Merge pull request #173 from CartoDB/better-unique-identifiers

Better unique identifiers
2015-10-27 14:19:50 +01:00 · 2015-10-27 14:19:50 +01:00 · e99231252e
commit e99231252e
parent cf2587ca54 d171afb9f4
7 changed files with 374 additions and 67 deletions
--- a/4
+++ b/4
@ -1,7 +1,7 @@
 # cartodb/Makefile

 EXTENSION = cartodb
-EXTVERSION = 0.11.2
+EXTVERSION = 0.11.3

 SED = sed

@ -11,6 +11,7 @@ CDBSCRIPTS = \
  scripts-available/CDB_DDLTriggers.sql \
  scripts-available/CDB_ExtensionPost.sql \
  scripts-available/CDB_ExtensionUtils.sql \
+  scripts-available/CDB_Helper.sql \
  $(END)

 UPGRADABLE = \
@ -53,6 +54,7 @@ UPGRADABLE = \
  0.11.0 \
  0.11.1 \
  0.11.2 \
+  0.11.3 \
  $(EXTVERSION)dev \
  $(EXTVERSION)next \
  $(END)
--- a/NEWS.md
+++ b/NEWS.md
@ -1,6 +1,13 @@
 next (2015-mm-dd)
 -----------------

+0.11.3 (2015-10-27)
+-------------------
+* Added CDB_Helper.sql [#173](https://github.com/CartoDB/cartodb-postgresql/pull/173)
+* Added _CDB_Unique_Identifier for creating UTF8 aware unique identifiers
+* Added _CDB_Unique_Column_Identifier for creating UTF8 aware unique identifiers for columns
+* Added _CDB_Octet_Truncate that truncates text to a certain amount of octets.
+
 0.11.2 (2015-10-19)
 -------------------
 * Fix schema not being specified on pg_get_serial_sequence [#170](https://github.com/CartoDB/cartodb-postgresql/pull/170)
--- a/scripts-available/CDB_CartodbfyTable.sql
+++ b/scripts-available/CDB_CartodbfyTable.sql
@ -1,4 +1,5 @@
 -- Depends on:
+--   * CDB_Helper.sql
 --   * CDB_ExtensionUtils.sql
 --   * CDB_TransformToWebmercator.sql
 --   * CDB_TableMetadata.sql
@ -427,8 +428,9 @@ END;
 $$ LANGUAGE 'plpgsql';


-- Find a unique relation name in the given schema, starting from the 
-- template given. If the template is already unique, just return it; 
+-- DEPRECATED: Use _CDB_Unique_Identifier since it's UTF8 Safe and length
+-- aware. Find a unique relation name in the given schema, starting from the
+-- template given. If the template is already unique, just return it;
 -- otherwise, append an increasing integer until you find a unique variant.
 CREATE OR REPLACE FUNCTION _CDB_Unique_Relation_Name(schemaname TEXT, relationname TEXT)
 RETURNS TEXT
@ -439,36 +441,15 @@ DECLARE
  newrelname TEXT;
 BEGIN

-  i := 0;
-  newrelname := relationname;
-  LOOP
+  RAISE EXCEPTION '_CDB_Unique_Relation_Name is DEPRECATED. Use _CDB_Unique_Identifier(prefix TEXT, relname TEXT, suffix TEXT, schema TEXT DEFAULT NULL)';

-    SELECT c.relname, n.nspname
-    INTO rec
-    FROM pg_class c
-    JOIN pg_namespace n ON c.relnamespace = n.oid
-    WHERE c.relname = newrelname
-    AND n.nspname = schemaname;
-  
-    IF NOT FOUND THEN
-      RETURN newrelname;
-    END IF;
-    
-    i := i + 1;
-    newrelname := relationname || '_' || i;
-  
-    IF i > 100 THEN
-      PERFORM _CDB_Error('looping too far', '_CDB_Unique_Relation_Name');
-    END IF;
-  
-  END LOOP;
-  
 END;
 $$ LANGUAGE 'plpgsql';


-- Find a unique column name in the given relation, starting from the 
-- column name given. If the column name is already unique, just return it; 
+-- DEPRECATED: Use _CDB_Unique_Column_Identifier since it's UTF8 Safe and length
+-- aware. Find a unique column name in the given relation, starting from the
+-- column name given. If the column name is already unique, just return it;
 -- otherwise, append an increasing integer until you find a unique variant.
 CREATE OR REPLACE FUNCTION _CDB_Unique_Column_Name(reloid REGCLASS, columnname TEXT)
 RETURNS TEXT
@ -479,32 +460,8 @@ DECLARE
  newcolname TEXT;
 BEGIN

-  i := 0;
-  newcolname := columnname;
-  LOOP
+  RAISE EXCEPTION '_CDB_Unique_Column_Name is DEPRECATED. Use _CDB_Unique_Column_Identifier(prefix TEXT, relname TEXT, suffix TEXT, reloid REGCLASS DEFAULT NULL)';

-    SELECT a.attname
-    INTO rec
-    FROM pg_class c
-    JOIN pg_attribute a ON a.attrelid = c.oid
-    WHERE NOT a.attisdropped
-    AND a.attnum > 0
-    AND c.oid = reloid
-    AND a.attname = newcolname;
-  
-    IF NOT FOUND THEN
-      RETURN newcolname;
-    END IF;
-    
-    i := i + 1;
-    newcolname := columnname || '_' || i;
-  
-    IF i > 100 THEN
-      PERFORM _CDB_Error('looping too far', '_CDB_Unique_Column_Name');
-    END IF;
-  
-  END LOOP;
-  
 END;
 $$ LANGUAGE 'plpgsql';

@ -589,7 +546,7 @@ BEGIN
          PERFORM _CDB_SQL(
            Format('ALTER TABLE %s RENAME COLUMN %s TO %I',
              reloid::text, rec.attname,
-              _CDB_Unique_Column_Name(reloid, const.pkey)),
+              cartodb._CDB_Unique_Column_Identifier(NULL, const.pkey, NULL, reloid)),
            '_CDB_Has_Usable_Primary_ID');
        
        END IF;
@ -606,7 +563,7 @@ BEGIN

      PERFORM _CDB_SQL(
        Format('ALTER TABLE %s RENAME COLUMN %s TO %I',
-                reloid::text, rec.attname, _CDB_Unique_Column_Name(reloid, const.pkey)),
+                reloid::text, rec.attname, cartodb._CDB_Unique_Column_Identifier(NULL, const.pkey, NULL, reloid)),
                '_CDB_Has_Usable_Primary_ID');
    
    END IF;
@ -772,7 +729,7 @@ BEGIN
          WHEN others THEN
            IF SQLERRM = 'parse error - invalid geometry' THEN
              text_geom_column := false;
-              str := _CDB_Unique_Column_Name(reloid, r1.attname);
+              str := cartodb._CDB_Unique_Column_Identifier(NULL, r1.attname, NULL, reloid);
              sql := Format('ALTER TABLE %s RENAME COLUMN %s TO %I', reloid::text, r1.attname, str);
              PERFORM _CDB_SQL(sql,'_CDB_Has_Usable_Geom');
              RAISE DEBUG 'CDB(_CDB_Has_Usable_Geom): %', 
@ -784,7 +741,7 @@ BEGIN

      -- Just change its name so we can write a new column into that name.
      ELSE
-        str := _CDB_Unique_Column_Name(reloid, r1.attname);
+        str := cartodb._CDB_Unique_Column_Identifier(NULL, r1.attname, NULL, reloid);
        sql := Format('ALTER TABLE %s RENAME COLUMN %s TO %I', reloid::text, r1.attname, str);
        PERFORM _CDB_SQL(sql,'_CDB_Has_Usable_Geom');
        RAISE DEBUG 'CDB(_CDB_Has_Usable_Geom): %', 
@ -854,8 +811,7 @@ DECLARE
  destname TEXT;
  destseq TEXT;
  destseqmax INTEGER;
-    
-  salt TEXT := md5(random()::text || now());
+
  copyname TEXT;

  column_name_sql TEXT;
@ -961,19 +917,18 @@ BEGIN
  -- Put the primary key sequence in the right schema
  -- If the new table is not moving, better ensure the sequence name
  -- is unique
-  destseq := relname || '_' || const.pkey || '_seq';
-  destseq := _CDB_Unique_Relation_Name(destschema, destseq);
+  destseq := cartodb._CDB_Unique_Identifier(NULL, relname, '_' || const.pkey || '_seq', destschema);
  destseq := Format('%I.%I', destschema, destseq);
  PERFORM _CDB_SQL(Format('CREATE SEQUENCE %s', destseq), '_CDB_Rewrite_Table');

-  -- Salt a temporary table name if we are re-writing in place
+  -- Temporary table name if we are re-writing in place
  -- Note copyname is already escaped and safe to use as identifier
  IF destschema = relschema THEN
-    copyname := Format('%I.%I', destschema, Format('%s_%s', destname, salt));
+    copyname := Format('%I.%I', destschema, cartodb._CDB_Unique_Identifier(NULL, destname, NULL), destschema);
  ELSE
    copyname := Format('%I.%I', destschema, destname);
  END IF;
-  
+
  -- Start building the SQL!
  sql := Format('CREATE TABLE %s AS SELECT ', copyname);

--- a/scripts-available/CDB_Helper.sql
+++ b/scripts-available/CDB_Helper.sql
@ -0,0 +1,156 @@
+-- UTF8 safe and length aware. Find a unique identifier with a given prefix
+-- and/or suffix and withing a schema. If a schema is not specified, the identifier
+-- is guaranteed to be unique for all schemas.
+CREATE OR REPLACE FUNCTION cartodb._CDB_Unique_Identifier(prefix TEXT, relname TEXT, suffix TEXT, schema TEXT DEFAULT NULL)
+RETURNS TEXT
+AS $$
+DECLARE
+  maxlen CONSTANT INTEGER := 63;
+
+  rec RECORD;
+  usedspace INTEGER;
+  ident TEXT;
+  origident TEXT;
+  candrelname TEXT;
+
+  i INTEGER;
+BEGIN
+  -- Accounts for the _XX incremental suffix in case the identifier is taken
+  usedspace := 3;
+  usedspace := usedspace + coalesce(octet_length(prefix), 0);
+  usedspace := usedspace + coalesce(octet_length(suffix), 0);
+
+  candrelname := _CDB_Octet_Truncate(relname, maxlen - usedspace);
+
+  IF candrelname = '' THEN
+    PERFORM _CDB_Error('prefixes are to long to generate a valid identifier', '_CDB_Unique_Identifier');
+  END IF;
+
+  ident := coalesce(prefix, '') || candrelname || coalesce(suffix, '');
+
+  i := 0;
+  origident := ident;
+
+  WHILE i < 100 LOOP
+    IF schema IS NOT NULL THEN
+      SELECT c.relname, n.nspname
+      INTO rec
+      FROM pg_class c
+      JOIN pg_namespace n ON c.relnamespace = n.oid
+      WHERE c.relname = ident
+      AND n.nspname = schema;
+    ELSE
+      SELECT c.relname, n.nspname
+      INTO rec
+      FROM pg_class c
+      JOIN pg_namespace n ON c.relnamespace = n.oid
+      WHERE c.relname = ident;
+    END IF;
+
+    IF NOT FOUND THEN
+      RETURN ident;
+    END IF;
+
+    ident := origident || '_' || i;
+    i := i + 1;
+  END LOOP;
+
+  PERFORM _CDB_Error('looping too far', '_CDB_Unique_Identifier');
+END;
+$$ LANGUAGE 'plpgsql';
+
+
+-- UTF8 safe and length aware. Find a unique identifier for a column with a given prefix
+-- and/or suffix based on colname and within a relation specified via reloid.
+CREATE OR REPLACE FUNCTION cartodb._CDB_Unique_Column_Identifier(prefix TEXT, colname TEXT, suffix TEXT, reloid REGCLASS)
+RETURNS TEXT
+AS $$
+DECLARE
+  maxlen CONSTANT INTEGER := 63;
+
+  rec RECORD;
+  candcolname TEXT;
+  usedspace INTEGER;
+  ident TEXT;
+  origident TEXT;
+
+  i INTEGER;
+BEGIN
+  -- Accounts for the _XX incremental suffix in case the identifier is taken
+  usedspace := 3;
+  usedspace := usedspace + coalesce(octet_length(prefix), 0);
+  usedspace := usedspace + coalesce(octet_length(suffix), 0);
+
+  candcolname := _CDB_Octet_Truncate(colname, maxlen - usedspace);
+
+  IF candcolname = '' THEN
+    PERFORM _CDB_Error('prefixes are to long to generate a valid identifier', '_CDB_Unique_Column_Identifier');
+  END IF;
+
+  ident := coalesce(prefix, '') || candcolname || coalesce(suffix, '');
+
+  i := 0;
+  origident := ident;
+
+  WHILE i < 100 LOOP
+    SELECT a.attname
+    INTO rec
+    FROM pg_class c
+    JOIN pg_attribute a ON a.attrelid = c.oid
+    WHERE NOT a.attisdropped
+    AND a.attnum > 0
+    AND c.oid = reloid
+    AND a.attname = ident;
+
+    IF NOT FOUND THEN
+      RETURN ident;
+    END IF;
+
+    ident := origident || '_' || i;
+    i := i + 1;
+  END LOOP;
+
+  PERFORM _CDB_Error('looping too far', '_CDB_Unique_Column_Identifier');
+END;
+$$ LANGUAGE 'plpgsql';
+
+
+-- Truncates a given string to a max_octets octets taking care
+-- not to leave characters in half. UTF8 safe.
+CREATE OR REPLACE FUNCTION cartodb._CDB_Octet_Truncate(string TEXT, max_octets INTEGER)
+RETURNS TEXT
+AS $$
+DECLARE
+  extcharlen CONSTANT INTEGER := octet_length('ñ');
+
+  expected INTEGER;
+  examined INTEGER;
+  strlen INTEGER;
+
+  i INTEGER;
+BEGIN
+
+  IF max_octets <= 0 THEN
+    RETURN '';
+  ELSIF max_octets >= octet_length(string) THEN
+    RETURN string;
+  END IF;
+
+  strlen := char_length(string);
+
+  expected := char_length(string);
+  examined := octet_length(string);
+
+  IF expected = examined THEN
+    RETURN left(string, max_octets);
+  END IF;
+
+  i := max_octets / extcharlen;
+
+  WHILE octet_length(left(string, i)) <= max_octets LOOP
+    i := i + 1;
+  END LOOP;
+
+  RETURN left(string, (i - 1));
+END;
+$$ LANGUAGE 'plpgsql';
--- a/test/CDB_CartodbfyTableTest.sql
+++ b/test/CDB_CartodbfyTableTest.sql
@ -291,7 +291,7 @@ INSERT INTO test VALUES
  (NULL), 
  (3);
 SELECT CDB_CartodbfyTableCheck('test', 'Table with null cartodb_id #148');
-SELECT cartodb_id, cartodb_id_1 from test;
+SELECT cartodb_id, cartodb_id_0 from test;
 DROP TABLE test;

 -- Table with non unique cartodb_id
@ -303,7 +303,7 @@ INSERT INTO test VALUES
  (2), 
  (2);
 SELECT CDB_CartodbfyTableCheck('test', 'Table with non unique cartodb_id #148');
-SELECT cartodb_id, cartodb_id_1 from test;
+SELECT cartodb_id, cartodb_id_0 from test;
 DROP TABLE test;

 -- Table with non unique and null cartodb_id
@ -316,7 +316,7 @@ INSERT INTO test VALUES
  (NULL),
  (2);
 SELECT CDB_CartodbfyTableCheck('test', 'Table with non unique and null cartodb_id #148');
-SELECT cartodb_id, cartodb_id_1 from test;
+SELECT cartodb_id, cartodb_id_0 from test;
 DROP TABLE test;


--- a/test/CDB_HelperTest.sql
+++ b/test/CDB_HelperTest.sql
@ -0,0 +1,128 @@
+-- Test unique identifier creation with normal length normal relname
+SELECT * FROM cartodb._CDB_Unique_Identifier(NULL, 'relname', NULL);
+
+-- Test unique identifier creation with prefix with normal length normal relname
+SELECT * FROM cartodb._CDB_Unique_Identifier('prefix_', 'relname', NULL);
+
+-- Test unique identifier creation with suffix with normal length normal relname
+SELECT * FROM cartodb._CDB_Unique_Identifier(NULL, 'relname', '_suffix');
+
+-- Test unique identifier creation with long length normal relname
+SELECT * FROM cartodb._CDB_Unique_Identifier(NULL, 'largolargolargolargolargolargolargolargolargolargolargolargolar', NULL);
+
+-- Test unique identifier creation with prefix with long length normal relname
+SELECT * FROM cartodb._CDB_Unique_Identifier('prefix_', 'largolargolargolargolargolargolargolargolargolargolargolargolar', NULL);
+
+-- Test new identifier is found when name is taken from previous case
+CREATE TABLE prefix_largolargolargolargolargolargolargolargolargolargolar (name text);
+SELECT * FROM cartodb._CDB_Unique_Identifier('prefix_', 'largolargolargolargolargolargolargolargolargolargolargolargolar', NULL);
+DROP TABLE prefix_largolargolargolargolargolargolargolargolargolargolar;
+
+-- Test unique identifier creation with suffix with long length normal relname
+SELECT * FROM cartodb._CDB_Unique_Identifier(NULL, 'largolargolargolargolargolargolargolargolargolargolargolargolar', '_suffix');
+
+-- Test new identifier is found when name is taken from previous case
+CREATE TABLE largolargolargolargolargolargolargolargolargolargolar_suffix (name text);
+SELECT * FROM cartodb._CDB_Unique_Identifier(NULL, 'largolargolargolargolargolargolargolargolargolargolargolargolar', '_suffix');
+DROP TABLE largolargolargolargolargolargolargolargolargolargolar_suffix;
+
+-- Test unique identifier creation with normal length UTF8 relname
+SELECT * FROM cartodb._CDB_Unique_Identifier(NULL, 'piraña', NULL);
+
+-- Test unique identifier creation with prefix with normal length UTF8 relname
+SELECT * FROM cartodb._CDB_Unique_Identifier('prefix_', 'piraña', NULL);
+
+-- Test unique identifier creation with suffix with normal length UTF8 relname
+SELECT * FROM cartodb._CDB_Unique_Identifier(NULL, 'piraña', '_suffix');
+
+-- Test unique identifier creation with long length UTF8 relname
+SELECT * FROM cartodb._CDB_Unique_Identifier(NULL, 'piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpiñaácidpin', NULL);
+
+-- Test unique identifier creation with prefix with long length UTF8 relname
+SELECT * FROM cartodb._CDB_Unique_Identifier('prefix_', 'piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpiñaácidpin', NULL);
+
+-- Test new identifier is found when name is taken from previous case
+CREATE TABLE prefix_piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpi (name text);
+SELECT * FROM cartodb._CDB_Unique_Identifier('prefix_', 'piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpiñaácidpin', NULL);
+DROP TABLE prefix_piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpi;
+
+-- Test unique identifier creation with suffix with long length UTF8 relname
+SELECT * FROM cartodb._CDB_Unique_Identifier(NULL, 'piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpiñaácidpin', '_suffix');
+
+-- Test new identifier is found when name is taken from previous case
+CREATE TABLE piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpi_suffix (name text);
+SELECT * FROM cartodb._CDB_Unique_Identifier(NULL, 'piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpiñaácidpin', '_suffix');
+DROP TABLE piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpi_suffix;
+
+CREATE TABLE test (name text);
+-- Test unique identifier creation with normal length normal colname
+SELECT * FROM cartodb._CDB_Unique_Column_Identifier(NULL, 'colname', NULL, 'test'::regclass);
+
+-- Test unique identifier creation with prefix with normal length normal colname
+SELECT * FROM cartodb._CDB_Unique_Column_Identifier('prefix_', 'colname', NULL, 'test'::regclass);
+
+-- Test unique identifier creation with suffix with normal length normal colname
+SELECT * FROM cartodb._CDB_Unique_Column_Identifier(NULL, 'colname', '_suffix', 'test'::regclass);
+
+-- Test unique identifier creation with long length normal colname
+SELECT * FROM cartodb._CDB_Unique_Column_Identifier(NULL, 'largolargolargolargolargolargolargolargolargolargolargolargolar', NULL, 'test'::regclass);
+
+-- Test unique identifier creation with prefix with long length normal colname
+SELECT * FROM cartodb._CDB_Unique_Column_Identifier('prefix_', 'largolargolargolargolargolargolargolargolargolargolargolargolar', NULL, 'test'::regclass);
+DROP TABLE test;
+
+-- Test new identifier is found when name is taken from previous case
+CREATE TABLE test (prefix_largolargolargolargolargolargolargolargolargolargolar text);
+SELECT * FROM cartodb._CDB_Unique_Column_Identifier('prefix_', 'largolargolargolargolargolargolargolargolargolargolargolargolar', NULL, 'test'::regclass);
+DROP TABLE test;
+
+-- Test unique identifier creation with suffix with long length normal colname
+CREATE TABLE test (name text);
+SELECT * FROM cartodb._CDB_Unique_Column_Identifier(NULL, 'largolargolargolargolargolargolargolargolargolargolargolargolar', '_suffix', 'test'::regclass);
+DROP TABLE test;
+
+-- Test new identifier is found when name is taken from previous case
+CREATE TABLE test (largolargolargolargolargolargolargolargolargolargolar_suffix text);
+SELECT * FROM cartodb._CDB_Unique_Column_Identifier(NULL, 'largolargolargolargolargolargolargolargolargolargolargolargolar', '_suffix', 'test'::regclass);
+DROP TABLE test;
+
+CREATE TABLE test (name text);
+-- Test unique identifier creation with normal length UTF8 colname
+SELECT * FROM cartodb._CDB_Unique_Column_Identifier(NULL, 'piraña', NULL, 'test'::regclass);
+
+-- Test unique identifier creation with prefix with normal length UTF8 colname
+SELECT * FROM cartodb._CDB_Unique_Column_Identifier('prefix_', 'piraña', NULL, 'test'::regclass);
+
+-- Test unique identifier creation with suffix with normal length UTF8 colname
+SELECT * FROM cartodb._CDB_Unique_Column_Identifier(NULL, 'piraña', '_suffix', 'test'::regclass);
+
+-- Test unique identifier creation with long length UTF8 colname
+SELECT * FROM cartodb._CDB_Unique_Column_Identifier(NULL, 'piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpiñaácidpin', NULL, 'test'::regclass);
+
+-- Test unique identifier creation with prefix with long length UTF8 colname
+SELECT * FROM cartodb._CDB_Unique_Column_Identifier('prefix_', 'piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpiñaácidpin', NULL, 'test'::regclass);
+DROP TABLE test;
+
+-- Test new identifier is found when name is taken from previous case
+CREATE TABLE test (prefix_piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpi text);
+SELECT * FROM cartodb._CDB_Unique_Column_Identifier('prefix_', 'piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpiñaácidpin', NULL, 'test'::regclass);
+DROP TABLE test;
+
+-- Test unique identifier creation with suffix with long length UTF8 colname
+CREATE TABLE test (name text);
+SELECT * FROM cartodb._CDB_Unique_Column_Identifier(NULL, 'piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpiñaácidpin', '_suffix', 'test'::regclass);
+DROP TABLE test;
+
+-- Test new identifier is found when name is taken from previous case
+CREATE TABLE test (piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpi_suffix text);
+SELECT * FROM cartodb._CDB_Unique_Column_Identifier(NULL, 'piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpiñaácidpin', '_suffix', 'test'::regclass);
+DROP TABLE test;
+
+-- Test _CDB_Octet_Truncate simple case
+SELECT * FROM cartodb._CDB_Octet_Truncate('piraña', 5);
+
+-- Test _CDB_Octet_Truncate UTF8 case
+SELECT * FROM cartodb._CDB_Octet_Truncate('piraña', 6);
+
+-- Test _CDB_Octet_Truncate UTF8 case
+SELECT * FROM cartodb._CDB_Octet_Truncate('piraña', 7);
--- a/test/CDB_HelperTest_expect
+++ b/test/CDB_HelperTest_expect
@ -0,0 +1,59 @@
+relname
+prefix_relname
+relname_suffix
+largolargolargolargolargolargolargolargolargolargolargolargo
+prefix_largolargolargolargolargolargolargolargolargolargolar
+CREATE TABLE
+prefix_largolargolargolargolargolargolargolargolargolargolar_0
+DROP TABLE
+largolargolargolargolargolargolargolargolargolargolar_suffix
+CREATE TABLE
+largolargolargolargolargolargolargolargolargolargolar_suffix_0
+DROP TABLE
+piraña
+prefix_piraña
+piraña_suffix
+piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpiñaácid
+prefix_piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpi
+CREATE TABLE
+prefix_piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpi_0
+DROP TABLE
+piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpi_suffix
+CREATE TABLE
+piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpi_suffix_0
+DROP TABLE
+CREATE TABLE
+colname
+prefix_colname
+colname_suffix
+largolargolargolargolargolargolargolargolargolargolargolargo
+prefix_largolargolargolargolargolargolargolargolargolargolar
+DROP TABLE
+CREATE TABLE
+prefix_largolargolargolargolargolargolargolargolargolargolar_0
+DROP TABLE
+CREATE TABLE
+largolargolargolargolargolargolargolargolargolargolar_suffix
+DROP TABLE
+CREATE TABLE
+largolargolargolargolargolargolargolargolargolargolar_suffix_0
+DROP TABLE
+CREATE TABLE
+piraña
+prefix_piraña
+piraña_suffix
+piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpiñaácid
+prefix_piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpi
+DROP TABLE
+CREATE TABLE
+prefix_piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpi_0
+DROP TABLE
+CREATE TABLE
+piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpi_suffix
+DROP TABLE
+CREATE TABLE
+piñaácidpiñaácidpiñaácidpiñaácidpiñaácidpi_suffix_0
+DROP TABLE
+pira
+pirañ
+piraña