lsm/PdfPig
1
0
Fork 0
mirror of https://github.com/UglyToad/PdfPig.git synced 2025-10-15 19:54:52 +08:00
Code Issues Actions 5 Packages Projects Releases Wiki Activity

Adding handling when pageWords count = 0 for IPageSegmenters

Browse Source
This commit is contained in:
BobLd
2019-09-04 22:14:08 +01:00
parent 68e04603c0
commit d36dee0e25
2 changed files with 6 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
src/UglyToad.PdfPig/DocumentLayoutAnalysis/DocstrumBB.cs
View File

@@ -5,6 +5,7 @@ using System.Linq;
using System.Threading.Tasks;
using UglyToad.PdfPig.Content;
using UglyToad.PdfPig.Geometry;
using UglyToad.PdfPig.Util;
namespace UglyToad.PdfPig.DocumentLayoutAnalysis
{
@@ -47,6 +48,8 @@ namespace UglyToad.PdfPig.DocumentLayoutAnalysis
public IReadOnlyList<TextBlock> GetBlocks(IEnumerable<Word> pageWords, double wlAngleLB, double wlAngleUB,
double blAngleLB, double blAngleUB, double blMultiplier)
{
if (pageWords.Count() == 0) return EmptyArray<TextBlock>.Instance;
var pageWordsArr = pageWords.Where(w => !string.IsNullOrWhiteSpace(w.Text)).ToArray(); // remove white spaces
var withinLineDistList = new ConcurrentBag<double[]>();

3
src/UglyToad.PdfPig/DocumentLayoutAnalysis/RecursiveXYCut.cs
View File

@@ -2,6 +2,7 @@
using System.Collections.Generic;
using System.Linq;
using UglyToad.PdfPig.Content;
using UglyToad.PdfPig.Util;
namespace UglyToad.PdfPig.DocumentLayoutAnalysis
{
@@ -64,6 +65,8 @@ namespace UglyToad.PdfPig.DocumentLayoutAnalysis
Func<IEnumerable<decimal>, decimal> dominantFontWidthFunc,
Func<IEnumerable<decimal>, decimal> dominantFontHeightFunc)
{
if (pageWords.Count() == 0) return EmptyArray<TextBlock>.Instance;
XYLeaf root = new XYLeaf(pageWords); // Create a root node.
XYNode node = VerticalCut(root, minimumWidth, dominantFontWidthFunc, dominantFontHeightFunc);