nutchanalysis.java

爬虫数据的改进,并修正了一些bug

    if (jj_scan_token(2)) {
    jj_scanpos = xsp;
    if (jj_scan_token(3)) return true;
    }
    }
    return false;
  }

  final private boolean jj_3R_20() {
    if (jj_3R_11()) return true;
    Token xsp;
    while (true) {
      xsp = jj_scanpos;
      if (jj_3R_24()) { jj_scanpos = xsp; break; }
    }
    return false;
  }

  final private boolean jj_3R_9() {
    if (jj_3R_15()) return true;
    return false;
  }

  final private boolean jj_3R_22() {
    if (jj_3R_17()) return true;
    return false;
  }

  final private boolean jj_3R_19() {
    if (jj_3R_18()) return true;
    return false;
  }

  final private boolean jj_3R_12() {
    if (jj_3R_17()) return true;
    return false;
  }

  final private boolean jj_3R_10() {
    if (jj_3R_16()) return true;
    return false;
  }

  final private boolean jj_3_2() {
    Token xsp;
    if (jj_3R_10()) return true;
    while (true) {
      xsp = jj_scanpos;
      if (jj_3R_10()) { jj_scanpos = xsp; break; }
    }
    if (jj_3R_11()) return true;
    return false;
  }

  final private boolean jj_3R_14() {
    if (jj_scan_token(QUOTE)) return true;
    Token xsp;
    while (true) {
      xsp = jj_scanpos;
      if (jj_3R_19()) { jj_scanpos = xsp; break; }
    }
    while (true) {
      xsp = jj_scanpos;
      if (jj_3R_20()) { jj_scanpos = xsp; break; }
    }
    if (jj_scan_token(QUOTE)) return true;
    return false;
  }

  final private boolean jj_3R_17() {
    Token xsp;
    xsp = jj_scanpos;
    if (jj_scan_token(10)) {
    jj_scanpos = xsp;
    if (jj_scan_token(11)) {
    jj_scanpos = xsp;
    if (jj_scan_token(12)) {
    jj_scanpos = xsp;
    if (jj_scan_token(13)) {
    jj_scanpos = xsp;
    if (jj_scan_token(14)) return true;
    }
    }
    }
    }
    return false;
  }

  final private boolean jj_3R_25() {
    if (jj_3R_16()) return true;
    return false;
  }

  final private boolean jj_3R_8() {
    if (jj_3R_14()) return true;
    return false;
  }

  final private boolean jj_3R_21() {
    Token xsp;
    if (jj_3R_25()) return true;
    while (true) {
      xsp = jj_scanpos;
      if (jj_3R_25()) { jj_scanpos = xsp; break; }
    }
    if (jj_3R_11()) return true;
    return false;
  }

  final private boolean jj_3R_16() {
    Token xsp;
    xsp = jj_scanpos;
    if (jj_scan_token(7)) {
    jj_scanpos = xsp;
    if (jj_scan_token(8)) {
    jj_scanpos = xsp;
    if (jj_3R_22()) return true;
    }
    }
    return false;
  }

  final private boolean jj_3R_15() {
    if (jj_3R_11()) return true;
    Token xsp;
    while (true) {
      xsp = jj_scanpos;
      if (jj_3R_21()) { jj_scanpos = xsp; break; }
    }
    return false;
  }

  final private boolean jj_3R_23() {
    if (jj_3R_16()) return true;
    return false;
  }

  final private boolean jj_3_3() {
    Token xsp;
    xsp = jj_scanpos;
    if (jj_scan_token(15)) {
    jj_scanpos = xsp;
    if (jj_3R_12()) {
    jj_scanpos = xsp;
    if (jj_3R_13()) return true;
    }
    }
    return false;
  }

  final private boolean jj_3_1() {
    if (jj_scan_token(WORD)) return true;
    if (jj_scan_token(COLON)) return true;
    Token xsp;
    xsp = jj_scanpos;
    if (jj_3R_8()) {
    jj_scanpos = xsp;
    if (jj_3R_9()) return true;
    }
    return false;
  }

  final private boolean jj_3R_13() {
    Token xsp;
    xsp = jj_scanpos;
    if (jj_scan_token(7)) {
    jj_scanpos = xsp;
    if (jj_scan_token(8)) return true;
    }
    if (jj_3R_18()) return true;
    return false;
  }

  final private boolean jj_3R_18() {
    Token xsp;
    xsp = jj_scanpos;
    if (jj_scan_token(15)) {
    jj_scanpos = xsp;
    if (jj_3R_23()) {
    jj_scanpos = xsp;
    if (jj_scan_token(0)) return true;
    }
    }
    return false;
  }

  public NutchAnalysisTokenManager token_source;
  public Token token, jj_nt;
  private int jj_ntk;
  private Token jj_scanpos, jj_lastpos;
  private int jj_la;
  public boolean lookingAhead = false;
  private boolean jj_semLA;
  private int jj_gen;
  final private int[] jj_la1 = new int[14];
  static private int[] jj_la1_0;
  static {
      jj_la1_0();
   }
   private static void jj_la1_0() {
      jj_la1_0 = new int[] {0x38e,0x180,0x180,0x20e,0xfd81,0xe,0xfd81,0x7d80,0xe,0xfd81,0x180,0xfd80,0x7d80,0x7c00,};
   }
  final private JJCalls[] jj_2_rtns = new JJCalls[3];
  private boolean jj_rescan = false;
  private int jj_gc = 0;

  public NutchAnalysis(CharStream stream) {
    token_source = new NutchAnalysisTokenManager(stream);
    token = new Token();
    jj_ntk = -1;
    jj_gen = 0;
    for (int i = 0; i < 14; i++) jj_la1[i] = -1;
    for (int i = 0; i < jj_2_rtns.length; i++) jj_2_rtns[i] = new JJCalls();
  }

  public void ReInit(CharStream stream) {
    token_source.ReInit(stream);
    token = new Token();
    jj_ntk = -1;
    jj_gen = 0;
    for (int i = 0; i < 14; i++) jj_la1[i] = -1;
    for (int i = 0; i < jj_2_rtns.length; i++) jj_2_rtns[i] = new JJCalls();
  }

  public NutchAnalysis(NutchAnalysisTokenManager tm) {
    token_source = tm;
    token = new Token();
    jj_ntk = -1;
    jj_gen = 0;
    for (int i = 0; i < 14; i++) jj_la1[i] = -1;
    for (int i = 0; i < jj_2_rtns.length; i++) jj_2_rtns[i] = new JJCalls();
  }

  public void ReInit(NutchAnalysisTokenManager tm) {
    token_source = tm;
    token = new Token();
    jj_ntk = -1;
    jj_gen = 0;
    for (int i = 0; i < 14; i++) jj_la1[i] = -1;
    for (int i = 0; i < jj_2_rtns.length; i++) jj_2_rtns[i] = new JJCalls();
  }

  final private Token jj_consume_token(int kind) throws ParseException {
    Token oldToken;
    if ((oldToken = token).next != null) token = token.next;
    else token = token.next = token_source.getNextToken();
    jj_ntk = -1;
    if (token.kind == kind) {
      jj_gen++;
      if (++jj_gc > 100) {
        jj_gc = 0;
        for (int i = 0; i < jj_2_rtns.length; i++) {
          JJCalls c = jj_2_rtns[i];
          while (c != null) {
            if (c.gen < jj_gen) c.first = null;
            c = c.next;
          }
        }
      }
      return token;
    }
    token = oldToken;
    jj_kind = kind;
    throw generateParseException();
  }

  static private final class LookaheadSuccess extends java.lang.Error { }
  final private LookaheadSuccess jj_ls = new LookaheadSuccess();
  final private boolean jj_scan_token(int kind) {
    if (jj_scanpos == jj_lastpos) {
      jj_la--;
      if (jj_scanpos.next == null) {
        jj_lastpos = jj_scanpos = jj_scanpos.next = token_source.getNextToken();
      } else {
        jj_lastpos = jj_scanpos = jj_scanpos.next;
      }
    } else {
      jj_scanpos = jj_scanpos.next;
    }
    if (jj_rescan) {
      int i = 0; Token tok = token;
      while (tok != null && tok != jj_scanpos) { i++; tok = tok.next; }
      if (tok != null) jj_add_error_token(kind, i);
    }
    if (jj_scanpos.kind != kind) return true;
    if (jj_la == 0 && jj_scanpos == jj_lastpos) throw jj_ls;
    return false;
  }

  final public Token getNextToken() {
    if (token.next != null) token = token.next;
    else token = token.next = token_source.getNextToken();
    jj_ntk = -1;
    jj_gen++;
    return token;
  }

  final public Token getToken(int index) {
    Token t = lookingAhead ? jj_scanpos : token;
    for (int i = 0; i < index; i++) {
      if (t.next != null) t = t.next;
      else t = t.next = token_source.getNextToken();
    }
    return t;
  }

  final private int jj_ntk() {
    if ((jj_nt=token.next) == null)
      return (jj_ntk = (token.next=token_source.getNextToken()).kind);
    else
      return (jj_ntk = jj_nt.kind);
  }

  private java.util.Vector jj_expentries = new java.util.Vector();
  private int[] jj_expentry;
  private int jj_kind = -1;
  private int[] jj_lasttokens = new int[100];
  private int jj_endpos;

  private void jj_add_error_token(int kind, int pos) {
    if (pos >= 100) return;
    if (pos == jj_endpos + 1) {
      jj_lasttokens[jj_endpos++] = kind;
    } else if (jj_endpos != 0) {
      jj_expentry = new int[jj_endpos];
      for (int i = 0; i < jj_endpos; i++) {
        jj_expentry[i] = jj_lasttokens[i];
      }
      boolean exists = false;
      for (java.util.Enumeration e = jj_expentries.elements(); e.hasMoreElements();) {
        int[] oldentry = (int[])(e.nextElement());
        if (oldentry.length == jj_expentry.length) {
          exists = true;
          for (int i = 0; i < jj_expentry.length; i++) {
            if (oldentry[i] != jj_expentry[i]) {
              exists = false;
              break;
            }
          }
          if (exists) break;
        }
      }
      if (!exists) jj_expentries.addElement(jj_expentry);
      if (pos != 0) jj_lasttokens[(jj_endpos = pos) - 1] = kind;
    }
  }

  public ParseException generateParseException() {
    jj_expentries.removeAllElements();
    boolean[] la1tokens = new boolean[20];
    for (int i = 0; i < 20; i++) {
      la1tokens[i] = false;
    }
    if (jj_kind >= 0) {
      la1tokens[jj_kind] = true;
      jj_kind = -1;
    }
    for (int i = 0; i < 14; i++) {
      if (jj_la1[i] == jj_gen) {
        for (int j = 0; j < 32; j++) {
          if ((jj_la1_0[i] & (1<<j)) != 0) {
            la1tokens[j] = true;
          }
        }
      }
    }
    for (int i = 0; i < 20; i++) {
      if (la1tokens[i]) {
        jj_expentry = new int[1];
        jj_expentry[0] = i;
        jj_expentries.addElement(jj_expentry);
      }
    }
    jj_endpos = 0;
    jj_rescan_token();
    jj_add_error_token(0, 0);
    int[][] exptokseq = new int[jj_expentries.size()][];
    for (int i = 0; i < jj_expentries.size(); i++) {
      exptokseq[i] = (int[])jj_expentries.elementAt(i);
    }
    return new ParseException(token, exptokseq, tokenImage);
  }

  final public void enable_tracing() {
  }

  final public void disable_tracing() {
  }

  final private void jj_rescan_token() {
    jj_rescan = true;
    for (int i = 0; i < 3; i++) {
      JJCalls p = jj_2_rtns[i];
      do {
        if (p.gen > jj_gen) {
          jj_la = p.arg; jj_lastpos = jj_scanpos = p.first;
          switch (i) {
            case 0: jj_3_1(); break;
            case 1: jj_3_2(); break;
            case 2: jj_3_3(); break;
          }
        }
        p = p.next;
      } while (p != null);
    }
    jj_rescan = false;
  }

  final private void jj_save(int index, int xla) {
    JJCalls p = jj_2_rtns[index];
    while (p.gen > jj_gen) {
      if (p.next == null) { p = p.next = new JJCalls(); break; }
      p = p.next;
    }
    p.gen = jj_gen + xla - jj_la; p.first = token; p.arg = xla;
  }

  static final class JJCalls {
    int gen;
    Token first;
    int arg;
    JJCalls next;
  }

}