rroonga 4.0.8 → 5.0.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: d2c33b482dafe8f399af8b7470fd96e22550930d
-  data.tar.gz: 6d52fe1e3cceffd71e9b1b93ce64549d0669d806
+  metadata.gz: e080d48d06e47c3504eef74495e62193e30b3ff8
+  data.tar.gz: bd3b0cd6fe3b2a26e550d9c7b8dfc16e786710c3
 SHA512:
-  metadata.gz: 82734af6ca9ecc542d94d5880e2203668f8e90cf14113172d0312b71b8200a9436176054647681f708c29f89a9f175163971b053334d75c25552095166fe272d
-  data.tar.gz: e74bfda44591c57ac21bfa6ef6b6a107cb0ab0471b3c6fde5cbf187124e6689a5bac6c9a4a3c9c12d5ade497b3cf7dff72eb00a819d7236c74e83f58b5349bf4
+  metadata.gz: b5c5eb1bf9999062382965a37ee20dd8dcdfffd9afd50516c8a1882165f41231602c6ed1c80efab9fbaa97272d8b2a17aa07db599d7c339621e6285cbf8e6327
+  data.tar.gz: 77a6604d8fd4c3831a2ee42016c1e25a8273623812d9a8582830458163f08a6ec735fe5a64af064b7a67b3150bfa429d3b0212d1decc719612390fadffd1c318

data/README.md

@@ -40,7 +40,7 @@ contributed patches.)
 ## Dependencies
 
 * Ruby >= 1.9.3
-* Groonga >= 4.0.9
+* Groonga >= 5.0.0
 
 ## Install
 
@@ -59,7 +59,7 @@ contributed patches.)
 ## Thanks
 
 * Daijiro MORI: sent patches to support the latest Groonga.
-* Tasuku SUENAGA: sent bug reports.
+* Tasuku SUENAGA: sent patches and bug reports.
 * niku: sent bug reports.
 * dara:
   * wrote tests.

data/doc/text/news.textile

@@ -1,5 +1,24 @@
 h1. NEWS
 
+h2(#5-0-0). 5.0.0: 2014-02-16 (planned)
+
+h3. Improvements
+
+* Supported Groonga 5.0.0. Groonga 4.1.1 or older aren't supported.
+* Added flags for {Groonga::Normalizer.normalize}.
+  [GitHub#44] [Patch by Tasuku SUENAGA a.k.a. gunyarakun]
+  <pre>
+  Groonga::Normalizer.normalize("AbC Def　gh")                         #=> "abcdefgh"
+  Groonga::Normalizer.normalize("AbC Def　gh", :remove_blank => true)  #=> "abcdefgh"
+  Groonga::Normalizer.normalize("AbC Def　gh", :remove_blank => false) #=> "abc def gh"
+  </pre>
+* Supported drilldown by multiple keys in {Groonga::Table#group}.
+* Supported calculation for drilldown in {Groonga::Table#group}.
+
+h3. Thanks
+
+* Tasuku SUENAGA a.k.a. gunyarakun
+
 h2(#4-0-8). 4.0.8: 2014-01-08
 
 h3. Improvements

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

@@ -1,7 +1,7 @@
 /* -*- coding: utf-8; mode: C; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
 /*
+  Copyright (C) 2014-2015  Masafumi Yokoyama <yokoyama@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
@@ -97,6 +97,8 @@ static VALUE eGrnCASError;
 static VALUE eGrnUnsupportedCommandVersion;
 static VALUE eGrnNormalizerError;
 static VALUE eGrnTokenFilterError;
+static VALUE eGrnCommandError;
+static VALUE eGrnPluginError;
 
 VALUE
 rb_grn_rc_to_exception (grn_rc rc)
@@ -329,6 +331,12 @@ rb_grn_rc_to_exception (grn_rc rc)
       case GRN_TOKEN_FILTER_ERROR:
         exception = eGrnTokenFilterError;
         break;
+      case GRN_COMMAND_ERROR:
+        exception = eGrnCommandError;
+        break;
+      case GRN_PLUGIN_ERROR:
+        exception = eGrnPluginError;
+        break;
     }
 
     if (NIL_P(exception))
@@ -568,6 +576,12 @@ rb_grn_rc_to_message (grn_rc rc)
       case GRN_TOKEN_FILTER_ERROR:
         message = "token filter error";
         break;
+      case GRN_COMMAND_ERROR:
+        message = "command error";
+        break;
+      case GRN_PLUGIN_ERROR:
+        message = "plugin error";
+        break;
     }
 
     if (!message)
@@ -1227,4 +1241,24 @@ rb_grn_init_exception (VALUE mGrn)
      */
     eGrnTokenFilterError =
         rb_define_class_under(mGrn, "TokenFilterError", rb_eGrnError);
+
+    /*
+     * Document-class: Groonga::CommandError
+     *
+     * It is used when a command causes an error.
+     *
+     * @since 5.0.0
+     */
+    eGrnCommandError =
+        rb_define_class_under(mGrn, "CommandError", rb_eGrnError);
+
+    /*
+     * Document-class: Groonga::PluginError
+     *
+     * It is used when a plugin causes an error.
+     *
+     * @since 5.0.0
+     */
+    eGrnPluginError =
+        rb_define_class_under(mGrn, "PluginError", rb_eGrnError);
 }

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

@@ -35,32 +35,62 @@ VALUE rb_cGrnNormalizer;
  *   # Normalizes "ABC" with the default normalizer
  *   Groonga::Normalizer.normalize("AbC") # => "abc"
  *
- * @overload normalize(string)
- *   @return [String] The normalized string
+ * @overload normalize(string, options={:remove_blank => true})
  *   @param [String] string The original string
+ *
+ *   @param [::Hash] options The optional parameters.
+ *   @option options :remove_blank (true)
+ *     If it's `true`, all blank characters are removed.
+ *   @option options :remove_tokenized_delimiter (false)
+ *     If it's `true`, all tokenized delimiter characters are removed.
+ *     The character is `U+FFFE`.
+ *
+ *   @return [String] The normalized string
  */
 static VALUE
-rb_grn_normalizer_s_normalize (VALUE klass, VALUE rb_string)
+rb_grn_normalizer_s_normalize (int argc, VALUE *argv, VALUE klass)
 {
     VALUE rb_context = Qnil;
+    VALUE rb_string;
+    VALUE rb_options;
+    VALUE rb_remove_blank_p;
+    VALUE rb_remove_tokenized_delimiter_p;
     VALUE rb_encoded_string;
     VALUE rb_normalized_string;
     grn_ctx *context = NULL;
     grn_obj *grn_string;
     grn_obj *normalizer = GRN_NORMALIZER_AUTO;
-    /* TODO: make customizable */
-    int flags = GRN_STRING_REMOVE_BLANK;
+    int flags = 0;
     const char *normalized_string;
     unsigned int normalized_string_length;
 
+    rb_scan_args(argc, argv, "11", &rb_string, &rb_options);
+    rb_grn_scan_options(rb_options,
+                        "remove_blank", &rb_remove_blank_p,
+                        "remove_tokenized_delimiter_p",
+                        &rb_remove_tokenized_delimiter_p,
+                        NULL);
+
     context = rb_grn_context_ensure(&rb_context);
     rb_encoded_string = rb_grn_context_rb_string_encode(context, rb_string);
+    if (NIL_P(rb_remove_blank_p)) {
+        rb_remove_blank_p = Qtrue;
+    }
+    if (RVAL2CBOOL(rb_remove_blank_p)) {
+        flags |= GRN_STRING_REMOVE_BLANK;
+    }
+    if (NIL_P(rb_remove_tokenized_delimiter_p)) {
+        rb_remove_tokenized_delimiter_p = Qfalse;
+    }
+    if (RVAL2CBOOL(rb_remove_tokenized_delimiter_p)) {
+        flags |= GRN_STRING_REMOVE_TOKENIZED_DELIMITER;
+    }
     grn_string = grn_string_open(context,
                                  RSTRING_PTR(rb_encoded_string),
                                  RSTRING_LEN(rb_encoded_string),
                                  normalizer,
                                  flags);
-    rb_grn_context_check(context, rb_string);
+    rb_grn_context_check(context, argv[0]);
     grn_string_get_normalized(context, grn_string,
                               &normalized_string, &normalized_string_length,
                               NULL);
@@ -79,5 +109,5 @@ rb_grn_init_normalizer (VALUE mGrn)
     rb_cGrnNormalizer = rb_define_class_under(mGrn, "Normalizer", rb_cObject);
 
     rb_define_singleton_method(rb_cGrnNormalizer, "normalize",
-                               rb_grn_normalizer_s_normalize, 1);
+                               rb_grn_normalizer_s_normalize, -1);
 }

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

@@ -1,7 +1,7 @@
 /* -*- coding: utf-8; mode: C; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
 /*
+  Copyright (C) 2014-2015  Masafumi Yokoyama <yokoyama@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
@@ -35,7 +35,7 @@ static ID id_array_set;
 /*
  * Document-class: Groonga::Table < Groonga::Object
  *
- * rroongaが提供するテーブルのベースとなるクラス。このクラス
+ * Rroongaが提供するテーブルのベースとなるクラス。このクラス
  * から {Groonga::Array} , {Groonga::Hash} , {Groonga::PatriciaTrie}
  * が継承されている。
  */
@@ -1289,7 +1289,7 @@ rb_grn_table_sort (int argc, VALUE *argv, VALUE self)
     result = grn_table_create(context, NULL, 0, NULL, GRN_TABLE_NO_KEY,
                               NULL, table);
     /* use n_records that is return value from
-       grn_table_sort() when rroonga user become specifying
+       grn_table_sort() when Rroonga user become specifying
        output table. */
     grn_table_sort(context, table, offset, limit, result, keys, n_keys);
     exception = rb_grn_context_to_exception(context, self);
@@ -1306,13 +1306,55 @@ rb_grn_table_sort (int argc, VALUE *argv, VALUE self)
  * 値でグループ化する。多くの場合、キーにはカラムを指定する。
  * カラムはカラム名（文字列）でも指定可能。
  *
+ * If one key is specified, you can get the grouping key by {Groonga::Record#key}:
+ *
+ * <pre>
+ * !!!ruby
+ * grouped_records = table.group("key1").collect do |group|
+ *   [group.key, group.n_sub_records]
+ * end
+ * </pre>
+ *
+ * If two or more keys are specified, the default value of _:max_n_sub_records_
+ * is _1_ (When one key is specified, the value is _0_.). You can get the grouping
+ * keys by {Groonga::Record#sub_records}:
+ *
+ * <pre>
+ * !!!ruby
+ * grouped_records = table.group(["key1", "key2"]).collect do |group|
+ *   sub_record = group.sub_records.first
+ *   [
+ *     sub_record.key1,
+ *     sub_record.key2,
+ *     group.n_sub_records,
+ *   ]
+ * end
+ * </pre>
+ *
  * @overload group([key1, key2, ...], options={})
- *   @return [[Groonga::Hash, ...]]
+ *   @return [Groonga::Hash]
+ *
+ *   @!macro [new] table.group.options
+ *     @option options :max_n_sub_records
+ *       グループ化した後のレコードのそれぞれについて最大 _:max_n_sub_records_ 件まで
+ *       そのグループに含まれる _table_ のレコードをサブレコードとして格納する。
+ *     @option options [String or Symbol] :calc_target
+ *       The target column name for _:calc_types_.
+ *     @option options [::Array] :calc_types
+ *       It specifies how to calculate (aggregate) values in grouped records by
+ *       a drilldown. You can specify multiple calculation types.
+ *
+ *       - +:max+ := Finding the maximum integer value from integer values in grouped records.
+ *       - +:min+ := Finding the minimum integer value from integer values in grouped records.
+ *       - +:sum+ := Summing integer values in grouped records.
+ *       - +:average+ := Averaging integer/float values in grouped records.
+ *
+ *   @!macro table.group.options
+ *
  * @overload group(key, options={})
  *   @return [Groonga::Hash]
- * @option options :max_n_sub_records
- *   グループ化した後のレコードのそれぞれについて最大 _:max_n_sub_records_ 件まで
- *   そのグループに含まれる _table_ のレコードをサブレコードとして格納する。
+ *
+ *   @!macro table.group.options
  */
 static VALUE
 rb_grn_table_group (int argc, VALUE *argv, VALUE self)
@@ -1320,13 +1362,13 @@ rb_grn_table_group (int argc, VALUE *argv, VALUE self)
     grn_ctx *context = NULL;
     grn_obj *table;
     grn_table_sort_key *keys;
-    grn_table_group_result *results;
-    int i, n_keys, n_results;
+    grn_table_group_result result;
+    int i, n_keys;
     unsigned int max_n_sub_records = 0;
     grn_rc rc;
     VALUE rb_keys, rb_options, rb_max_n_sub_records;
+    VALUE rb_calc_target, rb_calc_types;
     VALUE *rb_group_keys;
-    VALUE rb_results;
 
     rb_grn_table_deconstruct(SELF(self), &table, &context,
                              NULL, NULL,
@@ -1345,10 +1387,19 @@ rb_grn_table_group (int argc, VALUE *argv, VALUE self)
 
     rb_grn_scan_options(rb_options,
                         "max_n_sub_records", &rb_max_n_sub_records,
+                        "calc_target", &rb_calc_target,
+                        "calc_types", &rb_calc_types,
                         NULL);
 
-    if (!NIL_P(rb_max_n_sub_records))
+    if (NIL_P(rb_max_n_sub_records)) {
+        if (n_keys > 1) {
+            max_n_sub_records = 1;
+        } else {
+            max_n_sub_records = 0;
+        }
+    } else {
         max_n_sub_records = NUM2UINT(rb_max_n_sub_records);
+    }
 
     keys = ALLOCA_N(grn_table_sort_key, n_keys);
     for (i = 0; i < n_keys; i++) {
@@ -1380,34 +1431,54 @@ rb_grn_table_group (int argc, VALUE *argv, VALUE self)
         keys[i].flags = 0;
     }
 
-    n_results = n_keys;
-    results = ALLOCA_N(grn_table_group_result, n_results);
-    rb_results = rb_ary_new();
-    for (i = 0; i < n_results; i++) {
-        grn_obj *result;
-        VALUE rb_result;
-
-        result = grn_table_create_for_group(context, NULL, 0, NULL,
-                                            keys[i].key, table, max_n_sub_records);
-        results[i].table = result;
-        results[i].key_begin = 0;
-        results[i].key_end = 0;
-        results[i].limit = 0;
-        results[i].flags = 0;
-        results[i].op = GRN_OP_OR;
-
-        rb_result = GRNOBJECT2RVAL(Qnil, context, result, GRN_TRUE);
-        rb_ary_push(rb_results, rb_result);
+    result.table = NULL;
+    result.key_begin = 0;
+    result.key_end = n_keys - 1;
+    result.limit = 1;
+    result.flags = GRN_TABLE_GROUP_CALC_COUNT;
+    result.op = GRN_OP_OR;
+    result.max_n_subrecs = max_n_sub_records;
+    result.calc_target = NULL;
+
+    if (!NIL_P(rb_calc_target)) {
+        const char *name = NULL;
+        unsigned name_size = 0;
+        ruby_object_to_column_name(rb_calc_target, &name, &name_size);
+        result.calc_target = grn_obj_column(context, table, name, name_size);
+    }
+    if (result.calc_target) {
+        int i, n_calc_types;
+        VALUE *raw_calc_types;
+        raw_calc_types = RARRAY_PTR(rb_calc_types);
+        n_calc_types = RARRAY_LEN(rb_calc_types);
+        for (i = 0; i < n_calc_types; i++) {
+            VALUE rb_calc_type = raw_calc_types[i];
+            if (rb_grn_equal_option(rb_calc_type, "max")) {
+                result.flags |= GRN_TABLE_GROUP_CALC_MAX;
+            } else if (rb_grn_equal_option(rb_calc_type, "min")) {
+                result.flags |= GRN_TABLE_GROUP_CALC_MIN;
+            } else if (rb_grn_equal_option(rb_calc_type, "sum")) {
+                result.flags |= GRN_TABLE_GROUP_CALC_SUM;
+            } else if (rb_grn_equal_option(rb_calc_type, "average")) {
+                result.flags |= GRN_TABLE_GROUP_CALC_AVG;
+            } else {
+                grn_obj_unlink(context, result.calc_target);
+                rb_raise(rb_eArgError,
+                         "invalid calculation type: %s: "
+                         "available types: [:max, :min, :sum, :average]",
+                         rb_grn_inspect(rb_calc_type));
+            }
+        }
     }
 
-    rc = grn_table_group(context, table, keys, n_keys, results, n_results);
+    rc = grn_table_group(context, table, keys, n_keys, &result, 1);
+    if (result.calc_target) {
+        grn_obj_unlink(context, result.calc_target);
+    }
     rb_grn_context_check(context, self);
     rb_grn_rc_check(rc, self);
 
-    if (n_results == 1)
-        return rb_ary_pop(rb_results);
-    else
-        return rb_results;
+    return GRNOBJECT2RVAL(Qnil, context, result.table, GRN_TRUE);
 }
 
 /*
@@ -2380,7 +2451,7 @@ rb_grn_table_support_sub_records_p (VALUE self)
  * records for each group, the number of records is stored to
  * it. Records for each group are called as "sub records".
  *
- * Normally, you don't need to care about the space because rroonga
+ * Normally, you don't need to care about the space because Rroonga
  * creates a table with the space automatically. Normal tables
  * (persistent tables) don't have the space because they don't need
  * it.

data/ext/groonga/rb-grn.h

@@ -1,5 +1,6 @@
 /* -*- coding: utf-8; mode: C; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
 /*
+  Copyright (C) 2015  Masafumi Yokoyama <yokoyama@clear-code.com>
   Copyright (C) 2009-2014  Kouhei Sutou <kou@clear-code.com>
 
   This library is free software; you can redistribute it and/or
@@ -90,9 +91,9 @@ RB_GRN_BEGIN_DECLS
 #  define debug(...)
 #endif
 
-#define RB_GRN_MAJOR_VERSION 4
+#define RB_GRN_MAJOR_VERSION 5
 #define RB_GRN_MINOR_VERSION 0
-#define RB_GRN_MICRO_VERSION 8
+#define RB_GRN_MICRO_VERSION 0
 
 #define RB_GRN_QUERY_DEFAULT_MAX_EXPRESSIONS 32
 

data/lib/groonga/record.rb

@@ -1,6 +1,6 @@
 # -*- coding: utf-8 -*-
 #
-# Copyright (C) 2014  Masafumi Yokoyama <myokoym@gmail.com>
+# Copyright (C) 2014-2015  Masafumi Yokoyama <yokoyama@clear-code.com>
 # Copyright (C) 2009-2014  Kouhei Sutou <kou@clear-code.com>
 #
 # This library is free software; you can redistribute it and/or
@@ -226,6 +226,50 @@ module Groonga
       @table.support_sub_records?
     end
 
+    # The maximum integer value from integer values in grouped records.
+    # It can be used when specifying _:calc_target_ and _:calc_types_
+    # to {Groonga::Table#group}.
+    #
+    # @return [Integer] The maximum integer value from integer values in grouped records.
+    #
+    # @since 5.0.0
+    def max
+      self["_max"]
+    end
+
+    # The minimum integer value from integer values in grouped records.
+    # It can be used when specifying _:calc_target_ and _:calc_types_
+    # to {Groonga::Table#group}.
+    #
+    # @return [Integer] The minimum integer value from integer values in grouped records.
+    #
+    # @since 5.0.0
+    def min
+      self["_min"]
+    end
+
+    # The sum of integer values in grouped records.
+    # It can be used when specifying _:calc_target_ and _:calc_types_
+    # to {Groonga::Table#group}.
+    #
+    # @return [Integer] The sum of integer values in grouped records.
+    #
+    # @since 5.0.0
+    def sum
+      self["_sum"]
+    end
+
+    # The average of integer/float values in grouped records.
+    # It can be used when specifying _:calc_target_ and _:calc_types_
+    # to {Groonga::Table#group}.
+    #
+    # @return [Float] The average of integer/float values in grouped records.
+    #
+    # @since 5.0.0
+    def average
+      self["_avg"]
+    end
+
     # レコードの値を返す。
     def value
       @table.value(@id, :id => true)