Adaptation to newer Lucene 8.7

.github/workflows/build.yml

@@ -9,13 +9,13 @@ jobs:
     steps:
     - uses: actions/setup-java@v1
       with:
-        java-version: '8'
+        java-version: '21'
     - uses: actions/checkout@v2
       with:
         repository: MIR-MU/MathMLCan
     - run: mvn clean install
     - run: mvn integration-test
-    - uses: actions/upload-artifact@v2
+    - uses: actions/upload-artifact@v7
       with:
         name: MathMLCan
         path: .
@@ -24,13 +24,13 @@ jobs:
     steps:
     - uses: actions/setup-java@v1
       with:
-        java-version: '8'
+        java-version: '21'
     - uses: actions/checkout@v2
       with:
-        repository: MIR-MU/MathMLUnificator
+        repository: fsi-open/MathMLUnificator
     - run: mvn clean install
     - run: mvn integration-test
-    - uses: actions/upload-artifact@v2
+    - uses: actions/upload-artifact@v7
       with:
         name: MathMLUnificator
         path: .
@@ -40,21 +40,21 @@ jobs:
     steps:
     - uses: actions/setup-java@v1
       with:
-        java-version: '8'
-    - uses: actions/download-artifact@v2
+        java-version: '21'
+    - uses: actions/download-artifact@v7
       with:
         name: MathMLCan
         path: MathMLCan
     - run: mvn install -f MathMLCan/pom.xml
-    - uses: actions/download-artifact@v2
+    - uses: actions/download-artifact@v7
       with:
         name: MathMLUnificator
         path: MathMLUnificator
     - run: mvn install -f MathMLUnificator/pom.xml
     - uses: actions/checkout@v2
     - run: mvn clean install
     - run: mvn integration-test
-    - uses: actions/upload-artifact@v2
+    - uses: actions/upload-artifact@v7
       with:
         name: MIaSMath
         path: .

pom.xml

@@ -13,27 +13,19 @@
     <properties>
         <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
     </properties>
-	
+
     <licenses>
         <license>
             <name>Apache License, Version 2.0</name>
             <url>http://www.apache.org/licenses/LICENSE-2.0</url>
         </license>
     </licenses>
 
-    <repositories>
-        <repository>
-            <id>maven-restlet</id>
-            <name>Public online Restlet repository</name>
-            <url>https://maven.restlet.com</url>
-        </repository>
-    </repositories>
-
     <dependencies>
         <dependency>
             <groupId>org.apache.solr</groupId>
             <artifactId>solr-core</artifactId>
-            <version>4.10.4</version>
+            <version>9.8.0</version>
         </dependency>
         <dependency>
             <groupId>commons-io</groupId>
@@ -43,7 +35,7 @@
         <dependency>
             <groupId>cz.muni.fi.mir</groupId>
             <artifactId>mathml-canonicalizer</artifactId>
-            <version>1.4.0</version>
+            <version>1.3.1</version>
             <type>jar</type>
         </dependency>
         <dependency>
@@ -74,9 +66,15 @@
             <artifactId>log4j-core</artifactId>
             <version>2.13.2</version>
         </dependency>
-
+        <dependency>
+            <groupId>org.hamcrest</groupId>
+            <artifactId>hamcrest-all</artifactId>
+            <version>1.3</version>
+            <scope>test</scope>
+        </dependency>
+
     </dependencies>
-    
+
     <build>
         <plugins>
             <plugin>

src/main/java/cz/muni/fi/mias/MathTokenizerFactory.java

@@ -16,10 +16,9 @@
 package cz.muni.fi.mias;
 
 import cz.muni.fi.mias.math.MathTokenizer;
-import java.io.Reader;
 import java.util.Map;
 import org.apache.lucene.analysis.Tokenizer;
-import org.apache.lucene.analysis.util.TokenizerFactory;
+import org.apache.lucene.analysis.TokenizerFactory;
 import org.apache.lucene.util.AttributeFactory;
 
 /**
@@ -48,17 +47,13 @@
  */
 public class MathTokenizerFactory extends TokenizerFactory {
 
-    private final boolean subformulae;
-
-    public MathTokenizerFactory(Map<String, String> args) {
+    public MathTokenizerFactory(Map<String, String> args)
+    {
         super(args);
-        String subforms = args.get("subformulae");
-        subformulae = Boolean.parseBoolean(subforms);
     }
 
     @Override
-    public Tokenizer create(AttributeFactory af, Reader reader) {
-        return new MathTokenizer(reader, subformulae, MathTokenizer.MathMLType.BOTH);
+    public final Tokenizer create(AttributeFactory af) {
+        return new MathTokenizer(af, true, MathTokenizer.MathMLType.BOTH);
     }
-
 }

src/main/java/cz/muni/fi/mias/PayloadSimilarity.java

@@ -16,7 +16,7 @@
 package cz.muni.fi.mias;
 
 import cz.muni.fi.mias.math.PayloadHelper;
-import org.apache.lucene.search.similarities.DefaultSimilarity;
+import org.apache.lucene.search.similarities.ClassicSimilarity;
 import org.apache.lucene.util.BytesRef;
 
 /**
@@ -26,9 +26,8 @@
  *
  * @author Martin Liska
  */
-public class PayloadSimilarity extends DefaultSimilarity {
+public class PayloadSimilarity extends ClassicSimilarity {
 
-    @Override
     public float scorePayload(int docId, int start, int end, BytesRef byteRef) {
         float score = 1.0F;
         if (byteRef.bytes != BytesRef.EMPTY_BYTES) {

src/main/java/cz/muni/fi/mias/math/MathTokenizer.java

@@ -34,6 +34,7 @@
 import org.apache.lucene.analysis.tokenattributes.CharTermAttribute;
 import org.apache.lucene.analysis.tokenattributes.PayloadAttribute;
 import org.apache.lucene.analysis.tokenattributes.PositionIncrementAttribute;
+import org.apache.lucene.util.AttributeFactory;
 import org.apache.lucene.util.BytesRef;
 import org.jdom2.output.DOMOutputter;
 import org.w3c.dom.Document;
@@ -158,8 +159,8 @@ public enum MathMLType {
      * reduced</em>.
      * @param type type of MathML that should be processed
      */
-    public MathTokenizer(Reader input, boolean subformulae, MathMLType type) {
-        this(input, subformulae, type, subformulae);
+    public MathTokenizer(AttributeFactory af, boolean subformulae, MathMLType type) {
+        this(af, subformulae, type, subformulae);
     }
 
     /**
@@ -170,8 +171,8 @@ public MathTokenizer(Reader input, boolean subformulae, MathMLType type) {
      * subformulae (i.e. variants with unified variables, constants, extracted
      * subformulae etc.)
      */
-    public MathTokenizer(Reader input, boolean subformulae, MathMLType type, boolean reduceWeighting) {
-        super(input);
+    public MathTokenizer(AttributeFactory af, boolean subformulae, MathMLType type, boolean reduceWeighting) {
+        super(af);
 
         this.mmlType = type;
         this.subformulae = subformulae;

src/test/java/cz/muni/fi/mias/math/MathTokenizerTest.java

@@ -7,6 +7,7 @@
 import java.io.UnsupportedEncodingException;
 import org.apache.commons.compress.utils.CharsetNames;
 import org.apache.commons.io.IOUtils;
+import org.apache.lucene.util.AttributeFactory;
 import org.junit.*;
 
 public class MathTokenizerTest {
@@ -17,7 +18,8 @@ public class MathTokenizerTest {
     @Test
     public void stressTest() throws Exception {
         Reader reader = readResource(HUGE_MATHML);
-        MathTokenizer tokenizer = new MathTokenizer(reader, true, MathTokenizer.MathMLType.BOTH);
+        MathTokenizer tokenizer = new MathTokenizer(AttributeFactory.DEFAULT_ATTRIBUTE_FACTORY, true, MathTokenizer.MathMLType.BOTH);
+        tokenizer.setReader(reader);
 
         TimeLogger logger = new TimeLogger();