Merge pull request #111 from Myyyvothrr/main

PDF/bytes reader

src/main/java/org/texttechnologylab/DockerUnifiedUIMAInterface/DUUIComposer.java

@@ -1683,7 +1683,7 @@ public TypeSystemDescription instantiate_pipeline() throws Exception {
     private JCas run_pipeline(String name, JCas jc, long documentWaitTime, Vector<PipelinePart> pipeline) throws Exception {
         progress.set(0);
 
-        DUUIDocument document = new DUUIDocument("Text", "Text", jc.getDocumentText().getBytes(StandardCharsets.UTF_8));
+        DUUIDocument document = new DUUIDocument("Text", "Text", jc);
         if (JCasUtil.select(jc, DocumentMetaData.class).isEmpty()) {
             DocumentMetaData dmd = DocumentMetaData.create(jc);
             dmd.setDocumentId(document.getName());

src/main/java/org/texttechnologylab/DockerUnifiedUIMAInterface/document_handler/DUUIDocument.java

@@ -53,6 +53,24 @@ public DUUIDocument(String name, String path, byte[] bytes) {
         this.size = bytes.length;
     }
 
+    public DUUIDocument(String name, String path, JCas jCas) {
+        if (jCas.getDocumentText() != null) {
+            this.bytes = jCas.getDocumentText().getBytes(StandardCharsets.UTF_8);
+        }
+        else if (jCas.getSofaDataStream() != null) {
+            try {
+                this.bytes = jCas.getSofaDataStream().readAllBytes();
+            }
+            catch (Exception e) {
+                e.printStackTrace();
+            }
+        }
+
+        this.name = name;
+        this.path = path;
+        this.size = bytes.length;
+    }
+
     @Override
     public boolean equals(Object o) {
         if (o == this) {

src/main/java/org/texttechnologylab/DockerUnifiedUIMAInterface/io/reader/bytes/DUUIBytesReader.java

@@ -0,0 +1,30 @@
+package org.texttechnologylab.DockerUnifiedUIMAInterface.io.reader.bytes;
+
+import org.apache.uima.collection.CollectionException;
+import org.apache.uima.fit.descriptor.ConfigurationParameter;
+import org.apache.uima.jcas.JCas;
+import org.apache.uima.jcas.cas.ByteArray;
+import org.dkpro.core.api.io.JCasResourceCollectionReader_ImplBase;
+import org.dkpro.core.api.resources.CompressionUtils;
+
+import java.io.InputStream;
+
+public class DUUIBytesReader extends JCasResourceCollectionReader_ImplBase {
+    public static final String PARAM_MIME_TYPE = "mimeType";
+    @ConfigurationParameter(name = PARAM_MIME_TYPE, mandatory = false)
+    protected String mimeType;
+
+    @Override
+    public void getNext(JCas jCas) throws CollectionException {
+        Resource res = nextFile();
+        initCas(jCas, res);
+        try (InputStream is = CompressionUtils.getInputStream(res.getLocation(), res.getInputStream())) {
+            byte[] content = is.readAllBytes();
+            ByteArray data = new ByteArray(jCas, content.length);
+            data.copyFromArray(content, 0, 0, content.length);
+            jCas.setSofaDataArray(data, mimeType);
+        } catch (Exception e) {
+            throw new CollectionException(e);
+        }
+    }
+}

src/main/java/org/texttechnologylab/DockerUnifiedUIMAInterface/io/reader/pdf/DUUIPDFReader.java

@@ -0,0 +1,16 @@
+package org.texttechnologylab.DockerUnifiedUIMAInterface.io.reader.pdf;
+
+import org.apache.uima.UimaContext;
+import org.apache.uima.resource.ResourceInitializationException;
+import org.texttechnologylab.DockerUnifiedUIMAInterface.io.reader.bytes.DUUIBytesReader;
+
+public class DUUIPDFReader extends DUUIBytesReader {
+    @Override
+    public void initialize(UimaContext aContext) throws ResourceInitializationException {
+        super.initialize(aContext);
+
+        if (mimeType == null) {
+            mimeType = "application/pdf";
+        }
+    }
+}