private void AddPdfModel( string path, ClassifiedPdfModel newModel )
{
doneClassifications[ path ] = newModel;
}
void DoWorkClassifyAll( object sender, DoWorkEventArgs e )
{
BackgroundWorker worker = sender as BackgroundWorker;
string directory = e.Argument as string;
var dirInfo = new DirectoryInfo( directory );
FileInfo[] infos = dirInfo.GetFiles( "*.pdf" );
var numPDFs = infos.Length;
int progress = 0;
foreach ( var fileInfo in infos )
{
string fullPath = classifiedDocuments.MakePath( directory, fileInfo.Name );
string extension = System.IO.Path.GetExtension( fullPath );
if ( extension == ".pdf" )
{
org.pdfclown.files.File file = new org.pdfclown.files.File( fullPath );
document = file.Document;
if ( document == null )
continue;
try
{
string abstractFileName = fullPath.Replace(".pdf", ".abstract");
string abstractFileContent = System.IO.File.ReadAllText(abstractFileName);
abstractContent = abstractFileContent;
}
catch (Exception exception)
{
abstractContent = null;
Utility.Log("file " + fullPath + " loaded without an abstract...");
}
Utility.Log("Classifing " + fullPath);
var classificationResult = classifier.Classify( document, abstractContent );
if (classificationResult.DocumentClassificationRestult != null)
{
ClassifiedPdfModel newModel = new ClassifiedPdfModel();
foreach ( var documentClassResult in classificationResult.DocumentClassificationRestult )
{
ClassifiedItem newItem = new ClassifiedItem();
newItem.Category = documentClassResult.First as string;
newItem.Compatibility = documentClassResult.Second.ToString();
// abstract classification result
if (classificationResult.AbstractClassificationRestult != null)
{
Double abstractCompatibilty;
if (classificationResult.AbstractClassificationRestult.TryGetValue(documentClassResult.First, out abstractCompatibilty))
{
newItem.CompatibilityAbstract = abstractCompatibilty.ToString();
}
}
newModel.AddClassificationData( newItem );
}
classifiedDocuments.AddPdfModel( directory, fileInfo.Name, newModel );
}
worker.ReportProgress( (int)( 100 * progress++ / (double)numPDFs ) );
}
}
}
public void AddPdfModel( string directory, string fileName, ClassifiedPdfModel newModel )
{
doneClassifications [ MakePath( directory, fileName ) ] = newModel;
}
public void onClassificationFinish(object classificationRes, string name)
{
Classifier.ClassificationResult classificationResult = classificationRes as Classifier.ClassificationResult;
SetViewEnabled(true);
//LabelWait.Content = "";
if (classificationResult.DocumentClassificationRestult != null)
{
LabelClassifiedAs.Content = "classified as:";
ClassifiedPdfModel newModel = new ClassifiedPdfModel();
foreach (var documentClassResult in classificationResult.DocumentClassificationRestult)
{
ClassifiedItem newItem = new ClassifiedItem();
newItem.Category = documentClassResult.First as string;
newItem.Compatibility = documentClassResult.Second.ToString();
// abstract classification result
if (classificationResult.AbstractClassificationRestult != null)
{
Double abstractCompatibilty;
if (classificationResult.AbstractClassificationRestult.TryGetValue(documentClassResult.First, out abstractCompatibilty))
{
newItem.CompatibilityAbstract = abstractCompatibilty.ToString();
}
}
newModel.AddClassificationData( newItem );
//ListBoxClassificationResult.Items.Add(i + ". " + classResult.First + "\t\t\t\t" + classResult.Second);
}
classifiedDocuments.AddPdfModel( pdfsDirectoryTextBox.Text, name, newModel );
ListBoxClassificationResult.DataContext = newModel;
}
else
{
LabelClassifiedAs.Content = "classification failed";
}
}