SQL query results are combined by 100

sql2008 statement:

select name = case when num = 1 then a else''end, grades, subjects.

from (select name, num = row number() over(partition by name order by name desc) from table ) t

This will group the table by name to meet the requirements.

The listagg() function will group the query results of a column into a single line, and we can define the separators we want in each result set.

Example: select * from test;

select id,listagg(name,',') within group (order by name) from test group by id;

When the listagg() function merges the results, we can set the order of merging by and the conditions for merging by by group. Somewhat similar to sum() or count().

It's worth noting that the connectors of this function will only appear in the middle of the query results, and this merge will stack infinitely according to the grouping criteria. Therefore, when we need to insert the result into a row of the data table, please remember to add a length limit (usually we use the substr() function to remove the part that is too long) to avoid data loss, etc.

Try this.

create table new_table asselect * from a union allselect * from b

Here we talk about the relationship between union and union all, and the use of union is to merge two results.

However, union has a weight-ranking function, that is, there is a1 in the same two results, so that there is only one a1 after the query result; And Union All doesn't have a weight function.,There are two A1s in the final display result.

So you see which one you want to use.

With the left join you should be able to achieve what you require:

Set the table: T1 (columns C1, C2, C3), T2 (columns C1, C2), T3 (columns C1, C2, C3).

Result columns: C1, C2, C3, C4, C5, C6, C7, <>

Output merging can be done using the union all clause.

Two requirements for merging outputs:

1. Output the same number of query columns;

2. The data type and length of the corresponding column must be the same.

For example, select 1 as column1,'string list' as column2

union all

select 1,''

After the execution is executed, the output result is as follows:

However, I don't quite understand what you mean.

The idea is as follows:

1. SQL-SELECT with the HWO Type field as P0.

2. Record each item, go to the database to find whether there is a record of the same content in the field.

Logically, there are several possibilities.

1) t0-i0 (here there is no transition p0 in the middle) (2) p0-i0 (your 95035-p002 is this kind of data) (3) t0-p0-i0 (that is, your most general program, you want to remove the middle p0).

Then this is obviously three parts, if written separately.

Each part corresponds to the previous part, but it seems to be cumbersome, and when put together, I tried to write a statement (use case when, but there are too many contents in front of it, and each case when the words are easy to have problems, and there are too many places to test), I guess it can't be used even if it is written, so I only wrote this simplest one. (Calculated in chunks).