public DocumentWrapper addDocument(String text, String title, String path, String date)
{
Document doc = DatabaseAdaptor.addDocument(path, date, title);
DocumentWrapper wrapper = new DocumentWrapper(doc, new TextDocument(text, path, title, date));
docs.Add(wrapper);
return(wrapper);
}
public void clearAll()
{
DatabaseAdaptor.clearAllDocumentWordRelations();
DatabaseAdaptor.clearAllWords();
DatabaseAdaptor.clearAllDocuments();
int docCnt = DatabaseAdaptor.container.DocumentSet.Count();
int wordCnt = DatabaseAdaptor.container.WordSet.Count();
int relCnt = DatabaseAdaptor.container.DocumentWordRelationshipSet.Count();
}
public List <DocumentWrapper> getAllDocuments()
{
var list = DatabaseAdaptor.getAllDocuments();
List <DocumentWrapper> ret = new List <DocumentWrapper>();
foreach (Document doc in list)
{
ret.Add(getDocumentWrapper(doc));
}
return(ret);
}
public List <DocumentWrapper> getDocumentsContainingWord(String word)
{
List <DocumentWordRelationship> list = DatabaseAdaptor.getDocumentsContainingWord(word);
List <DocumentWrapper> ret = new List <DocumentWrapper>();
foreach (DocumentWordRelationship rel in list)
{
ret.Add(getDocumentWrapper(rel.Document));
}
return(ret);
}
private IndexManager()
{
//to force reindexing, uncomment this line
//this.clearAll();
String datapath = "D:\\Data\\original";
docs = new List <DocumentWrapper>();
List <TextDocument> txtdocs = DirectoryReader.getDocuments(datapath);
List <Document> dbDocs = DatabaseAdaptor.getAllDocuments();
bool isAllIndexed = true;
//for each text document, search in the db for a document with the same path. if all documents are in the
// database, then no need to reindex.
foreach (TextDocument txtDoc in txtdocs)
{
// finding a document in the db with the same path means it is indexed
Document doc = DatabaseAdaptor.findDocumentByPath(txtDoc.getPath());
if (doc == null)
{
isAllIndexed = false;
break;
}
else
{
docs.Add(new DocumentWrapper(doc, txtDoc));
}
}
//there still a single case, if the database contains documents more than what we want to index, then reindex.
if (isAllIndexed == false || dbDocs.Count > txtdocs.Count)
{
this.clearAll();
this.docs = new List <DocumentWrapper>();
foreach (TextDocument txtDoc in txtdocs)
{
DocumentWrapper wrapper = this.addDocument(txtDoc.getText(), txtDoc.getTitle(), txtDoc.getPath(), txtDoc.getDate());
this.indexDocument(wrapper);
}
}
// in the following case, all documents are already indexed, so just load them
}
public void indexDocument(DocumentWrapper doc)
{
List <DocumentWord> docWords = doc.getTextDocument().getDocumentWords();
int all = docWords.Count;
int sofar = 0;
foreach (DocumentWord word in docWords)
{
if (doc.getDocument() == null)
{
var df = 1;
}
Word w = DatabaseAdaptor.addOrGetWord(word.getWord());
DatabaseAdaptor.setWordFreqInDocument(doc.getDocument(), w, word.getWordFreq(), true);
sofar++;
}
DatabaseAdaptor.saveChanges();
}
public int getDocumentFryequency(String w)
{
return(DatabaseAdaptor.getDocumentFrequency(w));
}
public int getTotalWordOccurences(String w)
{
return(DatabaseAdaptor.getTotalWordOccurences(w));
}