Fix shifted bounding boxes calculation (#80)

* Fix shifted bounding boxes calculation

* Avoid mutating bounding boxes while drawing annotations

Author: LonelyMidoriya

java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/pdf/PDFWriter.java

@@ -51,11 +51,13 @@ public class PDFWriter {
 
     private static final List<List<PDAnnotation>> annotations = new ArrayList<>();
     private static final Logger LOGGER = Logger.getLogger(PDFWriter.class.getCanonicalName());
+    private static final List<BoundingBox> pageBoundingBoxes = new ArrayList<>();
 
     public static void updatePDF(File inputPDF, String password, String outputFolder, List<List<IObject>> contents) throws IOException {
         try (PDDocument document = Loader.loadPDF(inputPDF, password)) {
             for (int pageNumber = 0; pageNumber < StaticContainers.getDocument().getNumberOfPages(); pageNumber++) {
                 annotations.add(new ArrayList<>());
+                pageBoundingBoxes.add(DocumentProcessor.getPageBoundingBox(pageNumber));
                 for (IObject content : contents.get(pageNumber)) {
                     drawContent(content, PDFLayer.CONTENT);
                 }
@@ -64,6 +66,7 @@ public static void updatePDF(File inputPDF, String password, String outputFolder
                 document.getPage(pageNumber).getAnnotations().addAll(annotations.get(pageNumber));
             }
             annotations.clear();
+            pageBoundingBoxes.clear();
             createOptContentsForAnnotations(document);
             document.setAllSecurityToBeRemoved(true);
 
@@ -136,9 +139,15 @@ private static void drawListItems(PDFList list, PDAnnotation annot) throws IOExc
 
     public static PDAnnotation draw(BoundingBox boundingBox, float[] colorArray,
                                     String contents, Long id, PDAnnotation linkedAnnot, String level, PDFLayer layerName) {
+
+        BoundingBox movedBoundingBox = new BoundingBox(boundingBox);
+        BoundingBox pageBoundingBox = pageBoundingBoxes.get(boundingBox.getPageNumber());
+        if (pageBoundingBox != null) {
+            movedBoundingBox.move(pageBoundingBox.getLeftX(), pageBoundingBox.getBottomY());
+        }
         PDAnnotationSquareCircle square = new PDAnnotationSquare();
-        square.setRectangle(new PDRectangle(getFloat(boundingBox.getLeftX()), getFloat(boundingBox.getBottomY()),
-                getFloat(boundingBox.getWidth()), getFloat(boundingBox.getHeight())));
+        square.setRectangle(new PDRectangle(getFloat(movedBoundingBox.getLeftX()), getFloat(movedBoundingBox.getBottomY()),
+                getFloat(movedBoundingBox.getWidth()), getFloat(movedBoundingBox.getHeight())));
         square.setConstantOpacity(0.4f);
         PDColor color = new PDColor(colorArray, PDDeviceRGB.INSTANCE);
         square.setColor(color);

java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/processors/ContentFilterProcessor.java

@@ -72,6 +72,7 @@ private static void filterOutOfPageContents(int pageNumber, List<IObject> conten
         if (pageBoundingBox == null) {
             return;
         }
+        pageBoundingBox.move(-pageBoundingBox.getLeftX(), -pageBoundingBox.getBottomY());
         for (int index = 0; index < contents.size(); index++) {
             IObject object = contents.get(index);
             if (object != null && pageBoundingBox.notOverlaps(object.getBoundingBox())) {