lsm/PdfPig
1
0
Fork 0
mirror of https://github.com/UglyToad/PdfPig.git synced 2025-09-18 18:27:55 +08:00
Code Issues Actions 5 Packages Projects Releases Wiki Activity

Remove 'orderFunc' from 'NearestNeighbourWordExtractor' to use the order found by clustering algo

Browse Source
This commit is contained in:
BobLd
2020-01-19 15:22:49 +00:00
committed by Eliot Jones
parent 3b90370f28
commit 0cbf3434bc
2 changed files with 16 additions and 24 deletions
Download Patch File Download Diff File Expand all files Collapse all files

23
src/UglyToad.PdfPig.DocumentLayoutAnalysis/ClusteringAlgorithms.cs
View File

@@ -292,26 +292,33 @@
/// Depth-first search
/// <para>https://en.wikipedia.org/wiki/Depth-first_search</para>
/// </summary>
private static HashSet<int> DfsIterative(int c, int[][] adj, ref bool[] isDone)
private static HashSet<int> DfsIterative(int s, int[][] adj, ref bool[] isDone)
{
HashSet<int> group = new HashSet<int>();
Stack<int> S = new Stack<int>();
S.Push(c);
S.Push(s);
while (S.Any())
{
var v = S.Pop();
if (!isDone[v])
var u = S.Pop();
if (!isDone[u])
{
group.Add(v);
isDone[v] = true;
foreach (var w in adj[v])
group.Add(u);
isDone[u] = true;
foreach (var v in adj[u])
{
S.Push(w);
S.Push(v);
}
}
}
return group;
}
public class DfsEdge : Tuple<int, int>
{
public DfsEdge(int item1, int item2) : base(item1, item2)
{
}
}
}
}

17
src/UglyToad.PdfPig.DocumentLayoutAnalysis/WordExtractor/NearestNeighbourWordExtractor.cs
View File

@@ -97,21 +97,6 @@
throw new ArgumentException("NearestNeighbourWordExtractor.GetWords(): Mixed Text Direction.");
}
// TO DO: orderFunc should also take in account the edge relationships found by 'ClusterNearestNeighbours'
Func<IEnumerable<Letter>, IReadOnlyList<Letter>> orderFunc = l => l.OrderBy(x => x.GlyphRectangle.Left).ToList();
if (textDirection == TextDirection.Rotate180)
{
orderFunc = l => l.OrderByDescending(x => x.GlyphRectangle.Right).ToList();
}
else if (textDirection == TextDirection.Rotate90)
{
orderFunc = l => l.OrderByDescending(x => x.GlyphRectangle.Top).ToList();
}
else if (textDirection == TextDirection.Rotate270)
{
orderFunc = l => l.OrderBy(x => x.GlyphRectangle.Bottom).ToList();
}
Letter[] letters = pageLetters.ToArray();
var groupedIndexes = ClusteringAlgorithms.ClusterNearestNeighbours(letters,
@@ -124,7 +109,7 @@
List<Word> words = new List<Word>();
for (int a = 0; a < groupedIndexes.Count(); a++)
{
words.Add(new Word(orderFunc(groupedIndexes[a].Select(i => letters[i]))));
words.Add(new Word(groupedIndexes[a].Select(i => letters[i]).ToList()));
}
return words;