Format code

nguyenq · Dec 3, 2023 · 0ec859f · 0ec859f
1 parent 62d7238
commit 0ec859f
Show file tree

Hide file tree

Showing 3 changed files with 6 additions and 35 deletions.
diff --git a/src/main/java/net/sourceforge/tess4j/util/Hocr2PdfParser.java b/src/main/java/net/sourceforge/tess4j/util/Hocr2PdfParser.java
@@ -18,8 +18,6 @@
 import javax.xml.parsers.SAXParserFactory;
 import java.awt.*;
 import java.io.*;
-import java.util.HashMap;
-import java.util.Map;
 
 import static java.lang.Float.valueOf;
 
@@ -255,7 +253,6 @@ public Hocr2PdfParser(String hocrFilepath, PDDocument pdDocument, boolean visibl
     }
 
     public void parse() throws SAXException, IOException, ParserConfigurationException {
-
         SAXParserFactory parserFactory = SAXParserFactory.newInstance();
         SAXParser parser = parserFactory.newSAXParser();
         XMLReader reader = parser.getXMLReader();

diff --git a/src/main/java/net/sourceforge/tess4j/util/PdfBoxUtilities.java b/src/main/java/net/sourceforge/tess4j/util/PdfBoxUtilities.java
@@ -15,14 +15,6 @@
  */
 package net.sourceforge.tess4j.util;
 
-import org.apache.pdfbox.multipdf.PDFMergerUtility;
-import org.apache.pdfbox.multipdf.Splitter;
-import org.apache.pdfbox.pdmodel.PDDocument;
-import org.apache.pdfbox.rendering.ImageType;
-import org.apache.pdfbox.rendering.PDFRenderer;
-import org.apache.pdfbox.tools.imageio.ImageIOUtil;
-import org.slf4j.LoggerFactory;
-
 import java.awt.image.BufferedImage;
 import java.io.File;
 import java.io.FilenameFilter;
@@ -231,20 +223,17 @@ public static void mergePdf(File[] inputPdfFiles, File outputPdfFile) {
 
     /**
      * Merge text from hocr file into a pdf
+     *
      * @param inputHocr input hocr file
      * @param inputPdfStr input pdf file
      * @param outputPdfStr ouput pdf file result of merging
      * @param visible does the text are visible or not
      * @throws Exception
      */
     public static void mergeHocrIntoAPdf(String inputHocr, String inputPdfStr, String outputPdfStr, boolean visible) throws Exception {
-        try (
-                PDDocument pdDocument = Loader.loadPDF(new File(inputPdfStr))
-        ) {
+        try (PDDocument pdDocument = Loader.loadPDF(new File(inputPdfStr))) {
             final Hocr2PdfParser hocr2PdfParser = new Hocr2PdfParser(inputHocr, pdDocument, visible, false, null);
-
             hocr2PdfParser.parse();
-
             pdDocument.save(outputPdfStr);
         }
     }

diff --git a/src/test/java/net/sourceforge/tess4j/util/PdfBoxUtilitiesTest.java b/src/test/java/net/sourceforge/tess4j/util/PdfBoxUtilitiesTest.java
@@ -29,61 +29,47 @@ public void setUp() {
     }
 
     @Test
-    void mergeHocrIntoAPdf_multiplePages() throws Exception {
+    public void mergeHocrIntoAPdf_multiplePages() throws Exception {
         String hOcrFilename = String.format("%s/%s", this.testResourcesDataPath, "multipage-img.hocr");
         String pdfFilename = String.format("%s/%s", this.testResourcesDataPath, "multipage-img.pdf");
         String outputPdf = "target/test-classes/test-results/multipage-img-with-hocr.pdf";
-
         PdfBoxUtilities.mergeHocrIntoAPdf(hOcrFilename, pdfFilename, outputPdf, false);
-
         assertPdfContainsText("Auf der Registerkarte 'Einflgen' enthalten", outputPdf);
     }
 
     @Test
-    void mergeHocrIntoAPdf_singlePage() throws Exception {
+    public void mergeHocrIntoAPdf_singlePage() throws Exception {
         String hOcrFilename = String.format("%s/%s", this.testResourcesDataPath, "eurotext.hocr");
         String pdfFilename = String.format("%s/%s", this.testResourcesDataPath, "eurotext.pdf");
         String outputPdf = "target/test-classes/test-results/eurotext-withHocr.pdf";
-
         PdfBoxUtilities.mergeHocrIntoAPdf(hOcrFilename, pdfFilename, outputPdf, false);
-
         assertPdfContainsText("The (quick) [brown]", outputPdf);
     }
 
     @Test
-    void mergeHocrIntoAPdf_createHocrThenMergeToPDF() throws Exception {
+    public void mergeHocrIntoAPdf_createHocrThenMergeToPDF() throws Exception {
         String pdfFilename = String.format("%s/%s", this.testResourcesDataPath, "eurotext.pdf");
         File imageFile1 = new File(pdfFilename);
         String outputbase1 = "target/test-classes/test-results/docrenderer2-1";
         String outputbase2 = "target/test-classes/test-results/docrenderer2-1-merge.pdf";
-
         List<ITesseract.RenderedFormat> formats = new ArrayList<>(Arrays.asList(ITesseract.RenderedFormat.HOCR));
-
         instance.createDocuments(new String[]{imageFile1.getPath()}, new String[]{outputbase1}, formats);
-
         assertTrue(new File(outputbase1 + ".hocr").exists());
-
         PdfBoxUtilities.mergeHocrIntoAPdf(outputbase1 + ".hocr", pdfFilename, outputbase2, false);
-
         assertTrue(new File(outputbase2).exists());
         assertPdfContainsText("The (quick) [brown]", outputbase2);
     }
 
     @Test
-    void mergeHocrIntoAPdf_createHocrOnMultipageThenMergeToPDF() throws Exception {
+    public void mergeHocrIntoAPdf_createHocrOnMultipageThenMergeToPDF() throws Exception {
         String pdfFilename = String.format("%s/%s", this.testResourcesDataPath, "multipage-img.pdf");
         File imageFile1 = new File(pdfFilename);
         String outputbase1 = "target/test-classes/test-results/docrenderer2-2";
         String outputbase2 = "target/test-classes/test-results/docrenderer2-2-merge.pdf";
-
         List<ITesseract.RenderedFormat> formats = new ArrayList<>(Arrays.asList(ITesseract.RenderedFormat.HOCR));
-
         instance.createDocuments(new String[]{imageFile1.getPath()}, new String[]{outputbase1}, formats);
-
         assertTrue(new File(outputbase1 + ".hocr").exists());
-
         PdfBoxUtilities.mergeHocrIntoAPdf(outputbase1 + ".hocr", pdfFilename, outputbase2, false);
-
         assertPdfContainsText("Auf der Registerkarte 'Einflgen' enthalten", outputbase2);
     }
 
@@ -92,5 +78,4 @@ private void assertPdfContainsText(String expectedString, String pdfFilepath) th
         String extractText = new PDFTextStripper().getText(doc);
         assertTrue(extractText.contains(expectedString));
     }
-
 }