Beispiel #1
0
 /**
  * Prepare a document reconstructor.
  *
  * @param reader IndexReader to read from.
  * @param fieldNames if non-null or not empty, data will be collected only from these fields,
  *     otherwise data will be collected from all fields
  * @param numTerms total number of terms in the index, or -1 if unknown (will be calculated)
  * @throws Exception
  */
 public DocReconstructor(IndexReader reader, String[] fieldNames, int numTerms) throws Exception {
   if (reader == null) {
     throw new Exception("IndexReader cannot be null.");
   }
   this.reader = reader;
   if (fieldNames == null || fieldNames.length == 0) {
     // collect fieldNames
     this.fieldNames = (String[]) reader.getFieldNames(FieldOption.ALL).toArray(new String[0]);
   } else {
     this.fieldNames = fieldNames;
   }
   if (numTerms == -1) {
     Fields fields = MultiFields.getFields(reader);
     numTerms = 0;
     FieldsEnum fe = fields.iterator();
     String fld = null;
     while ((fld = fe.next()) != null) {
       TermsEnum te = fe.terms();
       while (te.next() != null) {
         numTerms++;
       }
     }
     this.numTerms = numTerms;
   }
   deleted = MultiFields.getDeletedDocs(reader);
 }
  /*
   *  listTermVectors displays the term vectors for all of the fields
   *  in a document in an index (specified by reader).
   */
  static void listTermVectors(IndexReader reader, String docidString) throws IOException {

    System.out.println("\nTermVector:  docid " + docidString);

    int docid = Integer.parseInt(docidString);

    if ((docid < 0) || (docid >= reader.numDocs())) {
      System.out.println("ERROR:  " + docidString + " is a bad document id.");
      return;
    }
    ;

    /*
     *  Iterate over the fields in this document.
     */
    Fields fields = reader.getTermVectors(docid);
    Iterator<String> fieldIterator = fields.iterator();

    while (fieldIterator.hasNext()) {
      String fieldName = fieldIterator.next();
      System.out.println("  Field: " + fieldName);

      Terms terms = fields.terms(fieldName);
      termVectorDisplay(terms);
    }
    ;
  }
  public static void main(String[] args) throws IOException {

    IndexReader reader = null;

    /*
     *  Opening the index first simplifies the processing of the
     *  rest of the command line arguments.
     */
    for (int i = 0; i < args.length; i++) {
      if (("-index".equals(args[i])) && ((i + 1) < args.length)) {
        reader = DirectoryReader.open(FSDirectory.open(new File(args[i + 1])));

        if (reader == null) {
          System.err.println("Error:  Can't open index " + args[i + 1]);
          System.exit(1);
        }
        ;

        break;
      }
      ;
    }
    ;

    if (reader == null) {
      System.err.println(usage);
      System.exit(1);
    }
    ;

    /*
     *  Process the command line arguments sequentially.
     */
    for (int i = 0; i < args.length; i++) {

      if ("-index".equals(args[i])) {

        /*
         *  Handled in the previous loop, so just skip the argument.
         */
        i++;

      } else if ("-list-edocid".equals(args[i])) {

        System.out.println("-list-edocid:");

        if ((i + 1) >= args.length) {
          System.out.println(usage);
          break;
        }
        ;

        Document d = reader.document(Integer.parseInt(args[i + 1]));

        System.out.println(
            "Internal docid --> External docid: " + args[i + 1] + " --> " + d.get("externalId"));

        i += 1;
      } else if ("-list-docids".equals(args[i])) {

        System.out.println("-list-docids:");

        for (int j = 0; j < reader.numDocs(); j++) {
          Document d = reader.document(j);
          System.out.println("Internal --> external docid: " + j + " --> " + d.get("externalId"));
        }
        ;

      } else if ("-list-fields".equals(args[i])) {

        Fields fields = MultiFields.getFields(reader);

        System.out.print("\nNumber of fields:  ");

        if (fields == null) System.out.println("0");
        else {
          System.out.println(fields.size());

          Iterator<String> is = fields.iterator();

          while (is.hasNext()) {
            System.out.println("\t" + is.next());
          }
          ;
        }
        ;

      } else if ("-list-postings".equals(args[i])) {

        if ((i + 2) >= args.length) {
          System.out.println(usage);
          break;
        }
        ;

        listPostings(reader, args[i + 1], args[i + 2], Integer.MAX_VALUE);
        i += 2;

      } else if ("-list-postings-sample".equals(args[i])) {

        if ((i + 2) >= args.length) {
          System.out.println(usage);
          break;
        }
        ;

        listPostings(reader, args[i + 1], args[i + 2], 5);
        i += 2;

      } else if ("-list-stats".equals(args[i])) {

        System.out.println("Corpus statistics:");
        System.out.println("\tnumdocs\t\t" + reader.numDocs());
        System.out.println(
            "\turl:\t"
                + "\tnumdocs="
                + reader.getDocCount("url")
                + "\tsumTotalTF="
                + reader.getSumTotalTermFreq("url")
                + "\tavglen="
                + reader.getSumTotalTermFreq("url") / (float) reader.getDocCount("url"));

        System.out.println(
            "\tkeywords:"
                + "\tnumdocs="
                + reader.getDocCount("keywords")
                + "\tsumTotalTF="
                + reader.getSumTotalTermFreq("keywords")
                + "\tavglen="
                + reader.getSumTotalTermFreq("keywords") / (float) reader.getDocCount("keywords"));

        System.out.println(
            "\ttitle:\t"
                + "\tnumdocs="
                + reader.getDocCount("title")
                + "\tsumTotalTF="
                + reader.getSumTotalTermFreq("title")
                + "\tavglen="
                + reader.getSumTotalTermFreq("title") / (float) reader.getDocCount("title"));

        System.out.println(
            "\tbody:\t"
                + "\tnumdocs="
                + reader.getDocCount("body")
                + "\tsumTotalTF="
                + reader.getSumTotalTermFreq("body")
                + "\tavglen="
                + reader.getSumTotalTermFreq("body") / (float) reader.getDocCount("body"));

        System.out.println(
            "\tinlink:\t"
                + "\tnumdocs="
                + reader.getDocCount("inlink")
                + "\tsumTotalTF="
                + reader.getSumTotalTermFreq("inlink")
                + "\tavglen="
                + reader.getSumTotalTermFreq("inlink") / (float) reader.getDocCount("inlink"));

      } else if ("-list-terms".equals(args[i])) {

        if ((i + 1) >= args.length) {
          System.out.println(usage);
          break;
        }
        ;

        listTermDictionary(reader, args[i + 1]);
        i += 1;

      } else if ("-list-termvector".equals(args[i])) {

        if ((i + 1) >= args.length) {
          System.out.println(usage);
          break;
        }
        ;

        listTermVectors(reader, args[i + 1]);
        i += 1;

      } else if ("-list-termvector-field".equals(args[i])) {

        if ((i + 2) >= args.length) {
          System.out.println(usage);
          break;
        }
        ;

        listTermVectorField(reader, args[i + 1], args[i + 2]);
        i += 2;

      } else System.err.println("\nWarning:  Unknown argument " + args[i] + " ignored.");
    }
    ;

    /*
     *  Close the index and exit gracefully.
     */
    reader.close();
  }