Exemplo n.º 1
0
        public static void Main(string[] args)
        {
            if (args.Length != 2)
            {
                Console.WriteLine("usage: Spimi <folderpath> <DestinationIndexFilePath>");
                return;
            }
            string directory = args[0];
            string indexFilePath = args[1];
            SpimiIndexer indexer = new SpimiIndexer(new BasicLexer());

            DirectoryInfo dir = new DirectoryInfo(directory);
            foreach(FileInfo file in dir.GetFiles().Where(f => f.Extension.Equals(".sgm")))
                indexer.Index(file.FullName, file.Open(FileMode.Open));

            using (FileStream indexFileStream = File.Open(indexFilePath, FileMode.CreateNew))
            {
                indexer.CreateIndex(indexFileStream);
                FileIndex index = FileIndex.Open(indexFileStream);
                QueryEngine queryEngine = new QueryEngine(index);
                while (true)
                {
                    Console.Write("> ");
                    string query = Console.ReadLine();
                    foreach (string docId in queryEngine.Query(query.ToLower()))
                    {
                        Console.WriteLine(docId);
                    }
                }
            }
        }
Exemplo n.º 2
0
 public void testQuery()
 {
     QueryEngine engine = new QueryEngine(index, metadata);
     IList<long> foundPostings = engine.Query("foo bar", RankingMode.TFIDF);
     IList<Posting> expectedPostings = postingsWithFoo.Union(postingsWithBar).ToList();
     foreach (Posting posting in expectedPostings)
     {
         Assert.IsTrue(foundPostings.Contains(posting.DocumentId));
     }
 }
Exemplo n.º 3
0
    public IList<QueryResult> Query(string query, RankingMode rankingMode)
    {
        using (FileStream indexFileStream = File.Open(indexFilePath, FileMode.Open))
        {
            using (FileStream metadataFileStream = File.Open(metadataFilePath, FileMode.Open))
            {
                IndexMetadata indexMetadata = new IndexMetadata(metadataFileStream);
                TermIndex index = new TermIndex(indexFileStream);
                QueryEngine queryEngine = new QueryEngine(index, indexMetadata);

                IList<long> results = queryEngine.Query(query.ToLower(), rankingMode);
                IList<QueryResult> queryResults = new List<QueryResult>();

                int i = 1;
                Console.WriteLine("rank\tscore\ttitle");
                foreach (long docId in results.Take(500))
                {
                    DocumentInfo docInfo;
                    if (indexMetadata.TryGetDocumentInfo(docId, out docInfo))
                    {
                        QueryResult res = new QueryResult()
                        {
                            Title = docInfo.Title,
                            Uri = docInfo.Uri,
                            Score = queryEngine.Scores[docId]
                        };
                        queryResults.Add(res);
                    }
                    else
                    {
                        Console.WriteLine("Found document id in posting list that wasn't indexed in metadata: " + docId);
                    }
                }

                return queryResults;
            }
        }
    }