Closes #1462: Run experiments with Grobid as a potential Cermine repl…

…acement

WIP: first version of Grobid wrapper. Not working at the moment due to java class version mismatch requiring the whole IIS to be compiled with java11 or newer.Y

iis-wf/iis-wf-metadataextraction/pom.xml

@@ -80,6 +80,11 @@
 			<artifactId>cermine-impl</artifactId>
 		</dependency>
 
+        <dependency>
+            <groupId>org.grobid</groupId>
+            <artifactId>grobid-core</artifactId>
+        </dependency>
+
         <dependency>
             <groupId>org.jdom</groupId>
             <artifactId>jdom</artifactId>

iis-wf/iis-wf-metadataextraction/src/main/java/eu/dnetlib/iis/wf/metadataextraction/GrobidBasedContentExtractor.java

@@ -0,0 +1,68 @@
+package eu.dnetlib.iis.wf.metadataextraction;
+
+import java.io.File;
+import java.util.Collections;
+
+import org.grobid.core.*;
+import org.grobid.core.data.*;
+import org.grobid.core.factory.*;
+import org.grobid.core.main.GrobidHomeFinder;
+import org.grobid.core.main.LibraryLoader;
+import org.grobid.core.utilities.*;
+import org.grobid.core.engines.Engine;
+import org.grobid.core.engines.config.GrobidAnalysisConfig;
+
+/**
+ * Simple content extractor relying on Grobid library.
+ * 
+ * @author mhorst
+ */
+public class GrobidBasedContentExtractor {
+
+    private static String grobidHome = "/home/mhorst/grobid-home";
+
+    public static void initializeGrobid() {
+
+        GrobidHomeFinder grobidHomeFinder = new GrobidHomeFinder(Collections.singletonList(grobidHome));
+        GrobidProperties.getInstance(grobidHomeFinder);
+
+        // Load native libraries required for Grobid
+        LibraryLoader.load();
+    }
+
+    public static void main(String[] args) {
+        // TODO Auto-generated method stub
+        String fileLocation = "/home/mhorst/Downloads/HPM-35-1009.pdf";
+
+        extractMeta(fileLocation);
+        extractText(fileLocation);
+    }
+
+    private static void extractMeta(String fileLocation) {
+        Engine engine = GrobidFactory.getInstance().createEngine();
+        // 0 - no consolidation
+        int consolidate = 0;
+        try {
+            BiblioItem result = new BiblioItem();
+            engine.processHeader(fileLocation, consolidate, result);
+            System.out.println("Title: " + result.getTitle());
+            System.out.println("Authors: " + result.getAuthors());
+            System.out.println("Abstract: " + result.getAbstract());
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+    }
+
+    private static void extractText(String fileLocation) {
+        Engine engine = GrobidFactory.getInstance().createEngine();
+        try {
+            GrobidAnalysisConfig config = GrobidAnalysisConfig.builder().consolidateHeader(0).build();
+            File file = new File(fileLocation);
+            String fullText = engine.fullTextToTEI(file, config);
+            System.out.println("Extracted Text: " + fullText);
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+    }
+
+}

pom.xml

@@ -97,6 +97,11 @@
                 <enabled>false</enabled>
             </snapshots>
         </repository>
+        <repository>
+            <id>grobid</id>
+            <name>GROBID DIY repo</name>
+            <url>https://grobid.s3.eu-west-1.amazonaws.com/repo/</url>
+        </repository>
     </repositories>
 
     <dependencies>
@@ -728,6 +733,12 @@
                 <artifactId>simpleclient_pushgateway</artifactId>
                 <version>0.8.1</version>
             </dependency>
+
+            <dependency>
+                <groupId>org.grobid</groupId>
+                <artifactId>grobid-core</artifactId>
+                <version>0.8.0</version>
+            </dependency>
 
             <dependency>
                 <groupId>org.junit</groupId>