Package org.apache.nutch.searcher.Query

Examples of org.apache.nutch.searcher.Query.Clause

99
100
101
102
103
104
105
106
107
108
109
110
111
  /** Run all defined filters. */
  public BooleanQuery filter(Query input) throws QueryException {
    // first check that all field names are claimed by some plugin
    Clause[] clauses = input.getClauses();
    for (int i = 0; i < clauses.length; i++) {
      Clause c = clauses[i];
      if (!isField(c.getField()))
        throw new QueryException("Not a known field name:"+c.getField());
    }

    // then run each plugin
    BooleanQuery output = new BooleanQuery();
    for (int i = 0; i < this.queryFilters.length; i++) {
View Full Code Here
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
    throws QueryException {

    // examine each clause in the Nutch query
    Clause[] clauses = input.getClauses();
    for (int i = 0; i < clauses.length; i++) {
      Clause c = clauses[i];

      // skip non-matching clauses
      String fieldName = c.getField();
      if (!fieldNames.contains(fieldName)) {
        continue;
      }
      String value = c.getTerm().toString().toLowerCase();

      // add a Lucene TermQuery for this clause
      TermQuery clause = new TermQuery(new Term(fieldName, value));
      // set boost
      if (boosts.containsKey(fieldName)) {
        clause.setBoost(boosts.get(fieldName));
      }

      // add it as specified in query
      output.add(clause, (c.isProhibited() ? BooleanClause.Occur.MUST_NOT
        : (c.isRequired() ? BooleanClause.Occur.MUST
          : BooleanClause.Occur.SHOULD)));
    }

    // return the modified Lucene query
    return output;
View Full Code Here
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
     * @param output
     */
    private void buildQuery(Query input, BooleanQuery output, Configuration conf) {
        Clause[] clauses = input.getClauses();
        for (int i = 0; i < clauses.length; i++) {
          Clause c = clauses[i];

          if (!c.getField().equals(Clause.DEFAULT_FIELD))
            continue;                                 // skip non-default fields
                 
          String[] sterms = null;
          if (c.isPhrase()) {
            //sterms = new CommonGrams(conf).optimizePhrase(c.getPhrase(), FIELDS[f]);         
            sterms = new String[c.getPhrase().getTerms().length];
            for (int j=0; j<sterms.length; j++) {
              sterms[j]=c.getPhrase().getTerms()[j].toString();
            }
          }
          else {
            sterms = new String[1];
            sterms[0]=c.getTerm().toString();
          }

          for (int j=0; j<sterms.length; j++) {   
            BooleanQuery out = new BooleanQuery();
            for (int f=0; f<FIELDS.length; f++) {                                     
              out.add(new TermQuery(new Term(FIELDS[f], sterms[j])), BooleanClause.Occur.SHOULD);
            }
            output.add(out, (c.isProhibited()
                      ? BooleanClause.Occur.MUST_NOT
                      : (c.isRequired()
                          ? BooleanClause.Occur.MUST
                          : BooleanClause.Occur.SHOULD
                        )));
          }         
        }       
View Full Code Here
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
    // Examine each clause in the Nutch query
    Clause [] clauses = input.getClauses();
   
    for (int i = 0; i < clauses.length; i++)
    {
      Clause c = clauses[i];

      // Skip if not date clauses
      if (!c.getField().equals(FIELD_NAME))
      {
        continue;
      }

      String dateTerm = c.getTerm().toString();
      Matcher matcher = pattern.matcher(dateTerm);
     
      if (matcher == null || !matcher.matches())
      {
        String message = "Wrong query syntax " + FIELD_NAME
          + ":" + dateTerm + ". Must be standalone 14 digit " +
          " IA format date or a range with a hyphen between.";
        LOGGER.error(message);
       
        throw new QueryException(message);
      }

      // So, date is in one of 2 possible formats.  First is standalone
      // 14 character IA date.
      String d = matcher.group(1);
     
      if (d != null)
      {
        LOGGER.debug("Found single date: " + d);

        // This is not a range query. Take the passed date and convert
        // it to seconds-since-epoch.
        BooleanQuery bq = new BooleanQuery();
       
        bq.add(new TermQuery(getTerm(getSeconds(pad(d)))),
          BooleanClause.Occur.SHOULD);
       
        output.add(bq,
          (c.isRequired() == true && c.isProhibited() == false)?
          BooleanClause.Occur.MUST:
          (c.isRequired() == false && c.isProhibited() == false)?
          BooleanClause.Occur.SHOULD:
          BooleanClause.Occur.MUST_NOT);

        continue;
      }
View Full Code Here
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
   BooleanClause.Occur occurinner=null;
   int nClauses=0;
   // examine each clause in the Nutch query
   Clause[] clauses = input.getClauses();
   for (int i = 0; i < clauses.length; i++) {
     Clause c = clauses[i];

     // skip non-matching clauses
     if (!c.getField().equals(field))
       continue;

     // get the field value from the clause
     // raw fields are guaranteed to be Terms, not Phrases
     String value = c.getTerm().toString();
     if (lowerCase)
       value = value.toLowerCase();

     // add a Lucene TermQuery for this clause
     TermQuery clause = new TermQuery(new Term(FINAL_KEY, value));   
     // add it as specified in query       
     outinner.add(clause, BooleanClause.Occur.SHOULD);   
     if (nClauses==0) {
       occurinner=c.isProhibited()
              ? BooleanClause.Occur.MUST_NOT
                    : (c.isRequired()
                        ? BooleanClause.Occur.MUST
                        : BooleanClause.Occur.SHOULD
                       );
     }
     nClauses++;
View Full Code Here
98
99
100
101
102
103
104
105
106
107
108
109
110
  /** Run all defined filters. */
  public BooleanQuery filter(Query input) throws QueryException {
    // first check that all field names are claimed by some plugin
    Clause[] clauses = input.getClauses();
    for (int i = 0; i < clauses.length; i++) {
      Clause c = clauses[i];
      if (!isField(c.getField()))
        throw new QueryException("Not a known field name:"+c.getField());
    }

    // then run each plugin
    BooleanQuery output = new BooleanQuery();
    for (int i = 0; i < this.queryFilters.length; i++) {
View Full Code Here
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
    // Examine each clause in the Nutch query
    Clause [] clauses = input.getClauses();
   
    for (int i = 0; i < clauses.length; i++)
    {
      Clause c = clauses[i];

      // Skip if not date clauses
      if (!c.getField().equals(FIELD_NAME))
      {
        continue;
      }

      String dateTerm = c.getTerm().toString();
      Matcher matcher = pattern.matcher(dateTerm);
     
      if (matcher == null || !matcher.matches())
      {
        String message = "Wrong query syntax " + FIELD_NAME
          + ":" + dateTerm + ". Must be standalone 14 digit " +
          " IA format date.";
        LOGGER.error(message);
       
        throw new QueryException(message);
      }

      // So, date is in one format.
      // 14 character IA date.
      String d = matcher.group(1);
     
      if (d != null)
      {
        LOGGER.debug("Found single date: " + d);

        // This is not a range query. Take the passed date and convert
        // it to seconds-since-epoch.     
        output.add(new PwaClosestQuery(getTerm(getSeconds(pad(d)))),
                (c.isProhibited()
                    ? BooleanClause.Occur.MUST_NOT
                    : (c.isRequired()
                        ? BooleanClause.Occur.MUST
                        : BooleanClause.Occur.SHOULD
                       )
                 ));
       
View Full Code Here
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
    throws QueryException {
   
    // examine each clause in the Nutch query
    Clause[] clauses = input.getClauses();
    for (int i = 0; i < clauses.length; i++) {
      Clause c = clauses[i];

      // skip non-matching clauses
      if (!c.getField().equals(field))
        continue;

      // get the field value from the clause
      // raw fields are guaranteed to be Terms, not Phrases
      String value = c.getTerm().toString();
      if (lowerCase)
        value = value.toLowerCase();

      // add a Lucene TermQuery for this clause
      TermQuery clause = new TermQuery(new Term(field, value));
      // set boost
      clause.setBoost(boost);
      // add it as specified in query
     
      output.add(clause,
          (c.isProhibited()
              ? BooleanClause.Occur.MUST_NOT
              : (c.isRequired()
                  ? BooleanClause.Occur.MUST
                  : BooleanClause.Occur.SHOULD
                 )
           ));
    }
View Full Code Here
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
    throws QueryException {
   
    // examine each clause in the Nutch query
    Clause[] clauses = input.getClauses();
    for (int i = 0; i < clauses.length; i++) {
      Clause c = clauses[i];

      // skip non-matching clauses
      if (!c.getField().equals(field))
        continue;

      // optimize phrase clause
      if (c.isPhrase()) {
        String[] opt = this.commonGrams.optimizePhrase(c.getPhrase(), field);
        if (opt.length==1) {
          c = new Clause(new Query.Term(opt[0]),
                         c.isRequired(), c.isProhibited(), getConf());
        } else {
          c = new Clause(new Phrase(opt), c.isRequired(), c.isProhibited(), getConf());
        }
      }

      // construct appropriate Lucene clause
      org.apache.lucene.search.Query luceneClause;
      if (c.isPhrase()) {
        Phrase nutchPhrase = c.getPhrase();
        Query.Term[] terms = nutchPhrase.getTerms();
        PhraseQuery lucenePhrase = new PhraseQuery();
        for (int j = 0; j < terms.length; j++) {
          lucenePhrase.add(new Term(field, terms[j].toString()));
        }
        luceneClause = lucenePhrase;
      } else {
        luceneClause = new TermQuery(new Term(field, c.getTerm().toString()));
      }

      // set boost
      luceneClause.setBoost(boost);
      // add it as specified in query
     
      output.add(luceneClause,
          (c.isProhibited()
              ? BooleanClause.Occur.MUST_NOT
              : (c.isRequired()
                  ? BooleanClause.Occur.MUST
                  : BooleanClause.Occur.SHOULD
                 )
           ));
    }
View Full Code Here
96
97
98
99
100
101
102
103
104
105
106
107
108
  /** Run all defined filters. */
  public BooleanQuery filter(Query input) throws QueryException {
    // first check that all field names are claimed by some plugin
    Clause[] clauses = input.getClauses();
    for (int i = 0; i < clauses.length; i++) {
      Clause c = clauses[i];
      if (!isField(c.getField()))
        throw new QueryException("Not a known field name:"+c.getField());
    }

    // then run each plugin
    BooleanQuery output = new BooleanQuery();
    for (int i = 0; i < this.queryFilters.length; i++) {
View Full Code Here
TOP

Related Classes of org.apache.nutch.searcher.Query.Clause

  • org.apache.nutch.searcher.custom.CustomFieldQueryFilter
  • org.apache.nutch.searcher.FieldQueryFilter
  • org.apache.nutch.searcher.more.DateQueryFilter
  • org.apache.nutch.searcher.QueryFilters
  • org.apache.nutch.searcher.RawFieldQueryFilter
  • org.archive.access.nutch.searcher.WaxDateClosestQueryFilter
  • org.archive.access.nutch.searcher.WaxDateQueryFilter
  • org.archive.access.nutch.searcher.WaxExacturlExpandQueryFilter
  • org.archive.nutch.trec.TRECFeaturesExtractor
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.