rroonga 4.0.5 → 4.0.6

data/ext/groonga/rb-grn-column.c

@@ -758,7 +758,7 @@ rb_grn_column_get_indexes (int argc, VALUE *argv, VALUE self)
 }
 
 /*
- * Renames the table to name.
+ * Renames the column to name.
  * @since 1.3.0
  * @overload rename(name)
  *   @param name [String] the new name
@@ -782,7 +782,8 @@ rb_grn_column_rename (VALUE self, VALUE rb_name)
     rc = grn_column_rename(context, column, name, name_size);
     rb_grn_context_check(context, self);
     rb_grn_rc_check(rc, self);
-
+    rb_grn_named_object_set_name(RB_GRN_NAMED_OBJECT(DATA_PTR(self)),
+                                 name, name_size);
     return self;
 }
 

data/ext/groonga/rb-grn-context.c

@@ -670,13 +670,28 @@ rb_grn_context_support_zlib_p (VALUE self)
 }
 
 /*
- * groongaがLZOサポート付きでビルドされていれば +true+ 、そう
- * でなければ +false+ を返す。
+ * If Groonga supports LZO compression, it returns +true+,
+ * otherwise it returns +false+.
+ *
+ * Groonga does not support LZO compression 4.0.7 or later.
+ * This method always returns +false+.
  *
  * @overload support_lzo?
  */
 static VALUE
 rb_grn_context_support_lzo_p (VALUE self)
+{
+    return Qfalse;
+}
+
+/*
+ * If Groonga supports LZ4 compression, it returns +true+,
+ * otherwise it returns +false+.
+ *
+ * @overload support_lz4?
+ */
+static VALUE
+rb_grn_context_support_lz4_p (VALUE self)
 {
     VALUE rb_support_p;
     grn_ctx *context;
@@ -684,7 +699,7 @@ rb_grn_context_support_lzo_p (VALUE self)
 
     context = SELF(self);
     GRN_BOOL_INIT(&support_p, 0);
-    grn_obj_get_info(context, NULL, GRN_INFO_SUPPORT_LZO, &support_p);
+    grn_obj_get_info(context, NULL, GRN_INFO_SUPPORT_LZ4, &support_p);
     rb_support_p = CBOOL2RVAL(GRN_BOOL_VALUE(&support_p));
     GRN_OBJ_FIN(context, &support_p);
 
@@ -974,6 +989,8 @@ rb_grn_init_context (VALUE mGrn)
                      rb_grn_context_support_zlib_p, 0);
     rb_define_method(cGrnContext, "support_lzo?",
                      rb_grn_context_support_lzo_p, 0);
+    rb_define_method(cGrnContext, "support_lz4?",
+                     rb_grn_context_support_lz4_p, 0);
 
     rb_define_method(cGrnContext, "database", rb_grn_context_get_database, 0);
 

data/ext/groonga/rb-grn-double-array-trie.c

@@ -219,6 +219,10 @@ rb_grn_double_array_trie_bind (RbGrnDoubleArrayTrie *rb_grn_double_array_trie,
  *       {Groonga::IndexColumn} を定義する場合は
  *       @"TokenBigram"@ などを指定する必要がある。
  *
+ *     @option options [::Array<String, Groonga::Procedure>, nil]
+ *       :token_filters (nil) The token filters to be used in the
+ *       table.
+ *
  *     @option options :sub_records
  *       +true+ を指定すると {#group} でグループ化したときに、
  *       {Groonga::Record#n_sub_records} でグループに含まれるレコー
@@ -246,7 +250,9 @@ rb_grn_double_array_trie_s_create (int argc, VALUE *argv, VALUE klass)
     VALUE options, rb_context, rb_name, rb_path, rb_persistent;
     VALUE rb_key_normalize, rb_key_with_sis, rb_key_type;
     VALUE rb_value_type;
-    VALUE rb_default_tokenizer, rb_sub_records;
+    VALUE rb_default_tokenizer;
+    VALUE rb_token_filters;
+    VALUE rb_sub_records;
     VALUE rb_normalizer;
 
     rb_scan_args(argc, argv, "01", &options);
@@ -261,6 +267,7 @@ rb_grn_double_array_trie_s_create (int argc, VALUE *argv, VALUE klass)
                         "key_type", &rb_key_type,
                         "value_type", &rb_value_type,
                         "default_tokenizer", &rb_default_tokenizer,
+                        "token_filters", &rb_token_filters,
                         "sub_records", &rb_sub_records,
                         "normalizer", &rb_normalizer,
                         NULL);
@@ -308,7 +315,9 @@ rb_grn_double_array_trie_s_create (int argc, VALUE *argv, VALUE klass)
     if (!NIL_P(rb_default_tokenizer))
         rb_funcall(rb_table, rb_intern("default_tokenizer="), 1,
                    rb_default_tokenizer);
-
+    if (!NIL_P(rb_token_filters))
+        rb_funcall(rb_table, rb_intern("token_filters="), 1,
+                   rb_token_filters);
     if (!NIL_P(rb_normalizer))
         rb_funcall(rb_table, rb_intern("normalizer="), 1,
                    rb_normalizer);

data/ext/groonga/rb-grn-exception.c

@@ -1,6 +1,7 @@
 /* -*- coding: utf-8; mode: C; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
 /*
-  Copyright (C) 2009-2012  Kouhei Sutou <kou@clear-code.com>
+  Copyright (C) 2009-2014  Kouhei Sutou <kou@clear-code.com>
+  Copyright (C) 2014  Masafumi Yokoyama <myokoym@gmail.com>
 
   This library is free software; you can redistribute it and/or
   modify it under the terms of the GNU Lesser General Public
@@ -83,7 +84,7 @@ static VALUE eGrnNotSocket;
 static VALUE eGrnOperationNotSupported;
 static VALUE eGrnAddressIsInUse;
 static VALUE eGrnZLibError;
-static VALUE eGrnLZOError;
+static VALUE eGrnLZ4Error;
 static VALUE eGrnStackOverFlow;
 static VALUE eGrnSyntaxError;
 static VALUE eGrnRetryMax;
@@ -95,6 +96,7 @@ static VALUE eGrnTooSmallLimit;
 static VALUE eGrnCASError;
 static VALUE eGrnUnsupportedCommandVersion;
 static VALUE eGrnNormalizerError;
+static VALUE eGrnTokenFilterError;
 
 VALUE
 rb_grn_rc_to_exception (grn_rc rc)
@@ -288,8 +290,8 @@ rb_grn_rc_to_exception (grn_rc rc)
       case GRN_ZLIB_ERROR:
         exception = eGrnZLibError;
         break;
-      case GRN_LZO_ERROR:
-        exception = eGrnLZOError;
+      case GRN_LZ4_ERROR:
+        exception = eGrnLZ4Error;
         break;
       case GRN_STACK_OVER_FLOW:
         exception = eGrnStackOverFlow;
@@ -324,6 +326,9 @@ rb_grn_rc_to_exception (grn_rc rc)
       case GRN_NORMALIZER_ERROR:
         exception = eGrnNormalizerError;
         break;
+      case GRN_TOKEN_FILTER_ERROR:
+        exception = eGrnTokenFilterError;
+        break;
     }
 
     if (NIL_P(exception))
@@ -524,8 +529,8 @@ rb_grn_rc_to_message (grn_rc rc)
       case GRN_ZLIB_ERROR:
         message = "ZLib error";
         break;
-      case GRN_LZO_ERROR:
-        message = "LZO error";
+      case GRN_LZ4_ERROR:
+        message = "LZ4 error";
         break;
       case GRN_STACK_OVER_FLOW:
         message = "stack over flow";
@@ -560,6 +565,9 @@ rb_grn_rc_to_message (grn_rc rc)
       case GRN_NORMALIZER_ERROR:
         message = "normalizer error";
         break;
+      case GRN_TOKEN_FILTER_ERROR:
+        message = "token filter error";
+        break;
     }
 
     if (!message)
@@ -1114,12 +1122,12 @@ rb_grn_init_exception (VALUE mGrn)
         rb_define_class_under(mGrn, "ZLibError", rb_eGrnError);
 
     /*
-     * Document-class: Groonga::LZOError
+     * Document-class: Groonga::LZ4Error
      *
-     * LZOに問題があるときに発生する。
+     * LZ4に問題があるときに発生する。
      */
-    eGrnLZOError =
-        rb_define_class_under(mGrn, "LZOError", rb_eGrnError);
+    eGrnLZ4Error =
+        rb_define_class_under(mGrn, "LZ4Error", rb_eGrnError);
 
     /*
      * Document-class: Groonga::StackOverFlow
@@ -1209,4 +1217,14 @@ rb_grn_init_exception (VALUE mGrn)
      */
     eGrnNormalizerError =
         rb_define_class_under(mGrn, "NormalizerError", rb_eGrnError);
+
+    /*
+     * Document-class: Groonga::TokenFilterError
+     *
+     * It is used when a token filter causes an error.
+     *
+     * @since 4.0.6
+     */
+    eGrnTokenFilterError =
+        rb_define_class_under(mGrn, "TokenFilterError", rb_eGrnError);
 }

data/ext/groonga/rb-grn-expression.c

@@ -1,6 +1,7 @@
 /* -*- coding: utf-8; mode: C; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
 /*
   Copyright (C) 2009-2014  Kouhei Sutou <kou@clear-code.com>
+  Copyright (C) 2014  Masafumi Yokoyama <myokoym@gmail.com>
 
   This library is free software; you can redistribute it and/or
   modify it under the terms of the GNU Lesser General Public
@@ -493,6 +494,37 @@ rb_grn_expression_compile (VALUE self)
     return Qnil;
 }
 
+/*
+ * Dump execution plan of the `expression` in string.
+ *
+ * @overload dump_plan
+ * @since 4.0.8
+ */
+static VALUE
+rb_grn_expression_dump_plan (VALUE self)
+{
+/* TODO: Remove me after Groogna 4.0.8 is released. */
+#ifdef HAVE_GRN_EXPR_DUMP_PLAN
+    grn_ctx *context = NULL;
+    grn_obj *expression;
+    grn_obj dumped_plan;
+    VALUE rb_dumped_plan;
+
+    rb_grn_expression_deconstruct(SELF(self), &expression, &context,
+                                  NULL, NULL,
+                                  NULL, NULL, NULL);
+
+    GRN_TEXT_INIT(&dumped_plan, 0);
+    grn_expr_dump_plan(context, expression, &dumped_plan);
+    rb_dumped_plan = GRNBULK2RVAL(context, &dumped_plan, NULL, self);
+    grn_obj_unlink(context, &dumped_plan);
+
+    return rb_dumped_plan;
+#else
+    return rb_str_new2("");
+#endif
+}
+
 /*
  * _expression_ で使用可能な変数のうち、名前が _name_
  * または _offset_ 番目に {Expression#append_object}
@@ -717,6 +749,47 @@ rb_grn_expression_snippet (int argc, VALUE *argv, VALUE self)
     return GRNOBJECT2RVAL(Qnil, context, snippet, GRN_TRUE);
 }
 
+/*
+ * Extracts keywords from _expression_. The keywords order isn't
+ * guaranteed.
+ *
+ * @example
+ *   expression.parse("Ruby OR Groonga")
+ *   expression.keywords  #=> ["Groonga", "Ruby"]
+ *
+ * @overload keywords
+ *   @return [::Array<String>] the extracted keywords
+ *
+ * @since 4.0.6
+ */
+static VALUE
+rb_grn_expression_get_keywords (VALUE self)
+{
+    grn_ctx *context = NULL;
+    grn_obj *expression;
+    grn_obj keywords;
+    VALUE rb_keywords = rb_ary_new();
+
+    rb_grn_expression_deconstruct(SELF(self), &expression, &context,
+                                  NULL, NULL,
+                                  NULL, NULL, NULL);
+
+    GRN_PTR_INIT(&keywords, GRN_OBJ_VECTOR, GRN_ID_NIL);
+    grn_expr_get_keywords(context, expression, &keywords);
+    {
+        int i, n_keywords;
+        n_keywords = GRN_BULK_VSIZE(&keywords) / sizeof(grn_obj *);
+        for (i = 0; i < n_keywords; i++) {
+            grn_obj *keyword = GRN_PTR_VALUE_AT(&keywords, i);
+            rb_ary_push(rb_keywords,
+                        GRNBULK2RVAL(context, keyword, NULL, self));
+        }
+    }
+    GRN_OBJ_FIN(context, &keywords);
+
+    return rb_keywords;
+}
+
 void
 rb_grn_init_expression (VALUE mGrn)
 {
@@ -741,6 +814,8 @@ rb_grn_init_expression (VALUE mGrn)
                      rb_grn_expression_execute, 0);
     rb_define_method(rb_cGrnExpression, "compile",
                      rb_grn_expression_compile, 0);
+    rb_define_method(rb_cGrnExpression, "dump_plan",
+                     rb_grn_expression_dump_plan, 0);
 
     rb_define_method(rb_cGrnExpression, "[]",
                      rb_grn_expression_array_reference, 1);
@@ -748,6 +823,9 @@ rb_grn_init_expression (VALUE mGrn)
     rb_define_method(rb_cGrnExpression, "snippet",
                      rb_grn_expression_snippet, -1);
 
+    rb_define_method(rb_cGrnExpression, "keywords",
+                     rb_grn_expression_get_keywords, 0);
+
     rb_define_method(rb_cGrnExpression, "inspect",
                      rb_grn_expression_inspect, 0);
 }

data/ext/groonga/rb-grn-hash.c

@@ -131,6 +131,11 @@ VALUE rb_cGrnHash;
  *       デフォルトでは何も設定されていないので、テーブルに
  *       {Groonga::IndexColumn} を定義する場合は
  *       @"TokenBigram"@ などを指定する必要がある。
+ *
+ *     @option options [::Array<String, Groonga::Procedure>, nil]
+ *       :token_filters (nil) The token filters to be used in the
+ *       table.
+ *
  *     @option options [Groonga::Record#n_sub_records] :sub_records
  *       +true+ を指定すると {#group} でグループ化したときに、
  *       {Groonga::Record#n_sub_records} でグループに含まれるレコー
@@ -157,6 +162,7 @@ rb_grn_hash_s_create (int argc, VALUE *argv, VALUE klass)
     VALUE rb_table;
     VALUE options, rb_context, rb_name, rb_path, rb_persistent;
     VALUE rb_key_normalize, rb_key_type, rb_value_type, rb_default_tokenizer;
+    VALUE rb_token_filters;
     VALUE rb_sub_records;
     VALUE rb_normalizer;
 
@@ -171,6 +177,7 @@ rb_grn_hash_s_create (int argc, VALUE *argv, VALUE klass)
                         "key_type", &rb_key_type,
                         "value_type", &rb_value_type,
                         "default_tokenizer", &rb_default_tokenizer,
+                        "token_filters", &rb_token_filters,
                         "sub_records", &rb_sub_records,
                         "normalizer", &rb_normalizer,
                         NULL);
@@ -215,6 +222,9 @@ rb_grn_hash_s_create (int argc, VALUE *argv, VALUE klass)
     if (!NIL_P(rb_default_tokenizer))
         rb_funcall(rb_table, rb_intern("default_tokenizer="), 1,
                    rb_default_tokenizer);
+    if (!NIL_P(rb_token_filters))
+        rb_funcall(rb_table, rb_intern("token_filters="), 1,
+                   rb_token_filters);
     if (!NIL_P(rb_normalizer))
         rb_funcall(rb_table, rb_intern("normalizer="), 1,
                    rb_normalizer);

data/ext/groonga/rb-grn-logger.c

@@ -193,13 +193,14 @@ rb_grn_logger_fin (grn_ctx *ctx, void *user_data)
  * groongaがログを出力する度に呼び出されるブロックを登録する。
  *
  * @overload register(options={})
- *   @yield [level, time, title, message, location]
- *     _level_ はSymbol、それ以外は全て文字列で渡される。 _level_ 以外
+ *   @yield [event, level, time, title, message, location]
+ *     _event_ と _level_ はSymbol、それ以外は全て文字列で渡される。
+ *     _event_ と _level_ 以外
  *     の4つについては _options_ で +false+ を指定することでブロックに
  *     渡さないようにすることができ、その場合は空文字列が実際には渡される。
  *   @param options [::Hash] The name and value
  *     pairs. Omitted names are initialized as the default value.
- *   @option options :level (:notice)
+ *   @option options :max_level (:notice)
  *     ログのレベルを +:none+ ,  +:emergency+ ,  +:alert+ ,
  *     +:critical+ , +:error+ , +:warning+ , +:notice+ , +:info+ ,
  *     +:debug+ ,  +:dump+ のいずれかで指定する。それより重要度が
@@ -234,6 +235,9 @@ rb_grn_logger_s_register (int argc, VALUE *argv, VALUE klass)
     rb_scan_args(argc, argv, "02&", &rb_logger, &rb_options, &rb_callback);
 
     if (rb_block_given_p()) {
+        if (!NIL_P(rb_logger)) {
+            rb_options = rb_logger;
+        }
         rb_logger = rb_funcall(cGrnCallbackLogger, id_new, 1, rb_callback);
     }
 

data/ext/groonga/rb-grn-object.c

@@ -1,6 +1,7 @@
 /* -*- coding: utf-8; mode: C; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
 /*
   Copyright (C) 2009-2014  Kouhei Sutou <kou@clear-code.com>
+  Copyright (C) 2014  Masafumi Yokoyama <myokoym@gmail.com>
 
   This library is free software; you can redistribute it and/or
   modify it under the terms of the GNU Lesser General Public
@@ -901,8 +902,8 @@ rb_grn_object_inspect_content_flags_with_label (VALUE inspected,
       case GRN_COLUMN_VAR_SIZE:
         if (flags & GRN_OBJ_COMPRESS_ZLIB)
             rb_ary_push(inspected_flags, rb_str_new2("COMPRESS_ZLIB"));
-        if (flags & GRN_OBJ_COMPRESS_LZO)
-            rb_ary_push(inspected_flags, rb_str_new2("COMPRESS_LZO"));
+        if (flags & GRN_OBJ_COMPRESS_LZ4)
+            rb_ary_push(inspected_flags, rb_str_new2("COMPRESS_LZ4"));
         break;
       case GRN_COLUMN_INDEX:
         if (flags & GRN_OBJ_WITH_SECTION)

data/ext/groonga/rb-grn-patricia-trie.c

@@ -1,7 +1,7 @@
 /* -*- coding: utf-8; mode: C; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
 /* vim: set sts=4 sw=4 ts=8 noet: */
 /*
-  Copyright (C) 2009  Kouhei Sutou <kou@clear-code.com>
+  Copyright (C) 2009-2014  Kouhei Sutou <kou@clear-code.com>
 
   This library is free software; you can redistribute it and/or
   modify it under the terms of the GNU Lesser General Public
@@ -141,6 +141,11 @@ VALUE rb_cGrnPatriciaTrie;
  *       デフォルトでは何も設定されていないので、テーブルに
  *       {Groonga::IndexColumn} を定義する場合は
  *       @"TokenBigram"@ などを指定する必要がある。
+ *
+ *     @option options [::Array<String, Groonga::Procedure>, nil]
+ *       :token_filters (nil) The token filters to be used in the
+ *       table.
+ *
  *     @option options :sub_records
  *       +true+ を指定すると {#group} でグループ化したときに、
  *       {Groonga::Record#n_sub_records} でグループに含まれるレコー
@@ -168,7 +173,9 @@ rb_grn_patricia_trie_s_create (int argc, VALUE *argv, VALUE klass)
     VALUE options, rb_context, rb_name, rb_path, rb_persistent;
     VALUE rb_key_normalize, rb_key_with_sis, rb_key_type;
     VALUE rb_value_type;
-    VALUE rb_default_tokenizer, rb_sub_records;
+    VALUE rb_default_tokenizer;
+    VALUE rb_token_filters;
+    VALUE rb_sub_records;
     VALUE rb_normalizer;
 
     rb_scan_args(argc, argv, "01", &options);
@@ -183,6 +190,7 @@ rb_grn_patricia_trie_s_create (int argc, VALUE *argv, VALUE klass)
                         "key_type", &rb_key_type,
                         "value_type", &rb_value_type,
                         "default_tokenizer", &rb_default_tokenizer,
+                        "token_filters", &rb_token_filters,
                         "sub_records", &rb_sub_records,
                         "normalizer", &rb_normalizer,
                         NULL);
@@ -230,6 +238,9 @@ rb_grn_patricia_trie_s_create (int argc, VALUE *argv, VALUE klass)
     if (!NIL_P(rb_default_tokenizer))
         rb_funcall(rb_table, rb_intern("default_tokenizer="), 1,
                    rb_default_tokenizer);
+    if (!NIL_P(rb_token_filters))
+        rb_funcall(rb_table, rb_intern("token_filters="), 1,
+                   rb_token_filters);
     if (!NIL_P(rb_normalizer))
         rb_funcall(rb_table, rb_intern("normalizer="), 1,
                    rb_normalizer);
@@ -340,7 +351,7 @@ rb_grn_patricia_trie_search (int argc, VALUE *argv, VALUE self)
  *
  * @example
  *   words = Groonga::PatriciaTrie.create(:key_type => "ShortText",
- *                                        :key_normalize => true)
+ *                                        :normalizer => "NormalizerAuto")
  *   words.add("リンク")
  *   adventure_of_link = words.add('リンクの冒険')
  *   words.add('冒険')
@@ -753,10 +764,11 @@ rb_grn_patricia_trie_open_rk_cursor (int argc, VALUE *argv, VALUE self)
 
 static grn_table_cursor *
 rb_grn_patricia_trie_open_grn_near_cursor (int argc, VALUE *argv, VALUE self,
-                                             grn_ctx **context, int flags)
+                                           grn_ctx **context, int flags)
 {
     grn_obj *table;
-    grn_obj *key_p = NULL, casted_key;
+    grn_id domain_id;
+    grn_obj *key, *domain;
     grn_table_cursor *cursor;
     unsigned min_size = 0;
     int offset = 0, limit = -1;
@@ -765,10 +777,10 @@ rb_grn_patricia_trie_open_grn_near_cursor (int argc, VALUE *argv, VALUE self,
 
     flags |= GRN_CURSOR_PREFIX;
 
-    rb_grn_table_deconstruct((RbGrnTable *)SELF(self), &table, context,
-                             NULL, NULL,
-                             NULL, NULL, NULL,
-                             NULL);
+    rb_grn_table_key_support_deconstruct(SELF(self), &table, context,
+                                         &key, &domain_id, &domain,
+                                         NULL, NULL, NULL,
+                                         NULL);
 
     rb_scan_args(argc, argv, "11", &rb_key, &options);
 
@@ -780,15 +792,7 @@ rb_grn_patricia_trie_open_grn_near_cursor (int argc, VALUE *argv, VALUE self,
                         "less_than", &rb_less_than,
                         NULL);
 
-    key_p = RVAL2GRNBULK_WITH_TYPE(rb_key, *context, key_p,
-                                   table->header.domain,
-                                   grn_ctx_at(*context, table->header.domain));
-    GRN_OBJ_INIT(&casted_key, GRN_BULK, 0, table->header.domain);
-    if (key_p->header.domain != table->header.domain) {
-        grn_obj_cast(*context, key_p, &casted_key, GRN_FALSE);
-        grn_obj_unlink(*context, key_p);
-        key_p = &casted_key;
-    }
+    RVAL2GRNBULK_WITH_TYPE(rb_key, *context, key, domain_id, domain);
 
     if (!NIL_P(rb_min_size))
         min_size = NUM2UINT(rb_min_size);
@@ -804,9 +808,8 @@ rb_grn_patricia_trie_open_grn_near_cursor (int argc, VALUE *argv, VALUE self,
 
     cursor = grn_table_cursor_open(*context, table,
                                    NULL, min_size,
-                                   GRN_BULK_HEAD(key_p), GRN_BULK_VSIZE(key_p),
+                                   GRN_BULK_HEAD(key), GRN_BULK_VSIZE(key),
                                    offset, limit, flags);
-    GRN_OBJ_FIN(*context, &casted_key);
     rb_grn_context_check(*context, self);
 
     return cursor;