<!-- #请认真填写以下信息，否则可能由于无法定位，导致issue无法解决而被取消 -->

【标题描述】: 部分窗口函数预期结果发生变化

【测试类型：SQL功能】【测试版本：7.0.0-RC1】【CI】 部分窗口函数预期结果发生变化

【操作系统和硬件信息】(查询命令: cat /etc/system-release, uname -a): ARM+openEuler

【测试环境】(单机/1主x备x级联备): 一主两备

【被测功能】: 窗口函数

【测试类型】: 功能测试

【数据库版本】(查询命令: gaussdb -V):

【预置条件】: 无

【操作步骤】(请填写详细的操作步骤):

1. cume_dist窗口函数，查询结果浮点精度为0时，精度值不存在；

测试SQL如下：

to_timestamp('2019-01-04 16:33:47.123456','yyyy-mm-dd hh24:mi:ss.ffffff'),

<!-- #请认真填写以下信息，否则可能由于无法定位，导致issue无法解决而被取消 -->

【标题描述】: 部分窗口函数预期结果发生变化

【测试类型：SQL功能】【测试版本：7.0.0-RC1】【CI】 部分窗口函数预期结果发生变化

【操作系统和硬件信息】(查询命令: cat /etc/system-release, uname -a): ARM+openEuler

【测试环境】(单机/1主x备x级联备): 一主两备

【被测功能】: 窗口函数

【测试类型】: 功能测试

【数据库版本】(查询命令: gaussdb -V):

【预置条件】: 无

【操作步骤】(请填写详细的操作步骤):

1. cume_dist窗口函数，查询结果浮点精度为0时，精度值不存在；

测试SQL如下：

```

create table t_ustore_function_cumedist_0054(

col_1 integer,col_2 bigint, col_3 float8,col_4 decimal(12,6),

col_5 bool,col_6 char(30),col_7 varchar2(50),col_8 varchar(30),

col_9 interval day to second,col_10 timestamp,col_11 date,

col_12 date,col_13 timestamp without time zone, col_14 blob,

col_15 clob,col_16 int[]) with (storage_type=ustore);

drop sequence if exists s_ustore_function_cumedist_0054;

create sequence s_ustore_function_cumedist_0054 increment by 1 start with 10;

begin

for i in 1..50 loop

insert into t_ustore_function_cumedist_0054 values

(i,s_ustore_function_cumedist_0054.nextval,i+445.255,

98*0.99*i,true,lpad('abc','30','@'),lpad('abc','30','b'),

rpad('abc','30','e'),

(interval '4 5:12:10.222' day to second(3)),

to_timestamp('2019-01-03 14:58:54.000000','yyyy-mm-dd hh24:mi:ss.ffffff'),

to_timestamp('2019-01-04 16:33:47.123456','yyyy-mm-dd hh24:mi:ss.ffffff'),

to_timestamp('2019-01-04 16:33:47.123456','yyyy-mm-dd hh24:mi:ss.ffffff'),

to_timestamp('2019-01-04 16:33:47.123456','yyyy-mm-dd hh24:mi:ss.ffffff'),

lpad('10','12','01010')::blob,rpad('abc','9','a@123&^%djgk')::clob,

'{32,535,5645645,6767,76,67,56,48,979,978,7}');

end loop;

end;

/

select col_3,col_4,cume_dist() over(partition by col_1,col_2 order by col_1 nulls first,col_2 nulls last)

from t_ustore_function_cumedist_0054

group by col_1,col_2,col_3,col_4 order by 1,2,3 limit 5;

```

2. dense_rank窗口函数，存在入参时函数报错，当前版本执行成功；

测试SQL如下：

```

create table t_ustore_function_denserank_0015(id int,name char(30),score decimal(6,2), course int,examedate date,gender bool) with (storage_type=ustore);

insert into t_ustore_function_denserank_0015 values(5,'elic', 70,1,to_date('2019-12-17', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(4,'dock', 100,1,to_date('2019-12-17', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(3,'clark', 80,1,to_date('2019-12-17', 'yyyy-mm-dd'),'false');

insert into t_ustore_function_denserank_0015 values(2,'bob', 90,1,to_date('2019-12-17', 'yyyy-mm-dd'),'false');

insert into t_ustore_function_denserank_0015 values(1,'alice',60,1,to_date('2019-12-17', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(10,'jacky',80,2,to_date('2019-12-18', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(9,'iris',60,2,to_date('2019-12-18', 'yyyy-mm-dd'),'false');

insert into t_ustore_function_denserank_0015 values(8,'hill',70,2,to_date('2019-12-18', 'yyyy-mm-dd'),'false');

insert into t_ustore_function_denserank_0015 values(7,'grace',50,2,to_date('2019-12-18', 'yyyy-mm-dd'),'false');

insert into t_ustore_function_denserank_0015 values(6,'frank',70,2,to_date('2019-12-18', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(11,'jak',100,2,to_date('2019-12-18', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(12,'merry',null,1,to_date('2019-12-18', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(13,'tom',null,2,to_date('2019-12-18', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(14,'tomy',null,1,to_date('2019-12-18', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(15,'tomyyy',null,1,to_date('2019-12-18', 'yyyy-mm-dd'),'false');

select dense_rank(2) within group (order by score) as dense_rank

from t_ustore_function_denserank_0015;

```

<!-- #请认真填写以下信息，否则可能由于无法定位，导致issue无法解决而被取消 -->

【标题描述】: 部分窗口函数预期结果发生变化

【测试类型：SQL功能】【测试版本：7.0.0-RC1】【CI】 部分窗口函数预期结果发生变化

【操作系统和硬件信息】(查询命令: cat /etc/system-release, uname -a): ARM+openEuler

【测试环境】(单机/1主x备x级联备): 一主两备

【被测功能】: 窗口函数

【测试类型】: 功能测试

【数据库版本】(查询命令: gaussdb -V):

【预置条件】: 无

【操作步骤】(请填写详细的操作步骤):

1. cume_dist窗口函数，查询结果浮点精度为0时，精度值不存在；

测试SQL如下：

```

create table t_ustore_function_cumedist_0054(

col_1 integer,col_2 bigint, col_3 float8,col_4 decimal(12,6),

col_5 bool,col_6 char(30),col_7 varchar2(50),col_8 varchar(30),

col_9 interval day to second,col_10 timestamp,col_11 date,

col_12 date,col_13 timestamp without time zone, col_14 blob,

col_15 clob,col_16 int[]) with (storage_type=ustore);

drop sequence if exists s_ustore_function_cumedist_0054;

create sequence s_ustore_function_cumedist_0054 increment by 1 start with 10;

begin

for i in 1..50 loop

insert into t_ustore_function_cumedist_0054 values

(i,s_ustore_function_cumedist_0054.nextval,i+445.255,

98*0.99*i,true,lpad('abc','30','@'),lpad('abc','30','b'),

rpad('abc','30','e'),

(interval '4 5:12:10.222' day to second(3)),

to_timestamp('2019-01-03 14:58:54.000000','yyyy-mm-dd hh24:mi:ss.ffffff'),

to_timestamp('2019-01-04 16:33:47.123456','yyyy-mm-dd hh24:mi:ss.ffffff'),

to_timestamp('2019-01-04 16:33:47.123456','yyyy-mm-dd hh24:mi:ss.ffffff'),

to_timestamp('2019-01-04 16:33:47.123456','yyyy-mm-dd hh24:mi:ss.ffffff'),

lpad('10','12','01010')::blob,rpad('abc','9','a@123&^%djgk')::clob,

'{32,535,5645645,6767,76,67,56,48,979,978,7}');

end loop;

end;

/

select col_3,col_4,cume_dist() over(partition by col_1,col_2 order by col_1 nulls first,col_2 nulls last)

from t_ustore_function_cumedist_0054

group by col_1,col_2,col_3,col_4 order by 1,2,3 limit 5;

```

2. dense_rank窗口函数，存在入参时函数报错，当前版本执行成功；

测试SQL如下：

```

create table t_ustore_function_denserank_0015(id int,name char(30),score decimal(6,2), course int,examedate date,gender bool) with (storage_type=ustore);

insert into t_ustore_function_denserank_0015 values(5,'elic', 70,1,to_date('2019-12-17', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(4,'dock', 100,1,to_date('2019-12-17', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(3,'clark', 80,1,to_date('2019-12-17', 'yyyy-mm-dd'),'false');

insert into t_ustore_function_denserank_0015 values(2,'bob', 90,1,to_date('2019-12-17', 'yyyy-mm-dd'),'false');

insert into t_ustore_function_denserank_0015 values(1,'alice',60,1,to_date('2019-12-17', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(10,'jacky',80,2,to_date('2019-12-18', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(9,'iris',60,2,to_date('2019-12-18', 'yyyy-mm-dd'),'false');

insert into t_ustore_function_denserank_0015 values(8,'hill',70,2,to_date('2019-12-18', 'yyyy-mm-dd'),'false');

insert into t_ustore_function_denserank_0015 values(7,'grace',50,2,to_date('2019-12-18', 'yyyy-mm-dd'),'false');

insert into t_ustore_function_denserank_0015 values(6,'frank',70,2,to_date('2019-12-18', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(11,'jak',100,2,to_date('2019-12-18', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(12,'merry',null,1,to_date('2019-12-18', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(13,'tom',null,2,to_date('2019-12-18', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(14,'tomy',null,1,to_date('2019-12-18', 'yyyy-mm-dd'),'true');

insert into t_ustore_function_denserank_0015 values(15,'tomyyy',null,1,to_date('2019-12-18', 'yyyy-mm-dd'),'false');

select dense_rank(2) within group (order by score) as dense_rank

from t_ustore_function_denserank_0015;

```

测试SQL如下：

```

create table t_percent_rank_001(id1 int,id2 int,id3 smallint,id4 int,num1 float,char1 varchar(500)) with (storage_type=ustore);

insert into t_percent_rank_001 values ( 1 , 4 , 4 , 4 , 4001.87754238 , ' 010012228 ' );

select id1,id3,percent_rank(null) over(partition by id1 order by id3 desc nulls last) f1 from t_percent_rank_003;

```

【实际输出】: 执行结果与原本预期不符

【原因分析】:

1. 这个问题的根因

2. 问题推断过程

3. 还有哪些原因可能造成类似现象

4. 该问题是否有临时规避措施

5. 问题解决方案

6. 预计修复问题时间

【日志信息】(请附上日志文件、截图、coredump信息):




【测试代码】: 不涉及