ferret 0.9.5 → 0.9.6

data/ext/ind.c

@@ -283,6 +283,7 @@ TopDocs *index_search_str(Index *self, char *qstr, int first_doc,
   Query *query;
   TopDocs *td;
   query = index_get_query(self, qstr); /* will ensure_searcher is open */
+  //printf(">> %s <<\n", query->to_s(query, ""));
   td = sea_search(self->sea, query, first_doc, num_docs, filter, sort);
   q_deref(query);
   return td;

data/ext/q_fuzzy.c

@@ -65,7 +65,6 @@ float fuzq_score(FuzzyQuery *fuzq, char *target)
   if (((m+1) * (n+1)) >= fuzq->da_capa) {
     fuzq->da_capa = ((m+1) * (n+1)) * 2;
     REALLOC_N(fuzq->da, int, fuzq->da_capa);
-    printf("making capa %d -> %d\n", fuzq->da_capa, (int)fuzq->da);
     d = fuzq->da;
   }
 
@@ -253,9 +252,7 @@ void fuzq_destroy(Query *self)
 {
   FuzzyQuery *fuzq = (FuzzyQuery *)self->data;
   if (self->destroy_all) term_destroy((Term *)fuzq->term);
-printf("freeing %d -> %d\n", fuzq->da_capa, (int)fuzq->da);
   free(fuzq->da);
-printf("success\n");
   free(fuzq);
   q_destroy_i(self);
 }

data/ext/q_parser.c

@@ -179,12 +179,16 @@ Query *get_range_q(char *field, char *from, char *to,
       sq = func;\
       if (sq) bq_add_query(q, sq, BC_SHOULD);\
     }\
+    if (((BooleanQuery *)q->data)->clause_cnt == 0) {\
+      q_deref(q);\
+      q = NULL;\
+    }\
   }\
 } while (0)
 
 
 /* Line 219 of yacc.c.  */
-#line 188 "y.tab.c"
+#line 192 "y.tab.c"
 
 #if ! defined (YYSIZE_T) && defined (__SIZE_TYPE__)
 # define YYSIZE_T __SIZE_TYPE__
@@ -423,12 +427,12 @@ static const yysigned_char yyrhs[] =
 /* YYRLINE[YYN] -- source line where rule number YYN was defined.  */
 static const unsigned char yyrline[] =
 {
-       0,    86,    86,    87,    89,    90,    91,    92,    94,    95,
-      96,    98,    99,   101,   102,   103,   104,   105,   106,   108,
-     109,   110,   112,   114,   114,   116,   116,   116,   119,   120,
-     122,   123,   124,   125,   127,   128,   129,   130,   131,   133,
-     134,   135,   136,   137,   138,   139,   140,   141,   142,   143,
-     144
+       0,    90,    90,    91,    93,    94,    95,    96,    98,    99,
+     100,   102,   103,   105,   106,   107,   108,   109,   110,   112,
+     113,   114,   116,   118,   118,   120,   120,   120,   123,   124,
+     126,   127,   128,   129,   131,   132,   133,   134,   135,   137,
+     138,   139,   140,   141,   142,   143,   144,   145,   146,   147,
+     148
 };
 #endif
 
@@ -1236,217 +1240,217 @@ yyreduce:
   switch (yyn)
     {
         case 2:
-#line 86 "src/query_parser/q_parser.y"
+#line 90 "src/query_parser/q_parser.y"
     { qp->result = (yyval.query) = NULL; }
     break;
 
   case 3:
-#line 87 "src/query_parser/q_parser.y"
+#line 91 "src/query_parser/q_parser.y"
     { qp->result = (yyval.query) = get_bool_q((yyvsp[0].array)); }
     break;
 
   case 4:
-#line 89 "src/query_parser/q_parser.y"
+#line 93 "src/query_parser/q_parser.y"
     { (yyval.array) = first_cls((yyvsp[0].bcls)); }
     break;
 
   case 5:
-#line 90 "src/query_parser/q_parser.y"
+#line 94 "src/query_parser/q_parser.y"
     { (yyval.array) = add_and_cls((yyvsp[-2].array), (yyvsp[0].bcls)); }
     break;
 
   case 6:
-#line 91 "src/query_parser/q_parser.y"
+#line 95 "src/query_parser/q_parser.y"
     { (yyval.array) = add_or_cls((yyvsp[-2].array), (yyvsp[0].bcls)); }
     break;
 
   case 7:
-#line 92 "src/query_parser/q_parser.y"
+#line 96 "src/query_parser/q_parser.y"
     { (yyval.array) = add_default_cls(qp, (yyvsp[-1].array), (yyvsp[0].bcls)); }
     break;
 
   case 8:
-#line 94 "src/query_parser/q_parser.y"
+#line 98 "src/query_parser/q_parser.y"
     { (yyval.bcls) = get_bool_cls((yyvsp[0].query), BC_MUST); }
     break;
 
   case 9:
-#line 95 "src/query_parser/q_parser.y"
+#line 99 "src/query_parser/q_parser.y"
     { (yyval.bcls) = get_bool_cls((yyvsp[0].query), BC_MUST_NOT); }
     break;
 
   case 10:
-#line 96 "src/query_parser/q_parser.y"
+#line 100 "src/query_parser/q_parser.y"
     { (yyval.bcls) = get_bool_cls((yyvsp[0].query), BC_SHOULD); }
     break;
 
   case 12:
-#line 99 "src/query_parser/q_parser.y"
+#line 103 "src/query_parser/q_parser.y"
     { if ((yyvsp[-2].query)) sscanf((yyvsp[0].str),"%f",&((yyvsp[-2].query)->boost)); (yyval.query)=(yyvsp[-2].query); }
     break;
 
   case 14:
-#line 102 "src/query_parser/q_parser.y"
+#line 106 "src/query_parser/q_parser.y"
     { (yyval.query) = get_bool_q((yyvsp[-1].array)); }
     break;
 
   case 19:
-#line 108 "src/query_parser/q_parser.y"
+#line 112 "src/query_parser/q_parser.y"
     { FLDS((yyval.query), get_term_q(qp, field, (yyvsp[0].str))); }
     break;
 
   case 20:
-#line 109 "src/query_parser/q_parser.y"
+#line 113 "src/query_parser/q_parser.y"
     { FLDS((yyval.query), get_fuzzy_q(qp, field, (yyvsp[-2].str), (yyvsp[0].str))); }
     break;
 
   case 21:
-#line 110 "src/query_parser/q_parser.y"
+#line 114 "src/query_parser/q_parser.y"
     { FLDS((yyval.query), get_fuzzy_q(qp, field, (yyvsp[-1].str), NULL)); }
     break;
 
   case 22:
-#line 112 "src/query_parser/q_parser.y"
+#line 116 "src/query_parser/q_parser.y"
     { FLDS((yyval.query), get_wild_q(qp, field, (yyvsp[0].str))); }
     break;
 
   case 23:
-#line 114 "src/query_parser/q_parser.y"
+#line 118 "src/query_parser/q_parser.y"
     { qp->fields = qp->def_fields; }
     break;
 
   case 24:
-#line 115 "src/query_parser/q_parser.y"
+#line 119 "src/query_parser/q_parser.y"
     { (yyval.query) = (yyvsp[-1].query); }
     break;
 
   case 25:
-#line 116 "src/query_parser/q_parser.y"
+#line 120 "src/query_parser/q_parser.y"
     { qp->fields = qp->all_fields; }
     break;
 
   case 26:
-#line 116 "src/query_parser/q_parser.y"
+#line 120 "src/query_parser/q_parser.y"
     {qp->fields = qp->def_fields;}
     break;
 
   case 27:
-#line 117 "src/query_parser/q_parser.y"
+#line 121 "src/query_parser/q_parser.y"
     { (yyval.query) = (yyvsp[-1].query); }
     break;
 
   case 28:
-#line 119 "src/query_parser/q_parser.y"
+#line 123 "src/query_parser/q_parser.y"
     { (yyval.hashset) = first_field(qp, (yyvsp[0].str)); }
     break;
 
   case 29:
-#line 120 "src/query_parser/q_parser.y"
+#line 124 "src/query_parser/q_parser.y"
     { (yyval.hashset) = add_field(qp, (yyvsp[0].str));}
     break;
 
   case 30:
-#line 122 "src/query_parser/q_parser.y"
+#line 126 "src/query_parser/q_parser.y"
     { (yyval.query) = get_phrase_q(qp, (yyvsp[-1].phrase), NULL); }
     break;
 
   case 31:
-#line 123 "src/query_parser/q_parser.y"
+#line 127 "src/query_parser/q_parser.y"
     { (yyval.query) = get_phrase_q(qp, (yyvsp[-3].phrase), (yyvsp[0].str)); }
     break;
 
   case 32:
-#line 124 "src/query_parser/q_parser.y"
+#line 128 "src/query_parser/q_parser.y"
     { (yyval.query) = NULL; }
     break;
 
   case 33:
-#line 125 "src/query_parser/q_parser.y"
+#line 129 "src/query_parser/q_parser.y"
     { (yyval.query) = NULL; }
     break;
 
   case 34:
-#line 127 "src/query_parser/q_parser.y"
+#line 131 "src/query_parser/q_parser.y"
     { (yyval.phrase) = ph_first_word((yyvsp[0].str)); }
     break;
 
   case 35:
-#line 128 "src/query_parser/q_parser.y"
+#line 132 "src/query_parser/q_parser.y"
     { (yyval.phrase) = ph_first_word(NULL); }
     break;
 
   case 36:
-#line 129 "src/query_parser/q_parser.y"
+#line 133 "src/query_parser/q_parser.y"
     { (yyval.phrase) = ph_add_word((yyvsp[-1].phrase), (yyvsp[0].str)); }
     break;
 
   case 37:
-#line 130 "src/query_parser/q_parser.y"
+#line 134 "src/query_parser/q_parser.y"
     { (yyval.phrase) = ph_add_word((yyvsp[-2].phrase), NULL); }
     break;
 
   case 38:
-#line 131 "src/query_parser/q_parser.y"
+#line 135 "src/query_parser/q_parser.y"
     { (yyval.phrase) = ph_add_multi_word((yyvsp[-2].phrase), (yyvsp[0].str));  }
     break;
 
   case 39:
-#line 133 "src/query_parser/q_parser.y"
+#line 137 "src/query_parser/q_parser.y"
     { FLDS((yyval.query), get_range_q(field, (yyvsp[-2].str),  (yyvsp[-1].str),  true,  true)); }
     break;
 
   case 40:
-#line 134 "src/query_parser/q_parser.y"
+#line 138 "src/query_parser/q_parser.y"
     { FLDS((yyval.query), get_range_q(field, (yyvsp[-2].str),  (yyvsp[-1].str),  true,  false)); }
     break;
 
   case 41:
-#line 135 "src/query_parser/q_parser.y"
+#line 139 "src/query_parser/q_parser.y"
     { FLDS((yyval.query), get_range_q(field, (yyvsp[-2].str),  (yyvsp[-1].str),  false, true)); }
     break;
 
   case 42:
-#line 136 "src/query_parser/q_parser.y"
+#line 140 "src/query_parser/q_parser.y"
     { FLDS((yyval.query), get_range_q(field, (yyvsp[-2].str),  (yyvsp[-1].str),  false, false)); }
     break;
 
   case 43:
-#line 137 "src/query_parser/q_parser.y"
+#line 141 "src/query_parser/q_parser.y"
     { FLDS((yyval.query), get_range_q(field, NULL,(yyvsp[-1].str),  false, false)); }
     break;
 
   case 44:
-#line 138 "src/query_parser/q_parser.y"
+#line 142 "src/query_parser/q_parser.y"
     { FLDS((yyval.query), get_range_q(field, NULL,(yyvsp[-1].str),  false, true)); }
     break;
 
   case 45:
-#line 139 "src/query_parser/q_parser.y"
+#line 143 "src/query_parser/q_parser.y"
     { FLDS((yyval.query), get_range_q(field, (yyvsp[-1].str),  NULL,true,  false)); }
     break;
 
   case 46:
-#line 140 "src/query_parser/q_parser.y"
+#line 144 "src/query_parser/q_parser.y"
     { FLDS((yyval.query), get_range_q(field, (yyvsp[-1].str),  NULL,false, false)); }
     break;
 
   case 47:
-#line 141 "src/query_parser/q_parser.y"
+#line 145 "src/query_parser/q_parser.y"
     { FLDS((yyval.query), get_range_q(field, NULL,(yyvsp[0].str),  false, false)); }
     break;
 
   case 48:
-#line 142 "src/query_parser/q_parser.y"
+#line 146 "src/query_parser/q_parser.y"
     { FLDS((yyval.query), get_range_q(field, NULL,(yyvsp[0].str),  false, true)); }
     break;
 
   case 49:
-#line 143 "src/query_parser/q_parser.y"
+#line 147 "src/query_parser/q_parser.y"
     { FLDS((yyval.query), get_range_q(field, (yyvsp[0].str),  NULL,true,  false)); }
     break;
 
   case 50:
-#line 144 "src/query_parser/q_parser.y"
+#line 148 "src/query_parser/q_parser.y"
     { FLDS((yyval.query), get_range_q(field, (yyvsp[0].str),  NULL,false, false)); }
     break;
 
@@ -1455,7 +1459,7 @@ yyreduce:
     }
 
 /* Line 1126 of yacc.c.  */
-#line 1459 "y.tab.c"
+#line 1463 "y.tab.c"
 
   yyvsp -= yylen;
   yyssp -= yylen;
@@ -1723,7 +1727,7 @@ yyreturn:
 }
 
 
-#line 146 "src/query_parser/q_parser.y"
+#line 150 "src/query_parser/q_parser.y"
 
 
 const char *special_char = "&:()[]{}!+\"~^-|<>=*?";

data/lib/ferret.rb

@@ -22,7 +22,7 @@
 #++
 # :include: ../TUTORIAL
 module Ferret
-  VERSION = '0.9.5'
+  VERSION = '0.9.6'
 end
 
 # try and load the C extension but it isn't necessary.

data/lib/ferret/query_parser/query_parser.tab.rb

@@ -11,7 +11,7 @@ module Ferret
 
   class QueryParser < Racc::Parser
 
-module_eval <<'..end lib/ferret/query_parser/query_parser.y modeval..id155b60f3fb', 'lib/ferret/query_parser/query_parser.y', 126
+module_eval <<'..end lib/ferret/query_parser/query_parser.y modeval..id07e7308361', 'lib/ferret/query_parser/query_parser.y', 126
   attr_accessor :default_field, :fields, :handle_parse_errors
 
   def initialize(default_field = "*", options = {})
@@ -323,7 +323,7 @@ module_eval <<'..end lib/ferret/query_parser/query_parser.y modeval..id155b60f3f
   end
 
   def add_and_clause(clauses, clause)
-    clauses.compact!
+    (clauses||=[]).compact!
     if (clauses.length == 1)
       last_cl = clauses[0]
       last_cl.occur = BooleanClause::Occur::MUST if not last_cl.prohibited?
@@ -379,7 +379,7 @@ module_eval <<'..end lib/ferret/query_parser/query_parser.y modeval..id155b60f3f
         q = yield(field)
         bq << BooleanClause.new(q) if q
       end
-      return bq                
+      return (bq.clauses.size == 0) ? nil : bq
     end
   end
 
@@ -398,7 +398,7 @@ module_eval <<'..end lib/ferret/query_parser/query_parser.y modeval..id155b60f3f
     return qp.parse(query)
   end
 
-..end lib/ferret/query_parser/query_parser.y modeval..id155b60f3fb
+..end lib/ferret/query_parser/query_parser.y modeval..id07e7308361
 
 ##### racc 1.4.4 generates ###
 

data/lib/rferret.rb

@@ -23,7 +23,7 @@ $: << File.dirname(__FILE__)
 #++
 # :include: ../TUTORIAL
 module Ferret
-  VERSION = '0.9.5'
+  VERSION = '0.9.6'
 end
 
 $ferret_pure_ruby = true

data/test/unit/index/tc_index.rb

@@ -624,4 +624,30 @@ class IndexTest < Test::Unit::TestCase
     index.flush
     index.close
   end
+
+  def test_stopwords
+    i = Ferret::Index::Index.new(
+            :occur_default => Ferret::Search::BooleanClause::Occur::MUST,
+            :default_search_field => '*')
+    d = Ferret::Document::Document.new
+ 
+    # adding this additional field to the document leads to failure below
+    # comment out this statement and all tests pass:
+    d << Ferret::Document::Field.new('id', '1',
+                                     Ferret::Document::Field::Store::YES,
+                                     Ferret::Document::Field::Index::UNTOKENIZED)
+ 
+    d << Ferret::Document::Field.new('content', 'Move or shake',
+                                     Ferret::Document::Field::Store::NO,
+                                     Ferret::Document::Field::Index::TOKENIZED,
+                                     Ferret::Document::Field::TermVector::NO,
+                                     false, 1.0)
+    i << d
+    hits = i.search 'move nothere shake'
+    assert_equal 0, hits.size
+    hits = i.search 'move shake'
+    assert_equal 1, hits.size
+    hits = i.search 'move or shake'
+    assert_equal 1, hits.size # fails when id field is present
+  end
 end

metadata

@@ -3,8 +3,8 @@ rubygems_version: 0.8.11
 specification_version: 1
 name: ferret
 version: !ruby/object:Gem::Version 
-  version: 0.9.5
-date: 2006-08-07 00:00:00 +09:00
+  version: 0.9.6
+date: 2006-08-23 00:00:00 +09:00
 summary: Ruby indexing library.
 require_paths: 
 - lib