/// <summary>
/// Adds 4 <see cref="Line"/>s forming a rectangle to the path.
/// </summary>
public void Rectangle(double x, double y, double width, double height)
{
currentPosition = new PdfPoint(x, y);
LineTo(x + width, y);
LineTo(x + width, y + height);
LineTo(x, y + height);
LineTo(x, y);
IsDrawnAsRectangle = true;
}
/// <summary>
/// log(n) implementation of find_intersections
/// based on http://people.csail.mit.edu/indyk/6.838-old/handouts/lec2.pdf
/// </summary>
/// <param name="horizontals"></param>
/// <param name="verticals"></param>
public static SortedDictionary <PdfPoint, Ruling[]> FindIntersections(IReadOnlyList <Ruling> horizontals, IReadOnlyList <Ruling> verticals)
{
//https://github.com/tabulapdf/tabula-java/blob/master/src/main/java/technology/tabula/Ruling.java#L312
List <SortObject> sos = new List <SortObject>();
SortedDictionary <Ruling, bool> tree = new SortedDictionary <Ruling, bool>(new TreeMapRulingComparer()); // TreeMap<Ruling, Boolean> tree
SortedDictionary <PdfPoint, Ruling[]> rv = new SortedDictionary <PdfPoint, Ruling[]>(new TreeMapPdfPointComparer()); // TreeMap<Point2D, Ruling[]> rv
foreach (Ruling h in horizontals)
{
sos.Add(new SortObject(SOType.HLEFT, h.Left - PERPENDICULAR_PIXEL_EXPAND_AMOUNT, h));
sos.Add(new SortObject(SOType.HRIGHT, h.Right + PERPENDICULAR_PIXEL_EXPAND_AMOUNT, h));
}
foreach (Ruling v in verticals)
{
sos.Add(new SortObject(SOType.VERTICAL, v.Left, v));
}
sos.Sort(new SortObjectComparer()); //Collections.sort(sos, new Comparator<SortObject>() ...
foreach (SortObject so in sos)
{
switch (so.type)
{
case SOType.VERTICAL:
foreach (var h in tree)
{
PdfPoint?i = h.Key.IntersectionPoint(so.ruling);
if (!i.HasValue)
{
continue;
}
rv[i.Value] = new Ruling[]
{
h.Key.Expand(PERPENDICULAR_PIXEL_EXPAND_AMOUNT),
so.ruling.Expand(PERPENDICULAR_PIXEL_EXPAND_AMOUNT)
};
}
break;
case SOType.HRIGHT:
tree.Remove(so.ruling);
break;
case SOType.HLEFT:
tree[so.ruling] = true;
break;
}
}
return(rv);
}
internal void LineTo(decimal x, decimal y)
{
if (currentPosition.HasValue)
{
var to = new PdfPoint(x, y);
commands.Add(new Line(currentPosition.Value, to));
currentPosition = to;
}
else
{
MoveTo(x, y);
}
}
internal void BezierCurveTo(decimal x1, decimal y1, decimal x2, decimal y2, decimal x3, decimal y3)
{
if (currentPosition.HasValue)
{
var to = new PdfPoint(x3, y3);
commands.Add(new BezierCurve(currentPosition.Value,
new PdfPoint(x1, y1), new PdfPoint(x2, y2), to));
currentPosition = to;
}
else
{
MoveTo(x3, y3);
}
}
internal void LineTo(decimal x, decimal y)
{
if (currentPosition.HasValue)
{
var to = new PdfPoint(x, y);
commands.Add(new Line(currentPosition.Value, to));
currentPosition = to;
}
else
{
// TODO: probably the wrong behaviour here, maybe line starts from (0, 0)?
MoveTo(x, y);
}
}
internal void BezierCurveTo(decimal x1, decimal y1, decimal x2, decimal y2, decimal x3, decimal y3)
{
if (currentPosition.HasValue)
{
var to = currentTransformationMatrix.Transform(new PdfPoint(x3, y3));
commands.Add(new BezierCurve(currentPosition.Value,
currentTransformationMatrix.Transform(new PdfPoint(x1, y1)), currentTransformationMatrix.Transform(new PdfPoint(x2, y2)), to));
currentPosition = to;
}
else
{
MoveTo(x3, y3);
}
}
/// <summary>
/// Add a <see cref="Line"/> command to the path.
/// </summary>
public void LineTo(double x, double y)
{
if (currentPosition.HasValue)
{
shoeLaceSum += ((x - currentPosition.Value.X) * (y + currentPosition.Value.Y));
var to = new PdfPoint(x, y);
commands.Add(new Line(currentPosition.Value, to));
currentPosition = to;
}
else
{
// PDF Reference 1.7 p226
throw new ArgumentNullException("LineTo(): currentPosition is null.");
}
}
/// <summary>
/// Add a <see cref="Line"/> command to the path.
/// </summary>
public void LineTo(double x, double y)
{
if (currentPosition.HasValue)
{
shoeLaceSum += ((x - currentPosition.Value.X) * (y + currentPosition.Value.Y));
var to = new PdfPoint(x, y);
commands.Add(new Line(currentPosition.Value, to));
currentPosition = to;
}
else
{
// TODO: probably the wrong behaviour here, maybe line starts from (0, 0)?
MoveTo(x, y);
}
}
internal void BezierCurveTo(decimal x1, decimal y1, decimal x2, decimal y2, decimal x3, decimal y3)
{
if (currentPosition.HasValue)
{
shoeLaceSum += (double)((x1 - currentPosition.Value.X) * (y1 + currentPosition.Value.Y));
shoeLaceSum += (double)((x2 - x1) * (y2 + y1));
shoeLaceSum += (double)((x3 - x2) * (y3 + y2));
var to = new PdfPoint(x3, y3);
commands.Add(new BezierCurve(currentPosition.Value, new PdfPoint(x1, y1), new PdfPoint(x2, y2), to));
currentPosition = to;
}
else
{
MoveTo(x3, y3);
}
}
/// <summary>
/// Add a <see cref="BezierCurve"/> to the path.
/// </summary>
public void BezierCurveTo(double x1, double y1, double x2, double y2, double x3, double y3)
{
if (currentPosition.HasValue)
{
shoeLaceSum += (x1 - currentPosition.Value.X) * (y1 + currentPosition.Value.Y);
shoeLaceSum += (x2 - x1) * (y2 + y1);
shoeLaceSum += (x3 - x2) * (y3 + y2);
var to = new PdfPoint(x3, y3);
commands.Add(new BezierCurve(currentPosition.Value, new PdfPoint(x1, y1), new PdfPoint(x2, y2), to));
currentPosition = to;
}
else
{
MoveTo(x3, y3);
}
}
/// <summary>
/// Add a <see cref="BezierCurve"/> to the path.
/// </summary>
public void BezierCurveTo(double x1, double y1, double x2, double y2, double x3, double y3)
{
if (currentPosition.HasValue)
{
shoeLaceSum += (x1 - currentPosition.Value.X) * (y1 + currentPosition.Value.Y);
shoeLaceSum += (x2 - x1) * (y2 + y1);
shoeLaceSum += (x3 - x2) * (y3 + y2);
var to = new PdfPoint(x3, y3);
commands.Add(new BezierCurve(currentPosition.Value, new PdfPoint(x1, y1), new PdfPoint(x2, y2), to));
currentPosition = to;
}
else
{
// PDF Reference 1.7 p226
throw new ArgumentNullException("BezierCurveTo(): currentPosition is null.");
}
}
/// <summary>
/// Extract the <see cref="PageArea"/>, with its text elements (letters) and rulings (processed PdfPath and PdfSubpath).
/// </summary>
/// <param name="pageNumber">The page number to extract.</param>
public PageArea ExtractPage(int pageNumber)
{
if (pageNumber > this.pdfDocument.NumberOfPages || pageNumber < 1)
{
throw new IndexOutOfRangeException("Page number does not exist");
}
Page p = this.pdfDocument.GetPage(pageNumber);
//ObjectExtractorStreamEngine se = new ObjectExtractorStreamEngine(p);
//se.processPage(p);
/**************** ObjectExtractorStreamEngine(PDPage page)*******************/
var rulings = new List <Ruling>();
foreach (var image in p.GetImages())
{
if (image.TryGetPng(out var png))
{
}
}
foreach (var path in p.ExperimentalAccess.Paths)
{
if (!path.IsFilled && !path.IsStroked)
{
continue; // strokeOrFillPath operator => filter stroke and filled
}
foreach (var subpath in path)
{
if (!(subpath.Commands[0] is Move first))
{
// skip paths whose first operation is not a MOVETO
continue;
}
if (subpath.Commands.Any(c => c is BezierCurve))
{
// or contains operations other than LINETO, MOVETO or CLOSE
// bobld: skip at subpath or path level?
continue;
}
// TODO: how to implement color filter?
PdfPoint? start_pos = RoundPdfPoint(first.Location, rounding);
PdfPoint? last_move = start_pos;
PdfPoint? end_pos = null;
PdfLine line;
PointComparer pc = new PointComparer();
foreach (var command in subpath.Commands)
{
if (command is Line linePath)
{
end_pos = RoundPdfPoint(linePath.To, rounding);
if (!start_pos.HasValue || !end_pos.HasValue)
{
break;
}
line = pc.Compare(start_pos.Value, end_pos.Value) == -1 ? new PdfLine(start_pos.Value, end_pos.Value) : new PdfLine(end_pos.Value, start_pos.Value);
// already clipped
Ruling r = new Ruling(line.Point1, line.Point2);
if (r.Length > 0.01)
{
rulings.Add(r);
}
}
else if (command is Move move)
{
start_pos = RoundPdfPoint(move.Location, rounding);
end_pos = start_pos;
}
else if (command is Close)
{
// according to PathIterator docs:
// "the preceding subpath should be closed by appending a line
// segment
// back to the point corresponding to the most recent
// SEG_MOVETO."
if (!start_pos.HasValue || !end_pos.HasValue)
{
break;
}
line = pc.Compare(end_pos.Value, last_move.Value) == -1 ? new PdfLine(end_pos.Value, last_move.Value) : new PdfLine(last_move.Value, end_pos.Value);
// already clipped
Ruling r = new Ruling(line.Point1, line.Point2); //.intersect(this.currentClippingPath());
if (r.Length > 0.01)
{
rulings.Add(r);
}
}
start_pos = end_pos;
}
}
}
/****************************************************************************/
TextStripper pdfTextStripper = new TextStripper(this.pdfDocument, pageNumber);
pdfTextStripper.Process();
Utils.Sort(pdfTextStripper.textElements, new TableRectangle.ILL_DEFINED_ORDER());
return(new PageArea(p.CropBox.Bounds,
p.Rotation.Value,
pageNumber,
p,
this.pdfDocument,
pdfTextStripper.textElements,
rulings,
pdfTextStripper.minCharWidth,
pdfTextStripper.minCharHeight,
pdfTextStripper.spatialIndex));
}
internal void MoveTo(decimal x, decimal y)
{
currentPosition = new PdfPoint(x, y);
commands.Add(new Move(currentPosition.Value));
}
/// <summary>
/// Add a <see cref="Move"/> command to the path.
/// </summary>
public void MoveTo(double x, double y)
{
currentPosition = new PdfPoint(x, y);
commands.Add(new Move(currentPosition.Value));
}
internal void MoveTo(decimal x, decimal y)
{
currentPosition = currentTransformationMatrix.Transform(new PdfPoint(x, y));
commands.Add(new Move(currentPosition.Value));
}