Пример #1
0
 public Searcher(bool stem)
 {
     stemmed  = stem;
     stemming = new StemmingSequence();
     parser   = new Parser(StartDialog.stopWordsPath, false);
     ranker   = new Ranker(StartDialog.indexPath, stemmed, @"C:\Users\welta\Desktop\New folder\glove.840B.300d\glove.840B.300d.vec" /*@"glove.6B.100dc.vec"*/);
 }
Пример #2
0
        /**
         * unused function for testing puposes
         */
        private void RunSingleThread()
        {
            var              watch   = System.Diagnostics.Stopwatch.StartNew();
            FileReader       fr      = new FileReader(this.path_to_corpus_show.Text);
            Parser           pr      = new Parser(this.path_to_stopfile_show.Text, true);
            StemmingSequence stemmer = new StemmingSequence();
            Indexer          ind     = new Indexer(this.path_to_index_show.Text);

            while (fr.hasNext())
            {
                Document[] docAr = fr.GetNextDocuments();
                for (int i = 0; i < docAr.Length; i++)
                {
                    Token[] parseResult = pr.processDoc(docAr[i]);
                    if (this.doStemming.Checked)
                    {
                        parseResult = stemmer.StemTokens(parseResult);
                    }
                    ind.ProcessBatch(parseResult);
                    Console.WriteLine(docAr[i].id);
                }
            }
            watch.Stop();
            Console.WriteLine("Total time: " + watch.ElapsedMilliseconds);
        }
Пример #3
0
        /**
         * function used to create calculation threads
         * executes all IO intencive tasks such as parsing and stemming
         * mid pipeline thread
         */
        private void CalculationTask(String stopWordsPath, ConcurrentQueue <Document[]> docQ, ConcurrentQueue <Token[]> tokenQ, Boolean doStemming)
        {
            Parser           parser  = new Parser(stopWordsPath, true);
            StemmingSequence stemmer = new StemmingSequence();

            while (!docQ.IsEmpty || !StartDialog.readFinished)
            {
                if (docQ.Count > 1000)
                {
                    Thread.Sleep(500);
                }
                Document[] docAr;
                StartDialog.qLok.WaitOne();
                bool ok = docQ.TryDequeue(out docAr);
                StartDialog.qLok.ReleaseMutex();
                if (!ok)
                {
                    continue;
                }
                for (int i = 0; i < docAr.Length; i++)
                {
                    Token[] parseResult = parser.processDoc(docAr[i]);
                    if (doStemming)
                    {
                        parseResult = stemmer.StemTokens(parseResult);
                    }
                    tokenQ.Enqueue(parseResult);
                    Console.WriteLine("Doc done :" + docAr[i].id);
                }
            }
            Console.WriteLine("Processing Complete");
            StartDialog.calculationsFinished = true;
        }
Пример #4
0
 public Searcher(bool stem)
 {
     stemmed  = stem;
     stemming = new StemmingSequence();
     parser   = new Parser(StartDialog.stopWordsPath, false);
     ranker   = new Ranker(StartDialog.indexPath, stemmed, @"glove.6B.100dc.vec");
 }