Support prefix RK search by '&^~' with prefix_search_ops_v2

pgroonga · Apr 7, 2016 · fed2054 · fed2054
1 parent 6903ceb
commit fed2054
Show file tree

Hide file tree

Showing 11 changed files with 376 additions and 49 deletions.
diff --git a/expected/prefix/text/prefix-rk/bitmapscan.out b/expected/prefix/text/prefix-rk/bitmapscan.out
@@ -0,0 +1,23 @@
+CREATE TABLE readings (
+  katakana text
+);
+INSERT INTO readings VALUES ('ポストグレスキューエル');
+INSERT INTO readings VALUES ('グルンガ');
+INSERT INTO readings VALUES ('ピージールンガ');
+INSERT INTO readings VALUES ('ピージーロジカル');
+CREATE INDEX pgrn_index ON readings
+  USING pgroonga (katakana pgroonga.prefix_search_ops_v2);
+SET enable_seqscan = off;
+SET enable_indexscan = off;
+SET enable_bitmapscan = on;
+SELECT katakana
+  FROM readings
+ WHERE katakana &^~ 'p';
+        katakana        
+------------------------
+ ポストグレスキューエル
+ ピージールンガ
+ ピージーロジカル
+(3 rows)
+
+DROP TABLE readings;
diff --git a/expected/prefix/text/prefix-rk/indexscan.out b/expected/prefix/text/prefix-rk/indexscan.out
@@ -0,0 +1,23 @@
+CREATE TABLE readings (
+  katakana text
+);
+INSERT INTO readings VALUES ('ポストグレスキューエル');
+INSERT INTO readings VALUES ('グルンガ');
+INSERT INTO readings VALUES ('ピージールンガ');
+INSERT INTO readings VALUES ('ピージーロジカル');
+CREATE INDEX pgrn_index ON readings
+  USING pgroonga (katakana pgroonga.prefix_search_ops_v2);
+SET enable_seqscan = off;
+SET enable_indexscan = on;
+SET enable_bitmapscan = off;
+SELECT katakana
+  FROM readings
+ WHERE katakana &^~ 'p';
+        katakana        
+------------------------
+ ピージールンガ
+ ピージーロジカル
+ ポストグレスキューエル
+(3 rows)
+
+DROP TABLE readings;
diff --git a/expected/prefix/text/prefix-rk/seqscan.out b/expected/prefix/text/prefix-rk/seqscan.out
@@ -0,0 +1,21 @@
+CREATE TABLE readings (
+  katakana text
+);
+INSERT INTO readings VALUES ('ポストグレスキューエル');
+INSERT INTO readings VALUES ('グルンガ');
+INSERT INTO readings VALUES ('ピージールンガ');
+INSERT INTO readings VALUES ('ピージーロジカル');
+SET enable_seqscan = on;
+SET enable_indexscan = off;
+SET enable_bitmapscan = off;
+SELECT katakana
+  FROM readings
+ WHERE katakana &^~ 'p';
+        katakana        
+------------------------
+ ポストグレスキューエル
+ ピージールンガ
+ ピージーロジカル
+(3 rows)
+
+DROP TABLE readings;
diff --git a/pgroonga.sql b/pgroonga.sql
@@ -206,7 +206,7 @@ CREATE FUNCTION pgroonga.options(internal)
 DELETE FROM pg_catalog.pg_am WHERE amname = 'pgroonga';
 INSERT INTO pg_catalog.pg_am VALUES(
 	'pgroonga',	-- amname
-	18,		-- amstrategies
+	19,		-- amstrategies
 	0,		-- amsupport
 	true,		-- amcanorder
 	true,		-- amcanorderbyop
@@ -427,6 +427,19 @@ CREATE OPERATOR &^ (
 	RIGHTARG = text
 );
 
+CREATE FUNCTION pgroonga.prefix_rk_text(text, text)
+	RETURNS bool
+	AS 'MODULE_PATHNAME', 'pgroonga_prefix_rk_text'
+	LANGUAGE C
+	IMMUTABLE
+	STRICT;
+
+CREATE OPERATOR &^~ (
+	PROCEDURE = pgroonga.prefix_rk_text,
+	LEFTARG = text,
+	RIGHTARG = text
+);
+
 CREATE FUNCTION pgroonga.script_text(text, text)
 	RETURNS bool
 	AS 'MODULE_PATHNAME', 'pgroonga_script_text'
@@ -473,10 +486,11 @@ CREATE OPERATOR CLASS pgroonga.text_full_text_search_ops_v2 FOR TYPE text
 		OPERATOR 12 &@,
 		OPERATOR 13 &?,
 		OPERATOR 14 &~?,
-		OPERATOR 16 &`,
-		OPERATOR 17 &@> (text, text[]),
-		OPERATOR 18 &?> (text, text[]);
+		OPERATOR 17 &`,
+		OPERATOR 18 &@> (text, text[]),
+		OPERATOR 19 &?> (text, text[]);
 
 CREATE OPERATOR CLASS pgroonga.prefix_search_ops_v2 FOR TYPE text
 	USING pgroonga AS
-		OPERATOR 15 &^;
+		OPERATOR 15 &^,
+		OPERATOR 16 &^~;
diff --git a/sql/prefix/text/prefix-rk/bitmapscan.sql b/sql/prefix/text/prefix-rk/bitmapscan.sql
@@ -0,0 +1,21 @@
+CREATE TABLE readings (
+  katakana text
+);
+
+INSERT INTO readings VALUES ('ポストグレスキューエル');
+INSERT INTO readings VALUES ('グルンガ');
+INSERT INTO readings VALUES ('ピージールンガ');
+INSERT INTO readings VALUES ('ピージーロジカル');
+
+CREATE INDEX pgrn_index ON readings
+  USING pgroonga (katakana pgroonga.prefix_search_ops_v2);
+
+SET enable_seqscan = off;
+SET enable_indexscan = off;
+SET enable_bitmapscan = on;
+
+SELECT katakana
+  FROM readings
+ WHERE katakana &^~ 'p';
+
+DROP TABLE readings;
diff --git a/sql/prefix/text/prefix-rk/indexscan.sql b/sql/prefix/text/prefix-rk/indexscan.sql
@@ -0,0 +1,21 @@
+CREATE TABLE readings (
+  katakana text
+);
+
+INSERT INTO readings VALUES ('ポストグレスキューエル');
+INSERT INTO readings VALUES ('グルンガ');
+INSERT INTO readings VALUES ('ピージールンガ');
+INSERT INTO readings VALUES ('ピージーロジカル');
+
+CREATE INDEX pgrn_index ON readings
+  USING pgroonga (katakana pgroonga.prefix_search_ops_v2);
+
+SET enable_seqscan = off;
+SET enable_indexscan = on;
+SET enable_bitmapscan = off;
+
+SELECT katakana
+  FROM readings
+ WHERE katakana &^~ 'p';
+
+DROP TABLE readings;
diff --git a/sql/prefix/text/prefix-rk/seqscan.sql b/sql/prefix/text/prefix-rk/seqscan.sql
@@ -0,0 +1,18 @@
+CREATE TABLE readings (
+  katakana text
+);
+
+INSERT INTO readings VALUES ('ポストグレスキューエル');
+INSERT INTO readings VALUES ('グルンガ');
+INSERT INTO readings VALUES ('ピージールンガ');
+INSERT INTO readings VALUES ('ピージーロジカル');
+
+SET enable_seqscan = on;
+SET enable_indexscan = off;
+SET enable_bitmapscan = off;
+
+SELECT katakana
+  FROM readings
+ WHERE katakana &^~ 'p';
+
+DROP TABLE readings;
diff --git a/src/pgrn_create.c b/src/pgrn_create.c
@@ -36,6 +36,20 @@ void
 PGrnCreateDataColumn(PGrnCreateData *data)
 {
 	grn_obj_flags flags = 0;
+	grn_obj *range;
+	grn_id rangeID;
+
+	if (data->forPrefixSearch) {
+		char lexiconName[GRN_TABLE_MAX_KEY_SIZE];
+
+		snprintf(lexiconName, sizeof(lexiconName),
+				 PGrnLexiconNameFormat, data->relNode, data->i);
+		range = PGrnLookup(lexiconName, ERROR);
+		rangeID = grn_obj_id(ctx, range);
+	} else {
+		rangeID = data->attributeTypeID;
+		range = grn_ctx_at(ctx, rangeID);
+	}
 
 	if (data->attributeFlags & GRN_OBJ_VECTOR)
 	{
@@ -47,7 +61,7 @@ PGrnCreateDataColumn(PGrnCreateData *data)
 
 		if (PGrnIsLZ4Available)
 		{
-			switch (data->attributeTypeID)
+			switch (rangeID)
 			{
 			case GRN_DB_SHORT_TEXT:
 			case GRN_DB_TEXT:
@@ -65,12 +79,12 @@ PGrnCreateDataColumn(PGrnCreateData *data)
 		PGrnCreateColumn(data->sourcesTable,
 						 columnName,
 						 flags,
-						 grn_ctx_at(ctx, data->attributeTypeID));
+						 range);
 	}
 }
 
 void
-PGrnCreateIndexColumn(PGrnCreateData *data)
+PGrnCreateLexicon(PGrnCreateData *data)
 {
 	grn_id typeID = GRN_ID_NIL;
 	char lexiconName[GRN_TABLE_MAX_KEY_SIZE];
@@ -124,14 +138,23 @@ PGrnCreateIndexColumn(PGrnCreateData *data)
 							 PGrnLookup(normalizerName, ERROR));
 		}
 	}
+}
 
-	{
-		grn_obj_flags flags = GRN_OBJ_COLUMN_INDEX;
-		if (data->forFullTextSearch || data->forRegexpSearch)
-			flags |= GRN_OBJ_WITH_POSITION;
-		PGrnCreateColumn(lexicon,
-						 PGrnIndexColumnName,
-						 flags,
-						 data->sourcesTable);
-	}
+void
+PGrnCreateIndexColumn(PGrnCreateData *data)
+{
+	char lexiconName[GRN_TABLE_MAX_KEY_SIZE];
+	grn_obj *lexicon;
+	grn_obj_flags flags = GRN_OBJ_COLUMN_INDEX;
+
+	snprintf(lexiconName, sizeof(lexiconName),
+			 PGrnLexiconNameFormat, data->relNode, data->i);
+	lexicon = PGrnLookup(lexiconName, ERROR);
+
+	if (data->forFullTextSearch || data->forRegexpSearch)
+		flags |= GRN_OBJ_WITH_POSITION;
+	PGrnCreateColumn(lexicon,
+					 PGrnIndexColumnName,
+					 flags,
+					 data->sourcesTable);
 }
diff --git a/src/pgrn_create.h b/src/pgrn_create.h
@@ -24,5 +24,6 @@ typedef struct PGrnCreateData
 
 void PGrnCreateSourcesCtidColumn(PGrnCreateData *data);
 void PGrnCreateSourcesTable(PGrnCreateData *data);
+void PGrnCreateLexicon(PGrnCreateData *data);
 void PGrnCreateDataColumn(PGrnCreateData *data);
 void PGrnCreateIndexColumn(PGrnCreateData *data);