WIP

pdsimplifyparser/src/main/scala/edu/psu/sagnik/research/pdsimplify/path/impl/BB.scala

@@ -9,27 +9,26 @@ import edu.psu.sagnik.research.data.RectangleOTL
  */
 
 object BB {
-  def Line(p0: Point2D.Float, p1: Point2D.Float) = {
+  def Line(p0: Point2D.Float, p1: Point2D.Float, pageHeight: Float) = {
     val xMin = scala.math.min(p0.x, p1.x)
     val yMin = scala.math.min(p0.y, p1.y)
     val xMax = scala.math.max(p0.x, p1.x)
-    val yMax = scala.math.max(p0.x, p1.x)
+    val yMax = scala.math.max(p0.y, p1.y)
     RectangleOTL(
       xTopLeft = xMin,
-      yTopLeft = yMin,
+      yTopLeft = pageHeight - yMax,
       widthRight = xMax - xMin,
       heightDown = yMax - yMin
     )
   }
 
-  def Curve(start: Point2D.Float, end: Point2D.Float, cp1: Point2D.Float, cp2: Point2D.Float) = {
+  def Curve(start: Point2D.Float, end: Point2D.Float, cp1: Point2D.Float, cp2: Point2D.Float, pageHeight: Float) = {
     val xs = List(start.x, end.x, cp1.x, cp2.x)
     val ys = List(start.y, end.y, cp1.y, cp2.y)
     val xTopLeft = xs.min
-    val yTopLeft = ys.min
     RectangleOTL(
       xTopLeft = xTopLeft,
-      yTopLeft = yTopLeft,
+      yTopLeft = pageHeight - ys.max,
       widthRight = xs.max - xs.min,
       heightDown = ys.max - ys.min
     )

pdsimplifyparser/src/main/scala/edu/psu/sagnik/research/pdsimplify/path/impl/ProcessPaths.scala

@@ -11,7 +11,6 @@ import org.apache.pdfbox.pdmodel.graphics.image.PDImage
 import org.apache.pdfbox.util.Matrix
 
 import java.util.logging.{ Level, Logger }
-import Level.{ INFO, FINE }
 
 /**
  * Created by schoudhury on 6/22/16.
@@ -48,10 +47,10 @@ class ProcessPaths(page: PDPage) extends PDFGraphicsStreamEngine(page: PDPage) {
     currentSubPath = Some(
       PDShape(
         segments = List(
-          PDLine(fp(p0), fp(p1), BB.Line(fp(p0), fp(p1))),
-          PDLine(fp(p1), fp(p2), BB.Line(fp(p1), fp(p2))),
-          PDLine(fp(p2), fp(p3), BB.Line(fp(p2), fp(p3))),
-          PDLine(fp(p3), fp(p0), BB.Line(fp(p3), fp(p0)))
+          PDLine(fp(p0), fp(p1), BB.Line(fp(p0), fp(p1), page.getBBox.getHeight)),
+          PDLine(fp(p1), fp(p2), BB.Line(fp(p1), fp(p2), page.getBBox.getHeight)),
+          PDLine(fp(p2), fp(p3), BB.Line(fp(p2), fp(p3), page.getBBox.getHeight)),
+          PDLine(fp(p3), fp(p0), BB.Line(fp(p3), fp(p0), page.getBBox.getHeight))
         ),
         fromReCommand = true
       )
@@ -111,13 +110,13 @@ class ProcessPaths(page: PDPage) extends PDFGraphicsStreamEngine(page: PDPage) {
       case Some(csp) => currentSubPath = Some(
         csp.copy(
           segments = csp.segments :+
-          PDLine(currentPoint, fp(new Point2D.Float(x, y)), BB.Line(currentPoint, fp(new Point2D.Float(x, y))))
+          PDLine(currentPoint, fp(new Point2D.Float(x, y)), BB.Line(currentPoint, fp(new Point2D.Float(x, y)), page.getBBox.getHeight))
         )
       )
       case _ => currentSubPath = Some( //current sub path is empty. We need to start a new PDShape i.e. subpath
         PDShape(
           segments = List(
-            PDLine(currentPoint, fp(new Point2D.Float(x, y)), BB.Line(currentPoint, fp(new Point2D.Float(x, y))))
+            PDLine(currentPoint, fp(new Point2D.Float(x, y)), BB.Line(currentPoint, fp(new Point2D.Float(x, y)), page.getBBox.getHeight))
           ),
           fromReCommand = false
         )
@@ -137,7 +136,7 @@ class ProcessPaths(page: PDPage) extends PDFGraphicsStreamEngine(page: PDPage) {
             endPoint = fp(new Point2D.Float(x3, y3)),
             controlPoint1 = fp(new Point2D.Float(x1, y1)),
             controlPoint2 = fp(new Point2D.Float(x2, y2)),
-            BB.Curve(currentPoint, fp(new Point2D.Float(x3, y3)), fp(new Point2D.Float(x1, y1)), fp(new Point2D.Float(x2, y2)))
+            BB.Curve(currentPoint, fp(new Point2D.Float(x3, y3)), fp(new Point2D.Float(x1, y1)), fp(new Point2D.Float(x2, y2)), page.getBBox.getHeight)
           )
         )
       )
@@ -149,7 +148,7 @@ class ProcessPaths(page: PDPage) extends PDFGraphicsStreamEngine(page: PDPage) {
               endPoint = fp(new Point2D.Float(x3, y3)),
               controlPoint1 = fp(new Point2D.Float(x1, y1)),
               controlPoint2 = fp(new Point2D.Float(x2, y2)),
-              BB.Curve(currentPoint, fp(new Point2D.Float(x3, y3)), fp(new Point2D.Float(x1, y1)), fp(new Point2D.Float(x2, y2)))
+              BB.Curve(currentPoint, fp(new Point2D.Float(x3, y3)), fp(new Point2D.Float(x1, y1)), fp(new Point2D.Float(x2, y2)), page.getBBox.getHeight)
             )
           ),
           fromReCommand = false
@@ -172,16 +171,17 @@ class ProcessPaths(page: PDPage) extends PDFGraphicsStreamEngine(page: PDPage) {
         val startPoint = csp.segments.head.startPoint
         currentSubPath = Some(
           csp.copy(
-            segments = csp.segments :+ PDLine(currentPoint, startPoint, BB.Line(currentPoint, startPoint))
+            segments = csp.segments :+ PDLine(currentPoint, startPoint, BB.Line(currentPoint, startPoint, page.getBBox.getHeight))
           )
         )
         currentPoint = startPoint
         subPathComplete()
 
-      case _ => {
+      case _ =>
         logger.warning("A path encountered a close operator before it even started. " +
-          "It will henceforth be known as Rickon Stark Blvd."); subPathComplete()
-      } //should never reach here}
+          "It will henceforth be known as Rickon Stark Blvd.")
+        subPathComplete()
+      //should never reach here}
 
     }
 

pdsimplifyparser/src/main/scala/edu/psu/sagnik/research/pdsimplify/raster/impl/ProcessRaster.scala

@@ -77,7 +77,7 @@ class ProcessRaster(page: PDPage) extends PDFGraphicsStreamEngine(page: PDPage)
 
       bb = RectangleOTL(
         xTopLeft = getCTM.getTranslateX - page.getCropBox.getLowerLeftX,
-        yTopLeft = getCTM.getTranslateY - page.getCropBox.getLowerLeftY,
+        yTopLeft = page.getBBox.getHeight - (getCTM.getTranslateY - page.getCropBox.getLowerLeftY) - getCTM.getScaleY,
         widthRight = getCTM.getScaleX,
         heightDown = getCTM.getScaleY
       )

pdsimplifyparser/src/main/scala/edu/psu/sagnik/research/pdsimplify/text/impl/CalculateBB.scala

@@ -10,10 +10,10 @@ object CalculateBB {
 
   def apply(texts: List[TextSegment]): Option[RectangleOTL] =
     if (texts.nonEmpty) {
-      val xTopLeft = texts.map(x => x.bb.xTopLeft).min
-      val yTopLeft = texts.map(x => x.bb.yTopLeft).min
-      val width = texts.map(t => t.bb.xTopLeft + t.bb.widthRight).max - xTopLeft
-      val height = texts.map(t => t.bb.yTopLeft + t.bb.heightDown).max - xTopLeft
+      val xTopLeft = texts.map(t => t.bb.xTopLeft).min
+      val yTopLeft = texts.map(t => t.bb.yTopLeft).min
+      val width = texts.map(t => (t.bb.xTopLeft + t.bb.widthRight)).max - xTopLeft
+      val height = texts.map(t => (t.bb.yTopLeft + t.bb.heightDown)).max - yTopLeft
       Some(
         RectangleOTL(
           xTopLeft = xTopLeft,

pdsimplifyparser/src/main/scala/edu/psu/sagnik/research/pdsimplify/text/impl/TextPositionBB.scala

@@ -15,7 +15,7 @@ object TextPositionBB {
 
   def approximate(tP: TextPosition, page: PDPage) = RectangleOTL(
     xTopLeft = tP.getXDirAdj, // text can be rotated, which will change the x,y coordinates and bounding boxes
-    yTopLeft = tP.getYDirAdj,
+    yTopLeft = tP.getYDirAdj - tP.getHeightDir,
     widthRight = tP.getWidthDirAdj,
     heightDown = tP.getHeightDir //(tP.getYDirAdj - tP.getHeightDir)+tP.getHeightDir
   )
@@ -64,11 +64,12 @@ object TextPositionBB {
     }
 
     val s = rotateAT.createTransformedShape(flipAT.createTransformedShape(at.createTransformedShape(rect)))
+    val heightDown = s.getBounds2D.getMaxY.toFloat - s.getBounds2D.getMinY.toFloat
     RectangleOTL(
       xTopLeft = s.getBounds2D.getMinX.toFloat,
-      yTopLeft = s.getBounds2D.getMinY.toFloat,
+      yTopLeft = s.getBounds2D.getMinY.toFloat - heightDown,
       widthRight = s.getBounds2D.getMaxX.toFloat - s.getBounds2D.getMinX.toFloat,
-      heightDown = s.getBounds2D.getMaxY.toFloat - s.getBounds2D.getMinY.toFloat
+      heightDown = heightDown
     )
   }
 
@@ -116,12 +117,13 @@ object TextPositionBB {
     else {
       val xTopLeft = afterPageTransformation.map(_.getBounds.x).min
       val yTopLeft = afterPageTransformation.map(_.getBounds.y).min
+      val heightDown = afterPageTransformation.map(a => a.getBounds.y + a.getBounds.height).max - yTopLeft
       Some(
         RectangleOTL(
           xTopLeft = xTopLeft,
-          yTopLeft = yTopLeft,
+          yTopLeft = yTopLeft - heightDown,
           widthRight = afterPageTransformation.map(a => a.getBounds.x + a.getBounds.width).max - xTopLeft,
-          heightDown = afterPageTransformation.map(a => a.getBounds.y + a.getBounds.height).max - yTopLeft
+          heightDown = heightDown
         )
       )
     }

writers/src/main/scala/edu/psu/sagnik/research/pdwriters/impl/ShowResults.scala

@@ -4,17 +4,14 @@ import java.awt.Color
 import java.io.File
 
 import org.apache.pdfbox.pdmodel.PDDocument
-import java.util.logging.{Level, Logger}
+import java.util.logging.{ Level, Logger }
 
 import edu.psu.sagnik.research.data.RectangleOTL
 import edu.psu.sagnik.research.pdsimplify.impl.ProcessDocument
-import edu.psu.sagnik.research.pdsimplify.path.model.PDPath
-import edu.psu.sagnik.research.pdsimplify.raster.model.PDRasterImage
-import edu.psu.sagnik.research.pdsimplify.text.model.PDParagraph
 import edu.psu.sagnik.research.pdwriters.writers.image.CreateMarkedPNG
 import edu.psu.sagnik.research.pdwriters.writers.pdf.CreateMarkedPDF
 
-import scala.util.{Failure, Success}
+import scala.util.{ Failure, Success }
 
 /**
  * Created by schoudhury on 6/27/16.
@@ -27,14 +24,14 @@ object ShowResults {
   def printExtractionResult(pdLoc: String, pageNum: Int, bbs: List[RectangleOTL], c: Color, qualifier: String) = {
     val document = PDDocument.load(new File(pdLoc))
     val page = document.getPage(pageNum)
-    CreateMarkedPDF(pdLoc, document, pageNum, page, bbs, c, qualifier)
+    CreateMarkedPNG(pdLoc, document, pageNum, page, bbs, c, qualifier)
     logger.fine(s"created ${qualifier.substring(0, qualifier.length - 1)} marked PDF")
   }
 
   def main(args: Array[String]): Unit = {
-    val DEFAULT_LOC = "/Users/schoudhury/codes/res-doc-sci/res-app-cmd/src/test/resources/008baad5-3fdb-4aea-9311-3ef7499e3f1f.pdf"
+    val DEFAULT_LOC = "src/test/resources/test1.pdf"
     //"/Users/schoudhury/hassan/C10-2042.pdf"
-    val DEFAULT_PAGE_NUM = 3
+    val DEFAULT_PAGE_NUM = 5
 
     val pdLoc = if (args.length > 1) args(0) else DEFAULT_LOC
     val pageNum = if (args.length == 2) args(1).toInt else DEFAULT_PAGE_NUM

writers/src/main/scala/edu/psu/sagnik/research/pdwriters/writers/image/CreateMarkedPNG.scala

@@ -1,13 +1,13 @@
 package edu.psu.sagnik.research.pdwriters.writers.image
 
 import java.awt.geom.Rectangle2D
-import java.awt.{BasicStroke, Color}
-import java.io.{File}
-import java.util.logging.{Level, Logger}
+import java.awt.{ BasicStroke, Color }
+import java.io.{ File }
+import java.util.logging.{ Level, Logger }
 import javax.imageio.ImageIO
 
 import edu.psu.sagnik.research.data.RectangleOTL
-import org.apache.pdfbox.pdmodel.{PDDocument, PDPage}
+import org.apache.pdfbox.pdmodel.{ PDDocument, PDPage }
 import org.apache.pdfbox.rendering.PDFRenderer
 
 /**
@@ -31,7 +31,7 @@ object CreateMarkedPNG {
     val pH = page.getBBox.getHeight
 
     for (bb <- bbs) {
-      logger.info(s"[$tElemType] ${(bb.xTopLeft, bb.yTopLeft, bb.widthRight, bb.heightDown)}")
+      //logger.info(s"[$tElemType] ${(bb.xTopLeft, bb.yTopLeft, bb.widthRight, bb.heightDown)}")
       g2d.draw(new Rectangle2D.Float(bb.xTopLeft, bb.yTopLeft, bb.widthRight, bb.heightDown))
     }
 

writers/src/main/scala/edu/psu/sagnik/research/pdwriters/writers/json/CreateJSON.scala

@@ -4,7 +4,6 @@ import java.awt.geom.Point2D
 
 import edu.psu.sagnik.research.data.RectangleOTL
 import edu.psu.sagnik.research.pdsimplify.model.PDDocumentSimple
-import edu.psu.sagnik.research.pdsimplify.model.Rectangle
 import edu.psu.sagnik.research.pdsimplify.path.model.PDLine
 import org.json4s.JsonDSL._
 import org.json4s.JsonDSL._

writers/src/main/scala/edu/psu/sagnik/research/pdwriters/writers/pdf/CreateMarkedPDF.scala

@@ -2,10 +2,10 @@ package edu.psu.sagnik.research.pdwriters.writers.pdf
 
 import java.awt.Color
 import java.io.IOException
-import java.util.logging.{Level, Logger}
+import java.util.logging.{ Level, Logger }
 
 import edu.psu.sagnik.research.data.RectangleOTL
-import org.apache.pdfbox.pdmodel.{PDDocument, PDPage, PDPageContentStream}
+import org.apache.pdfbox.pdmodel.{ PDDocument, PDPage, PDPageContentStream }
 import org.apache.pdfbox.util.Matrix
 
 /**
@@ -21,9 +21,9 @@ object CreateMarkedPDF {
   private def drawRect(content: PDPageContentStream, color: Color, rect: RectangleOTL, page: PDPage, fill: Boolean) {
     content.addRect(
       rect.xTopLeft + page.getCropBox.getLowerLeftX,
-      page.getCropBox.getHeight - (rect.yTopLeft + page.getCropBox.getLowerLeftY),
-      rect.x2 - rect.x1,
-      rect.y1 - rect.y2
+      page.getCropBox.getHeight - (rect.yTopLeft + page.getCropBox.getLowerLeftY) - rect.heightDown,
+      rect.widthRight,
+      rect.heightDown
     )
     //remember the addRect is drawing a rectangle with x,y at bottom left. Also, we adjusted the rect for cropbox before. Since we are not changing the
     //content stream, that adjustment has to be _re_adjusted.
@@ -36,7 +36,7 @@ object CreateMarkedPDF {
     }
   }
 
-  def apply(docLoc: String, document: PDDocument, pageNum: Int, page: PDPage, bbs: List[Rectangle], color: Color, tElemType: String): Unit = {
+  def apply(docLoc: String, document: PDDocument, pageNum: Int, page: PDPage, bbs: List[RectangleOTL], color: Color, tElemType: String): Unit = {
     bbs.foreach(bb => {
       val content = new PDPageContentStream(document, page, PDPageContentStream.AppendMode.APPEND, false)
       drawRect(content, color, bb, page, fill = false)

writers/src/main/scala/edu/psu/sagnik/research/pdwriters/writers/svg/TextHelper.scala

@@ -34,9 +34,9 @@ object TextHelper {
     "\""
 
   def getLocationString(c: PDChar, h: Float): String = "y=\"" +
-    (c.bb.y2) +
+    (c.bb.yTopLeft) +
     "\" x=\"" +
-    c.bb.x1 +
+    c.bb.xTopLeft +
     "\">"
 
   def getTransformString(c: PDChar): String = "transform=\"rotate(" + c.style.rotation.toString + ")\""