文档首页> 数据仓库服务 GaussDB(DWS)> 工具指南（巴黎区域）> DSC SQL语法迁移工具> Teradata语法迁移> 查询迁移操作符
更新时间：2023-03-17 GMT+08:00
查看PDF

查询迁移操作符

本节主要介绍Teradata查询迁移操作符的迁移语法。迁移语法决定了关键字/特性的迁移方式。

具体详见以下节点内容：

QUALIFY

ALIAS

FORMAT和CAST

缩写关键字迁移

以$开头的对象名称

QUALIFY

通常，QUALIFY子句和CSUM()、MDIFF()、ROW_NUMBER()、RANK()等分析函数（窗口函数）一同使用。子查询中会包含QUALIFY子句指定的窗口函数。迁移工具支持QUALIFY使用MDIFF()、RANK()和ROW_NUMBER()函数。QUALIFY是Teradata扩展项，不是标准ANSI语法。QUALIFY在WHERE和GROUP BY子句后执行，必须单独成行。

只有当SELECT语句显式包含列名和/或表达式时，DSC才允许在ORDER BY子句中指定该列名和/或表达式。

输入：QUALIFY

1
2
3
4
5
6
 
SELECT
        CUSTOMER_ID
       ,CUSTOMER_NAME
  FROM
       CUSTOMER_T QUALIFY row_number( ) Over( partition BY CUSTOMER_ID ORDER BY POSTAL_CODE DESC ) = 1
;

输出

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
 
SELECT
          CUSTOMER_ID
          ,CUSTOMER_NAME
     FROM
          (
               SELECT
                         CUSTOMER_ID
                         ,CUSTOMER_NAME
                         ,row_number( ) Over( partition BY CUSTOMER_ID ORDER BY POSTAL_CODE DESC ) AS ROW_NUM1
                    FROM
                         CUSTOMER_T
          ) Q1
     WHERE
          Q1.ROW_NUM1 = 1
;

输入：QUALIFY，使用MDIFF和RANK

1
2
3
4
5
6
7
8
9
 
SELECT
           material_name
          ,unit_of_measure * standard_cost AS tot_cost
     FROM
          raw_material_t m LEFT JOIN supplies_t s
               ON s.material_id = m.material_id 
          QUALIFY rank ( ) over( ORDER BY tot_cost DESC ) IN '5'
                  OR mdiff( tot_cost ,3 ,material_name ) IS NULL
;

输出

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
 
SELECT
          material_name
          ,tot_cost
     FROM
          (
               SELECT
                         material_name
                         ,unit_of_measure * standard_cost AS tot_cost
                         ,rank ( ) over( ORDER BY unit_of_measure * standard_cost DESC ) AS ROW_NUM1
                         ,unit_of_measure * standard_cost - (LAG( unit_of_measure * standard_cost ,3 ,NULL ) over( ORDER BY material_name )) AS ROW_NUM2
                    FROM
                         raw_material_t m LEFT JOIN supplies_t s
                              ON s.material_id = m.material_id
          ) Q1
     WHERE
          Q1.ROW_NUM1 = '5'
          OR Q1.ROW_NUM2 IS NULL
;

输入：QUALIFY，使用ORDER BY且ORDER BY中包含不存在于SELECT列表的列

1
2
3
4
5
6
 
SELECT Postal_Code
   FROM db_pvfc9_std.Customer_t t1
   GROUP BY Customer_Name ,Postal_Code
   QUALIFY ---comments
 (  Rank ( CHAR(Customer_Address) DESC  )  ) = 1
  ORDER BY t1.Customer_Name;

输出

1
2
3
4
5
6
7
8
 
SELECT Postal_Code FROM
           ( SELECT Customer_Name, Postal_Code
    , Rank () over( PARTITION BY Customer_Name, Postal_Code ORDER BY LENGTH(Customer_Address) DESC ) AS Rank_col
               FROM db_pvfc9_std.Customer_t t1
           ) Q1
      WHERE /*comments*/
    Q1.Rank_col = 1
  ORDER BY Q1.Customer_Name;

输入：QUALIFY，使用列别名且不应在SELECT列表中再次添加相应的列表达式

1
2
3
4
5
 
SELECT material_name, unit_of_measure * standard_cost as tot_cost,
        RANK() over(order by tot_cost desc) vendor_cnt
 FROM raw_material_t m left join supplies_t s
 ON s.material_id = m.material_id
 QUALIFY vendor_cnt < 5 or MDIFF(tot_cost, 3, material_name) IS NULL;

输出

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
 
SELECT material_name, tot_cost, vendor_cnt
   FROM  ( SELECT material_name
                       , unit_of_measure * standard_cost AS tot_cost
                      , rank () over (ORDER BY tot_cost DESC) vendor_cnt
                      , tot_cost - ( LAG(tot_cost ,3 ,NULL) over (ORDER BY material_name) ) AS anltfn
               FROM raw_material_t m LEFT JOIN supplies_t s
                  ON s.material_id = m.material_id
           ) Q1
      WHERE  Q1.vendor_cnt < 5 OR Q1.anltfn IS NULL
 ;

TITLE和QUALIFY

输入

REPLACE VIEW ${STG_VIEW}.LP06_BMCLIINFP${v_Table_Suffix_Inc}
(
  CLICLINBR
, CLICHNNAM
, CLICHNSHO
, CLICLIMNE
, CLIBNKCOD
)
AS
LOCKING ${STG_DATA}.LP06_BMCLIINFP${v_Table_Suffix_Inc} FOR ACCESS
SELECT
  CLICLINBR (title '    VARCHAR(20)')
, CLICHNNAM (title '        VARCHAR(200)')
, CLICHNSHO (title '        VARCHAR(20)')
, CLICLIMNE (title '      VARCHAR(10)')
, CLIBNKCOD (title '       VARCHAR(11)')
FROM
  ${STG_DATA}.LP06_BMCLIINFP${v_Table_Suffix_Inc} s1
QUALIFY
    ROW_NUMBER() OVER(PARTITION BY  CLICLINBR ORDER BY CLICLINBR  ) = 1
;

输出

CREATE OR REPLACE VIEW ${STG_VIEW}.LP06_BMCLIINFP${v_Table_Suffix_Inc}
(
  CLICLINBR
, CLICHNNAM
, CLICHNSHO
, CLICLIMNE
, CLIBNKCOD
)
AS
/* LOCKING ${STG_DATA}.LP06_BMCLIINFP${v_Table_Suffix_Inc} FOR ACCESS */
SELECT CLICLINBR
        , CLICHNNAM
       , CLICHNSHO
       , CLICLIMNE
      , CLIBNKCOD
 FROM (
           SELECT
                          CLICLINBR /* (title '    VARCHAR(20)') */
                        , CLICHNNAM /* (title '        VARCHAR(200)') */
                       , CLICHNSHO /* (title '        VARCHAR(20)') */
                       , CLICLIMNE /* (title '      VARCHAR(10)') */
                       , CLIBNKCOD /* (title '       VARCHAR(11)') */
                       , ROW_NUMBER() OVER(PARTITION BY  CLICLINBR ORDER BY CLICLINBR  ) AS ROWNUM1
FROM
  ${STG_DATA}.LP06_BMCLIINFP${v_Table_Suffix_Inc} s1 ) Q1
WHERE Q1.ROWNUM1 = 1
;

ALIAS

所有数据库都支持ALIAS。在Teradata中，定义ALIAS的语句允许其SELECT和WHERE子句引用ALIAS。但是目标数据库的SELECT和WHERE语句不支持ALIAS，因此MT将其迁移为实际的字段名称替换。

比较操作符LT、LE、GT、GE、EQ和NE不得用作表别名或列别名。

工具支持列的ALIAS名称。如果ALIAS名称与列名称相同，则仅为该列而非表中其他列指定ALIAS。在以下示例中，DATA_DT列名称与DATA_DT别名之间存在冲突，工具不支持。

1
 
SELECT DATA_DT,DATA_INT AS DATA_DT FROM KK WHERE DATA_DT=DATE;

输入：ALIAS

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
 
SELECT
          expression1 (
               TITLE 'Expression 1'
          ) AS alias1
          ,CASE
               WHEN alias1 + Cx >= z
               THEN 1
               ELSE 0
          END AS alias2
     FROM
          tab1
     WHERE
          alias1 = y
;

输出：tdMigrateALIAS = FALSE

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
 
SELECT
          expression1 AS alias1
          ,CASE
               WHEN alias1 + Cx >= z
               THEN 1
               ELSE 0
          END AS alias2
     FROM
          tab1
     WHERE
          alias1 = y
;

输出：tdMigrateALIAS = TRUE

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
 
SELECT
          expression1 AS alias1
          ,CASE
               WHEN expression1 + Cx >= z
               THEN 1
               ELSE 0
          END AS alias2
     FROM
          tab1
     WHERE
          expression1 = y
;

FORMAT和CAST

Teradata中，关键词FORMAT用于格式化列或表达式。例如，LPAD中FORMAT '9(n)'和'z(n)'分别用'0'和空格(' ')表示。

数据类型转换可使用CAST或直接数据类型（[like (expression1)(CHAR(n))]）进行。该功能使用CAST实现。详情参见类型转换和格式化。

输入：FORMAT和CAST 
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
 
SELECT
          CAST(TRIM( Agt_Num ) AS DECIMAL( 5 ,0 ) FORMAT '9(5)' )
     FROM
          C03_AGENT_BOND
;

SELECT
          CAST(CAST( Agt_Num AS INT FORMAT 'Z(17)' ) AS CHAR( 5 ) )
     FROM         
          C03_AGENT_BOND
;

SELECT
          CHAR(CAST( CAST( CND_VLU AS DECIMAL( 17 ,0 ) FORMAT 'Z(17)' ) AS VARCHAR( 17 ) ) )
     FROM
          C03_AGENT_BOND
;

输出

SELECT
          LPAD( CAST( TRIM( Agt_Num ) AS DECIMAL( 5 ,0 ) ) ,5 ,'0' ) AS Agt_Num
     FROM
          C03_AGENT_BOND
;
SELECT
CAST(CAST( Agt_Num AS INT FORMAT 'Z(17)' ) AS CHAR( 5 ) )
FROM
C03_AGENT_BOND
;

SELECT
          LENGTH( CAST( LPAD( CAST( CND_VLU AS DECIMAL( 17 ,0 ) ) ,17 ,' ' ) AS VARCHAR( 17 ) ) ) AS CND_VLU
     FROM
          C03_AGENT_BOND
;

输入：FORMAT 'Z(n)9'

1
2
3
4
5
6
 
SELECT
          standard_price (FORMAT 'Z(5)9') (CHAR( 6 ))
          ,max_price (FORMAT 'ZZZZZ9') (CHAR( 6 ))
     FROM
          product_t
;

输出

1
2
3
4
5
6
 
SELECT
          CAST( TO_CHAR( standard_price ,'999990' ) AS CHAR( 6 ) ) AS standard_price
          ,CAST( TO_CHAR( max_price ,'999990' ) AS CHAR( 6 ) ) AS max_price
     FROM
          product_t
;

输入：FORMAT 'z(m)9.9(n)'

1
2
3
4
5
 
SELECT
          standard_price (FORMAT 'Z(6)9.9(2)') (CHAR( 6 ))
     FROM
          product_t
;

输出

1
2
3
4
5
 
SELECT
          CAST( TO_CHAR( standard_price ,'9999990.00' ) AS CHAR( 6 ) ) AS standard_price
     FROM
          product_t
;

输入：CAST AS INTEGER

1
2
3
4
5
 
SELECT
          CAST( standard_price AS INTEGER )
     FROM
          product_t
;

输出

1
2
3
4
5
 
SELECT
         (standard_price)
     FROM
          product_t
;

输入：CAST AS INTEGER FORMAT

1
2
3
4
5
6
7
 
SELECT
          CAST( price11 AS INTEGER FORMAT 'Z(4)9' ) (
               CHAR( 10 )
          )
     FROM
          product_t
;

输出

1
2
3
4
5
 
SELECT
          CAST( TO_CHAR(  ( price11 ) ,'99990' ) AS CHAR( 10 ) ) AS price11
     FROM
          product_t
;

新增以下Gauss函数来转换为INTEGER：

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
 
CREATE OR REPLACE FUNCTION 
/*  This function is used to support "CAST AS INTEGER" of Teradata.
    It should be created in the "mig_td_ext" schema.
*/
     ( i_param                            TEXT )
RETURN INTEGER
AS
  v_castasint    INTEGER;
BEGIN

   v_castasint := CASE WHEN i_param IS NULL 
                                                THEN NULL         -- if NULL value is provided as input
                                                                WHEN TRIM(i_param) IS NULL 
                                                                                THEN 0                  -- if empty string with one or more spaces is provided 
                                                                ELSE TRUNC(CAST(i_param AS NUMBER))            -- if any numeric value is provided
                                END;

RETURN v_castasint;
END;

缩写关键字迁移

表1列出了Teradata支持的缩写关键字及其语法在GaussDB(DWS)中对应的语法。

表1 缩写关键字列表

Teradata语法

对应的GaussDB(DWS)语法

SEL

SELECT

INS

INSERT

UPD

UPDATE

DEL

DELETE

CT

CREATE TABLE

CV

CREATE VIEW

BT

START TRANSACTION

ET

COMMIT

输入：BT

BT;
--              
delete from ${BRTL_DCOR}.BRTL_CS_CUST_CID_UID_REL 
where DW_Job_Seq = ${v_Group_No};

.if ERRORCODE <> 0 then .quit 12;

--     
insert into ${BRTL_DCOR}.BRTL_CS_CUST_CID_UID_REL 
(
   Cust_Id
  ,Cust_UID
  ,DW_Upd_Dt
  ,DW_Upd_Tm
  ,DW_Job_Seq
  ,DW_Etl_Dt
)
select
   a.Cust_Id
  ,a.Cust_UID
  ,current_date as Dw_Upd_Dt
  ,current_time(0) as DW_Upd_Tm
  ,cast(${v_Group_No} as byteint) as DW_Job_Seq
  ,cast('${v_Trx_Dt}' as date format 'yyyy-mm-dd') as DW_Etl_Dt
from ${BRTL_VCOR}.BRTL_CS_CUST_CID_UID_REL_S a
where a.DW_Snsh_Dt = cast('${v_Trx_Dt}' as date format 'yyyy-mm-dd');

.if ERRORCODE <> 0 then .quit 12;

ET;cd ..

输出

BEGIN
--    
   BEGIN          
      delete from ${BRTL_DCOR}.BRTL_CS_CUST_CID_UID_REL  
       where DW_Job_Seq = ${v_Group_No};
         lv_mig_errorcode = 0;
   EXCEPTION
      WHEN OTHERS THEN
         lv_mig_errorcode = -1;
   END;

   IF lv_mig_errorcode <> 0 THEN
        RAISE EXCEPTION '12';
   END IF;

--     
   BEGIN
        insert into ${BRTL_DCOR}.BRTL_CS_CUST_CID_UID_REL 
        (
          Cust_Id
         ,Cust_UID
         ,DW_Upd_Dt
         ,DW_Upd_Tm
         ,DW_Job_Seq
         ,DW_Etl_Dt
       )
      select
          a.Cust_Id
         ,a.Cust_UID
         ,current_date as Dw_Upd_Dt
         ,current_time(0) as DW_Upd_Tm
         ,cast(${v_Group_No} as byteint) as DW_Job_Seq
         ,cast('${v_Trx_Dt}' as date format 'yyyy-mm-dd') as DW_Etl_Dt
       from ${BRTL_VCOR}.BRTL_CS_CUST_CID_UID_REL_S a
     where a.DW_Snsh_Dt = cast('${v_Trx_Dt}' as date format 'yyyy-mm-dd');
   EXCEPTION
      WHEN OTHERS THEN
         lv_mig_errorcode = -1;
   END;

   IF lv_mig_errorcode <> 0 THEN
        RAISE EXCEPTION '12';
   END IF;

END;

以$开头的对象名称

本节介绍如何迁移以$（美元符号）开头的对象名称。

下表具体描述了这些对象名称的迁移行为。这些行为可以通过tdMigrateDollar参数来设置。

详情请参见IN/NOT IN转换。

表2 以$开头的对象名称的迁移行为

tdMigrateDollar设置

对象名称

迁移为

true

$V_SQL

静态对象名称

"$V_SQL"

true

${V_SQL}

动态对象名称

${V_SQL}

无变化：不支持动态对象名称

false

$V_SQL

静态对象名称

$V_SQL

无变化：参数设为false

false

${V_SQL}

动态对象名称

${V_SQL}

无变化：参数设为false

任何以$开头的变量都被视为值。工具将通过添加ARRAY来进行迁移。

输入：以$开头的对象 
1
 
SELECT $C1 from p11 where $C1 NOT LIKE ANY ($sql1);

输出（tdMigrateDollar = TRUE）

1
2
3
4
5
6
7
8
9
 
SELECT
        "$C1"
FROM
        p11
WHERE
        "$C1" NOT LIKE ANY (
        ARRAY[ "$sql1" ]
        )
;

输出（tdMigrateDollar = FALSE）

1
2
3
4
5
6
7
8
9
 
SELECT
        $C1
FROM
        p11
WHERE
        $C1 NOT LIKE ANY (
        ARRAY[ $sql1 ]
        )
;
输入：LIKE ALL/LIKE ANY中以$开头的值 
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
 
SELECT * FROM T1
WHERE T1.Event_Dt>=ADD_MONTHS(CAST('${OUT_DATE}' AS DATE FORMAT 'YYYYMMDD')+1,(-1)*CAST(T7.Tm_Range_Month AS INTEGER)) 
   AND T1.Event_Dt<=CAST('${OUT_DATE}' AS DATE FORMAT 'YYYYMMDD')
   AND T1.Cntpty_Acct_Name NOT LIKE ALL ( SELECT Tx_Cntpty_Name_Key FROM TEMP_NAME )
   AND T1.Cntpty_Acct_Name NOT LIKE ANY ( SELECT Tx_Cntpty_Name_Key FROM TEMP_NAME )
   AND T1.Cntpty_Acct_Name LIKE ALL ( SELECT Tx_Cntpty_Name_Key FROM TEMP_NAME )
   AND T1.Cntpty_Acct_Name LIKE ANY ( SELECT Tx_Cntpty_Name_Key FROM TEMP_NAME )
   AND T1.Col1 NOT LIKE ANY ($sql1)
   AND T1.Col1 NOT LIKE ALL ($sql1)
   AND T1.Col1 LIKE ANY ($sql1)
   AND T1.Col1 LIKE ALL ($sql1);

输出

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
 
SELECT
          *
     FROM
          T1
     WHERE
          T1.Event_Dt >= ADD_MONTHS (CAST( '${OUT_DATE}' AS DATE ) + 1 ,(- 1 ) * CAST( T7.Tm_Range_Month AS INTEGER ))
          AND T1.Event_Dt <= CAST( '${OUT_DATE}' AS DATE )
          AND T1.Cntpty_Acct_Name NOT LIKE ALL (
               SELECT
                         Tx_Cntpty_Name_Key
                    FROM
                         TEMP_NAME
          )
          AND T1.Cntpty_Acct_Name NOT LIKE ANY (
               SELECT
                         Tx_Cntpty_Name_Key
                    FROM
                         TEMP_NAME
          )
          AND T1.Cntpty_Acct_Name LIKE ALL (
               SELECT
                         Tx_Cntpty_Name_Key
                    FROM
                         TEMP_NAME
          )
          AND T1.Cntpty_Acct_Name LIKE ANY (
               SELECT
                         Tx_Cntpty_Name_Key
                    FROM
                         TEMP_NAME
          )
          AND T1.Col1 NOT LIKE ANY (
               ARRAY[ "$sql1" ]
          )
          AND T1.Col1 NOT LIKE ALL (
               ARRAY[ "$sql1" ]
          )
          AND T1.Col1 LIKE ANY (
               ARRAY[ "$sql1" ]
          )
          AND T1.Col1 LIKE ALL (
               ARRAY[ "$sql1" ]
          )
;

QUALIFY、CASE和ORDER BY

输入

select 
   a.Cust_UID as Cust_UID          /*   UID */
  ,a.Rtl_Usr_Id as Ini_CM          /*        */
  ,a.Cntr_Aprv_Dt as Aprv_Pass_Tm          /*        */
  ,a.Blg_Org_Id as CM_BRN_Nbr          /*         */
  ,a.Mng_Chg_Typ_Cd  as MNG_CHG_TYP_CD          /*          */
  ,case when a.Blg_Org_Id = b.BRN_Org_Id and a.Mng_Chg_Typ_Cd= 'PMD' and a.Pst_Id in ('PB0101','PB0104') then 'Y' ----        ,        
       when a.Blg_Org_Id = b.BRN_Org_Id and a.Mng_Chg_Typ_Cd= 'DEVPMD' and a.Pst_Id ='PB0106'  then 'Y' ----            
       when a.Blg_Org_Id = b.BRN_Org_Id and a.Mng_Chg_Typ_Cd= 'DMD' and a.Pst_Id in ('PB0201','PB0204') then 'Y' ----        ,        
      when a.Blg_Org_Id = b.BRN_Org_Id and a.Mng_Chg_Typ_Cd= 'DEVDMD' and a.Pst_Id ='PB0109'  then 'Y' ----            ,
       else ''
end  as Pst_Flg          /*      */
  ,a.Pst_Id as Pst_Id          /*      */
  ,a.BBK_Org_Id  as BBK_Org_Id          /*        */
from VT_CUID_MND_NMN_CHG_INF as a          /* VT_         */
LEFT OUTER JOIN ${BRTL_VCOR}.BRTL_EM_USR_PST_REL_INF_S as b          /* EM_           */
  on  a.Rtl_Usr_Id = b.Rtl_Usr_Id          
  AND a.Blg_Org_Id  = b.BRN_Org_Id           
  AND a.Pst_Id = b.Pst_Id          
  AND b.Sys_Id = 'privatebanking'           
  AND b.pst_sts IN ('1','0','-2') /*     1   -2   0  */          
  AND b.DW_Snsh_Dt =  cast('${v_Trx_Dt}' as date format 'yyyy-mm-dd')          
qualify   row_number() over(partition by a.Cust_UID,a.bbk_org_id order by 
case when ( a.Mng_Chg_Typ_Cd= 'PMD'    and   a.Pst_Id in ('PB0101','PB0104')) or ( a.Mng_Chg_Typ_Cd= 'DEVPMD'    and   a.Pst_Id ='PB0106')   
then 0 when  (a.Mng_Chg_Typ_Cd= 'DMD' and a.Pst_Id in ('PB0201','PB0204')) or    (a.Mng_Chg_Typ_Cd= 'DEVDMD' and a.Pst_Id ='PB0109 ')   then 0 else 1 end  asc )  = 1         
;

输出

 SELECT
                     Cust_UID AS Cust_UID /*   UID */
                     ,Ini_CM /*        */
                     ,Aprv_Pass_Tm /*        */
                     ,CM_BRN_Nbr /*         */
                     ,MNG_CHG_TYP_CD /*          */
                     ,Pst_Flg /*      */
                     ,Pst_Id AS Pst_Id /*      */
                     ,BBK_Org_Id AS BBK_Org_Id /*        */
                FROM
                     ( SELECT
                          a.Cust_UID AS Cust_UID /*   UID */
                          ,a.Rtl_Usr_Id AS Ini_CM /*        */
                          ,a.Cntr_Aprv_Dt AS Aprv_Pass_Tm /*        */
                          ,a.Blg_Org_Id AS CM_BRN_Nbr /*         */
                          ,a.Mng_Chg_Typ_Cd AS MNG_CHG_TYP_CD /*          */
                          ,CASE WHEN a.Blg_Org_Id = b.BRN_Org_Id AND a.Mng_Chg_Typ_Cd = 'PMD' AND a.Pst_Id IN ( 'PB0101' ,'PB0104' )
                                    THEN 'Y' /*         , */
                               WHEN a.Blg_Org_Id = b.BRN_Org_Id AND a.Mng_Chg_Typ_Cd = 'DEVPMD' AND a.Pst_Id = 'PB0106'
                                    THEN 'Y' /*  */
                               WHEN a.Blg_Org_Id = b.BRN_Org_Id AND a.Mng_Chg_Typ_Cd = 'DMD' AND a.Pst_Id IN ( 'PB0201' ,'PB0204' )
                                    THEN 'Y' /*         , */
                               WHEN a.Blg_Org_Id = b.BRN_Org_Id AND a.Mng_Chg_Typ_Cd = 'DEVDMD' AND a.Pst_Id = 'PB0109'
                                    THEN 'Y' /*             , */
                          ELSE
                               ''
                          END AS Pst_Flg /*      */
                          ,a.Pst_Id AS Pst_Id /*      */
                          ,a.BBK_Org_Id AS BBK_Org_Id /*        */
                          ,row_number( ) over( partition BY a.Cust_UID
                          ,a.bbk_org_id
                     ORDER BY
                          CASE WHEN( a.Mng_Chg_Typ_Cd = 'PMD' AND Q1.Pst_Id IN ( 'PB0101' ,'PB0104' ) ) OR( Q1.Mng_Chg_Typ_Cd = 'DEVPMD' AND a.Pst_Id = 'PB0106' )
                                    THEN 0
                               WHEN( a.Mng_Chg_Typ_Cd = 'DMD' AND Q1.Pst_Id IN ( 'PB0201' ,'PB0204' ) ) OR( Q1.Mng_Chg_Typ_Cd = 'DEVDMD' AND a.Pst_Id = 'PB0109 ' )
                                    THEN 0
                          ELSE
                               1
                          END ASC ) AS ROW_NUM1
                     FROM
                          VT_CUID_MND_NMN_CHG_INF AS a /* VT_         */
                          LEFT OUTER JOIN BRTL_VCOR.BRTL_EM_USR_PST_REL_INF_S AS b /* EM_           */
                               ON a.Rtl_Usr_Id = b.Rtl_Usr_Id
                          AND a.Blg_Org_Id = b.BRN_Org_Id
                          AND a.Pst_Id = b.Pst_Id
                          AND b.Sys_Id = 'privatebanking'
                          AND b.pst_sts IN ( '1' ,'0' ,'-2' ) /*     1   -2   0  */
                          AND b.DW_Snsh_Dt = CAST( lv_mig_v_Trx_Dt AS DATE ) ) Q1
                     WHERE
                          Q1.ROW_NUM1 = 1 ;

父主题： Teradata语法迁移