public void Eu004()
{
using (PdfDocument document = PdfDocument.Open("Resources/icdar2013-dataset/competition-dataset-eu/eu-004.pdf", new ParsingOptions()
{
ClipPaths = true
}))
{
ObjectExtractor oe = new ObjectExtractor(document);
PageArea page = oe.Extract(3);
var detector = new SimpleNurminenDetectionAlgorithm();
var regions = detector.Detect(page);
var newArea = page.GetArea(regions[0].BoundingBox);
var sea = new SpreadsheetExtractionAlgorithm();
var tables = sea.Extract(newArea);
/*
* var detector = new SimpleNurminenDetectionAlgorithm();
* var regions = detector.Detect(page);
*
* foreach (var a in regions)
* {
* IExtractionAlgorithm ea = new BasicExtractionAlgorithm();
* var newArea = page.GetArea(a.BoundingBox);
* List<Table> tables = ea.Extract(newArea);
* }
*/
}
}
public void TestLinesToCells()
{
using (PdfDocument document = PdfDocument.Open("test3.pdf", new ParsingOptions()
{
ClipPaths = true
}))
{
ObjectExtractor oe = new ObjectExtractor(document);
PageArea page = oe.Extract(1);
SimpleNurminenDetectionAlgorithm detector = new SimpleNurminenDetectionAlgorithm();
var regions = detector.Detect(page);
foreach (var a in regions)
{
IExtractionAlgorithm ea = new BasicExtractionAlgorithm();
var newArea = page.GetArea(a.BoundingBox);
List <Table> tables = ea.Extract(newArea);
}
}
}