Search text and get the bound boxes

Message Topic Search Topic Options Post Reply Create New Topic Printable Version Translate Topic

   from our user Pinozzy:Thanks for sharing!

I'm using the Viewer SDK (with c#) for search some text in a document. 
I need to localize that text and get the co-ordinates of the single results. 

My approach is now this: 

SearchPDFText("my string");

what I obtain is the number of the occurrences. 
Now I need the bound rectangles of theese occurrences. 

How can I do that?

. . .

This is how I solved my problem. 
This method returns results from a search by Regex.

Bye!

----

public struct FindResult
    {
      public string Text { get; set; }
      public RectangleF Rectangle { get; set; }
      public int Page { get; set; }
    }

    public override List<FindResult> SearchPattern(int pageIndex, string pattern)
    {
      var retVal = new List<FindResult>();
      var dpl = Document.DPL;
      dpl.SetTextExtractionWordGap(1);
      dpl.SetTextExtractionOptions(3, 0);
      var regex = new Regex(pattern);
      for (var i = 0; i < Pages; i++)
      {
        if(pageIndex > 0 && (pageIndex - 1) != i) continue; 
        var id = dpl.ExtractPageTextBlocks(4);
        dpl.SelectPage(i);
        for (var f = 1; f <= dpl.GetTextBlockCount(id); f++)
        {
          var text = dpl.GetTextBlockText(id, f);
          var match = regex.Match(text);
          if (!match.Success) continue;
          var res = new FindResult
          {
            Rectangle = new RectangleF(
              (float)dpl.GetTextBlockBound(id, f, 7), 
              (float)dpl.GetTextBlockBound(id, f, 8),
              (float)dpl.GetTextBlockBound(id, f, 5) - (float)dpl.GetTextBlockBound(id, f, 7),
              (float)dpl.GetTextBlockBound(id, f, 6) - (float)dpl.GetTextBlockBound(id, f, 4)
            ),
            Page = i+1, Text = text
          };
          retVal.Add(res);
        }
        dpl.ReleaseTextBlocks(id);
      }
      return retVal;
    }

Edited by Ingo - 27 Mar 16 at 1:44PM

Author	Message Topic Search Topic Options Post Reply Create New Topic Printable Version Translate Topic
Ingo Members Profile Find Members Posts Moderator Group Joined: 29 Oct 05 Status: Offline Points: 3527	Post Options Post Reply Quote Ingo Report Post Thanks(0) Quote Reply Topic: Search text and get the bound boxes Posted: 27 Mar 16 at 1:41PM
	from our user Pinozzy: Thanks for sharing! I'm using the Viewer SDK (with c#) for search some text in a document. I need to localize that text and get the co-ordinates of the single results. My approach is now this: SearchPDFText("my string"); what I obtain is the number of the occurrences. Now I need the bound rectangles of theese occurrences. How can I do that? . . . This is how I solved my problem. This method returns results from a search by Regex. Bye! ---- public struct FindResult { public string Text { get; set; } public RectangleF Rectangle { get; set; } public int Page { get; set; } } public override List<FindResult> SearchPattern(int pageIndex, string pattern) { var retVal = new List<FindResult>(); var dpl = Document.DPL; dpl.SetTextExtractionWordGap(1); dpl.SetTextExtractionOptions(3, 0); var regex = new Regex(pattern); for (var i = 0; i < Pages; i++) { if(pageIndex > 0 && (pageIndex - 1) != i) continue; var id = dpl.ExtractPageTextBlocks(4); dpl.SelectPage(i); for (var f = 1; f <= dpl.GetTextBlockCount(id); f++) { var text = dpl.GetTextBlockText(id, f); var match = regex.Match(text); if (!match.Success) continue; var res = new FindResult { Rectangle = new RectangleF( (float)dpl.GetTextBlockBound(id, f, 7), (float)dpl.GetTextBlockBound(id, f, 8), (float)dpl.GetTextBlockBound(id, f, 5) - (float)dpl.GetTextBlockBound(id, f, 7), (float)dpl.GetTextBlockBound(id, f, 6) - (float)dpl.GetTextBlockBound(id, f, 4) ), Page = i+1, Text = text }; retVal.Add(res); } dpl.ReleaseTextBlocks(id); } return retVal; } Edited by Ingo - 27 Mar 16 at 1:44PM
	Cheers, Ingo