.lucene.index.IndexWriter;
 import org.apache.lucene.index.IndexWriterConfig;
 import org.apache.lucene.index.IndexWriterConfig.OpenMode;
 import org.apache.lucene.store.Directory;
 import org.apache.lucene.store.FSDirectory;
-import org.json.simple.JSONArray;
+import com.google.gson.Gson;
-import org.json.simple.JSONObject;
-import org.json.simple.JSONValue;
 
 public class Indexer {
     IndexWriter index;
@@ -57,13 +57,12 @@ public class Indexer {
         return files;
     }
 
-    JSONArray parseJSONFile(File file) throws IOException {
+    Paper parseJSONFile(File file) throws IOException {
         InputStream jsonFile = new FileInputStream(file);
         Reader readerJson = new InputStreamReader(jsonFile);
-        Object fileObject = JSONValue.parse(readerJson);
+        Gson gson = new Gson();
-        JSONArray arrayObject = new JSONArray();
+        Paper data = gson.fromJson(readerJson, Paper.class);
-        arrayObject.add(fileObject);
+        return data;
-        return arrayObject;
     }
 
     void createIndex() throws IOException {
@@ -73,11 +72,16 @@ public class Indexer {
         index = new IndexWriter(dir, config);
     }
 
-    void addDocuments(JSONArray jsonObjects) throws IOException {
+    void addDocument(Paper paper) throws IOException {
-        for (JSONObject object : (List<JSONObject>) jsonObjects) {
+        Document doc = new Document();
-            Document doc = new Document();
+        doc.add(new StringField("document_id", paper.paper_id, Field.Store.YES));
-            index.addDocument(doc);
+        doc.add(new TextField("title", paper.metadata.title, Field.Store.YES));
+        for (Author author : paper.metadata.authors) {
+            String authorName = author.first + " " + author.middle + " " + author.last;
+            authorName = authorName.replaceAll("\\p{P}", "");
+            doc.add(new TextField("authors", authorName, Field.Store.YES));
         }
+        index.addDocument(doc);
     }
 
     void commitChanges() throws IOException {
@@ -88,8 +92,8 @@ public class Indexer {
     void populateIndex() throws IOException, ParseException {
         createIndex();
         for (File file : files) {
-            JSONArray jsonObjects = parseJSONFile(file);
+            Paper paper = parseJSONFile(file);
-            addDocument(jsonObjects);
+            addDocument(paper);
         }
         commitChanges();
     }

src/main/java/org/RI/P2/Paper.java

@@ -0,0 +1,45 @@
+package org.RI.P2;
+
+import java.util.List;
+
+class Affiliation {
+    String laboratory;
+    String institution;
+}
+
+class Location {
+    String postCode;
+    String settlement;
+    String region;
+    String country;
+}
+
+class Author {
+    String first;
+    List<String> middle;
+    String last;
+    String suffix;
+    Affiliation affiliation;
+    Location location;
+    String email;
+}
+
+class Metadata {
+    String title;
+    List<Author> authors;
+}
+
+class Abstract {
+    String text;
+}
+
+class Body_Text {
+    String text;
+}
+
+public class Paper {
+    String paper_id;
+    Metadata metadata;
+    List<Abstract> _abstract;
+    List<Body_Text> body_text;
+}