From f95a5be5dc85f23bcf12830a7ca95e6625e1a731 Mon Sep 17 00:00:00 2001
From: noureldin-eg <noureldin@eg.ibm.com>
Date: Sat, 11 Feb 2023 16:50:28 +0200
Subject: [PATCH] PDFBOX-5487: Remove all space characters if contained within
 the adjacent letters

https://issues.apache.org/jira/browse/PDFBOX-5487
---
 .../apache/pdfbox/text/PDFTextStripper.java   | 32 ++++++++++++
 .../org/apache/pdfbox/text/TextPosition.java  | 50 +++++++++++++++++++
 2 files changed, 82 insertions(+)

diff --git a/pdfbox/src/main/java/org/apache/pdfbox/text/PDFTextStripper.java b/pdfbox/src/main/java/org/apache/pdfbox/text/PDFTextStripper.java
index 4be7defe333..7f2a8bf969e 100644
--- a/pdfbox/src/main/java/org/apache/pdfbox/text/PDFTextStripper.java
+++ b/pdfbox/src/main/java/org/apache/pdfbox/text/PDFTextStripper.java
@@ -508,6 +508,10 @@ protected void writePage() throws IOException
                 {
                     IterativeMergeSort.sort(textList, comparator);
                 }
+                finally
+                {
+                    removeContainedSpaces(textList);
+                }
             }
 
             startArticle();
@@ -709,6 +713,34 @@ private boolean overlap(float y1, float height1, float y2, float height2)
                 || y1 <= y2 && y1 >= y2 - height2;
     }
 
+    /**
+     * Remove all space characters if contained within the adjacent letters
+     */
+    private void removeContainedSpaces(List<TextPosition> textList)
+    {
+        TextPosition position, previousPosition;
+        Iterator<TextPosition> iterator = textList.iterator();
+        
+        if (!iterator.hasNext())
+        {
+            return;
+        }
+        previousPosition = iterator.next();
+        
+        while (iterator.hasNext()) 
+        {
+            position = iterator.next();
+            if (" ".equals(position.getUnicode()) && previousPosition.completelyContains(position))
+            {
+                iterator.remove();
+            }
+            else
+            {
+                previousPosition = position;
+            } 
+        }
+    }
+
     /**
      * Write the line separator value to the output stream.
      * 
diff --git a/pdfbox/src/main/java/org/apache/pdfbox/text/TextPosition.java b/pdfbox/src/main/java/org/apache/pdfbox/text/TextPosition.java
index f4ba18d0344..227d82f50d9 100644
--- a/pdfbox/src/main/java/org/apache/pdfbox/text/TextPosition.java
+++ b/pdfbox/src/main/java/org/apache/pdfbox/text/TextPosition.java
@@ -570,6 +570,56 @@ else if (tp2Xstart < thisXstart && tp2Xend < thisXend)
         return true;
     }
 
+    /**
+     * Determine if this TextPosition perfectly contains another (i.e. the other TextPosition
+     * overlaps 100% with this one and fits entirely inside its bounding box when they are rendered
+     * on top of each other).
+     *
+     * @param tp2 The other TestPosition to compare against
+     * @return True if tp2 is contained completely inside the bounding box of this text.
+     */
+    public boolean completelyContains(TextPosition tp2)
+    {
+        //  Note: (0, 0) is in the upper left and y-coordinate is top of TextPosition
+         
+        //      +---thisTop------------+
+        //      |    +--tp2Top---+     |
+        //      |    |           |     |
+        //  thisLeft |       tp2Right  |
+        //      | tp2Left        | thisRight
+        //      |    |           |     |
+        //      |    +-tp2Bottom-+     |
+        //      +---------thisBottom---+
+
+        float thisLeft = getXDirAdj();
+        float thisWidth = getWidthDirAdj();
+        float thisRight = thisLeft + thisWidth;
+
+        float tp2Left = tp2.getXDirAdj();
+        float tp2Width = tp2.getWidthDirAdj();
+        float tp2Right = tp2Left + tp2Width;
+
+        if (thisLeft > tp2Left || tp2Right > thisRight)
+        {
+            return false;
+        }
+
+        float thisTop = getYDirAdj();
+        float thisHeight = getHeightDir();
+        float thisBottom = thisTop + thisHeight;
+
+        float tp2Top = tp2.getYDirAdj();
+        float tp2Height = tp2.getHeightDir();
+        float tp2Bottom = tp2Top + tp2Height;
+        
+        if (thisTop > tp2Top || tp2Bottom > thisBottom)
+        {
+            return false;
+        }
+
+        return true;
+    }
+
     /**
      * Merge a single character TextPosition into the current object. This is to be used only for
      * cases where we have a diacritic that overlaps an existing TextPosition. In a graphical