I have table with certain number of columns.
I want to populate other table with the data of a particular column of Table1 as columns of table2 dynamically.
When I say dynamically I mean to say that when ever any data is added to the column of Table1 the table2 is populated with as many number of columns.
Changing the schema on the fly really isn't a good idea, for a number of reasons. From what you've described, I think you would be better off using a view for this. A view will give you the dynamic capabilities you're looking for with fewer side effects.
See this article:
How to create a view in SQL Server
I will once again repeat the disclaimer that this is a bad idea, many things can go wrong, and I'm certain there is a better solution to whatever underlying problem you're trying to solve. That said, to answer the explicit question anyway, here is an example of how to do this:
USE tempdb;
GO
CREATE TABLE dbo.Table1(Description VARCHAR(32));
CREATE TABLE dbo.Table2(ID INT);
GO
CREATE TRIGGER dbo.CatchNewTable1Data
ON dbo.Table1
FOR INSERT
AS
BEGIN
SET NOCOUNT ON;
DECLARE #sql NVARCHAR(MAX) = N'';
SELECT #sql += CHAR(13) + CHAR(10) +
'ALTER TABLE dbo.Table2 ADD '
+ QUOTENAME(d) + ' VARCHAR(255);' -- guessing on destination data type
FROM
(
SELECT DISTINCT d = LEFT([Description], 128) -- identifier <= 128
FROM inserted AS i
WHERE NOT EXISTS
(
SELECT 1 FROM sys.columns
WHERE name = LEFT(i.[Description], 128)
AND [object_id] = OBJECT_ID('dbo.Table2')
)
) AS x;
EXEC sp_executesql #sql;
END
GO
Now, let's try it out! Try a column that already exists, a multi-row insert where one of the columns already exists, a multi-row insert with dupes, etc. I am not posting a value > 255 nor am I dealing with any fancy characters that will cause a problem. Why? Because ultimately I don't want you to use this solution, I want to solve the real problem. But for the googlers I want to show that there is a solution to the stated problem.
-- does nothing:
INSERT dbo.Table1 SELECT 'ID';
-- only adds column 'foo':
INSERT dbo.Table1 SELECT 'ID'
UNION ALL SELECT 'foo';
-- adds both of these columns:
INSERT dbo.Table1 SELECT 'bar'
UNION ALL SELECT 'splan foob';
-- only adds one of these:
INSERT dbo.Table1 SELECT 'blat'
UNION ALL SELECT 'blat';
SELECT * FROM dbo.Table2;
Results:
ID foo bar splan foob blat
----------- ------------ ------------ ------------ ------------
Don't forget to clean up:
DROP TABLE dbo.Table1, dbo.Table2;
Related
I have the following columns in a SQLite DB.
id,ts,origin,product,bid,ask,nextts
1,2016-10-18 20:20:54.733,SourceA,Dow,1.09812,1.0982,
2,2016-10-18 20:20:55.093,SourceB,Oil,7010.5,7011.5,
3,2016-10-18 20:20:55.149,SourceA,Dow,18159.0,18161.0,
How can I populate the 'next timestamp' column (nextts) with the next timestamp for the same product (ts), from the same source? I've been trying the following, but I can't seem to put a subquery in an UPDATE statement.
UPDATE TEST a SET nextts = (select ts
from TEST b
where b.id> a.id and a.origin = b.origin and a.product = b.product
order by id asc limit 1);
If I call this, I can display it, but I haven't found a way of updating the value yet.
select a.*,
(select ts
from TEST b
where b.id> a.id and a.origin = b.origin and a.product = b.product
order by id asc limit 1) as nextts
from TEST a
order by origin, a.id;
The problem is that you're using table alias for table in UPDATE statement, which is not allowed. You can skip alias from there and use unaliased (but table-name prefixed) reference to its columns (while keeping aliased references for the SELECT), like this:
UPDATE TEST
SET nextts = (
SELECT b.ts
FROM TEST b
WHERE b.id > TEST.id AND
TEST.origin = b.origin AND
TEST.product = b.product
ORDER BY b.id ASC
LIMIT 1
);
Prefixing unaliased column references with the table name is necessary for SQLite to identify that you're referencing to unaliased table. Otherwise the id column whould be understood as the id from the closest[*] possible data source, in which case it's the aliased table (as b alias), while we're interested in the unaliased table, therefore we need to explicitly tell SQLite that.
[*] Closest data source is the one listed in the same query, or parent query, or parent's parent query, etc. SQLite is looking for the first data source (going from inner part to the outside) in the query hierarchy that defines this column.
I need to delete data from many tables based on one parameter
The problem is that two tables are related to each other so in order to delete data properly i need to store id's somewhere.
-- i would like to store temp data
-- this one is only for convienience to avoid repeating same select many times
create table ztTaryfa as select zt_taryfa from tw_zbiory_taryfy
where 1=2;
-- this one is mandatory but I dont know how to make it work
Create table wnioskiId as select poli_wnio_id_wniosku from polisy
where 1=2;
Begin
-- fill temp tables
insert into ztTaryfa (
select zt_taryfa from tw_zbiory_taryfy
where zt_zbior = :zbiorId);
insert into wnioskiId (
select poli_wnio_id_wniosku from polisy
where poli_taryfa_id in ztTaryfa);
- regular deletion
delete from POLISY_OT where ot_poli_id in (
select poli_id from polisy
where poli_taryfa_id in ztTaryfa);
commit;
delete from DANE_RAPORTOWE where DR_RPU_ID in (
select RPU_ID from ROZLICZ_PLIK_UBEZP where RPU_ROZLICZ_PLIK_ID in (
select RP_ID from ROZLICZ_PLIK
where RP_ZBIOR_ID = :zbiorId ));
commit;
-- and here we go I need to delete data from POLISY first
delete from POLISY where poli_taryfa_id in ztTaryfa;
commit;
-- but by doing it I lose ids which i need here,
-- so I have to store them somehow and use them here.
delete from WNIOSKI where wnio_id in wnioskiId;
commit;
End;
-- and now lets get rid off temp tables
drop table ztTaryfa;
commit;
drop table wnioskiId;
commit;
To sum up i just need to know how to store somewhere between Begin and End a result of a select query which I can later use in delete statement.
Sounds but I tried so many different methods and all seems to not work.
What u see above is just a 1/3 of the script so I rly would like to make it all simple to use with one parameter.
Thanks you in advance.
You can use global types as simple as this:
create or replace type myrec is object (myid number);
create or replace type mytemp_collection is table of myrec;
declare
v_temp_collection mytemp_collection;
begin
v_temp_collection := mytemp_collection();
select myrec (t.field_type_id ) bulk collect into v_temp_collection from fs_field_types t
where mod(t.field_type_id+1,3)=0; -- for example
FOR i IN 1 .. v_temp_collection.count LOOP
DBMS_OUTPUT.put_line(v_temp_collection(i).myid);
End loop;
delete fs_field_types_back t where t.field_type_id in (select myid from table(v_temp_collection));
end;
Change select and where clause according to your business.
This works:
sqlite> .databases
seq name file
--- --------------- ----------------------------------------------------------
0 main /path/to/db
2 uni /path/to/db
also this:
sqlite> pragma main.table_info(tsv_storage);
0|id|int|0||0
1|seqid|text|0||0
...
and this:
sqlite> select count(*) from main.tsv_storage;
198159
and also the attached database works:
sqlite> select * from uni.fasta_storage where uni.fasta_storage.id = 1;
1 MASNTVSAQ... Q197F8.1 002R_IIV3 Uncharacterized protein 002R Q197F8
but this not:
sqlite> select main.tsv_storage.seqid where main.tsv_storage.id=8;
Error: no such column: main.tsv_storage.seqid
EDIT:
and I have also problems with this, do I have to join the tables?
insert into main.tsv_storage(seqlength) select length(fasta) from
uni.fasta_storage where uni.fasta_storage.title = main.tsv_storage.seqid;
Error: no such column: main.tsv_storage.seqid
It happens for all columns, not only seqid. I think I did everything that is explained here: http://sqlite.awardspace.info/syntax/sqlitepg12.htm
What am I missing?
sqlite> select * from main.tsv_storage.seqid where main.tsv_storage.id=8;
You have not defined where to look for the selection. You need to tell the query what fields to search within the table, then define which table you are searching. The select * portion tells the query to look in all fields within the table. The from portion of the query tells the processes what table to look in. And lastly the where portion tells the query what to match when looking.
When using INSERT ... SELECT ..., the SELECT part must be valid query.
You cannot access a column like main.tsv_storage without having the table in the FROM clause:
INSERT INTO main.tsv_storage(seqlength)
SELECT length(fasta)
FROM uni.fasta_storage, main.tsv_storage
WHERE uni.fasta_storage.title = main.tsv_storage.seqid;
And the entire commands looks suspicious.
Are you sure you don't want to update the values in the seqlength column for existing records?
In that case, you would use something like this:
UPDATE main.tsv_storage
SET seqlength = (SELECT length(fasta)
FROM uni.fasta_storage
WHERE uni.fasta_storage.title = main.tsv_storage.seqid);
I want to select Column from Table based on Its Order
like
create Table Products
(
ProductId Int,
ProductName varchar(50)
)
lets Say I don't Know the name of the second column.
How I can get it like :
Select Col1,Col2 From Product
For SQL Server:
You can't do this in the SELECT clause. You can't select based on the order number of the column. You have to list the columns' names you need to select explicitly, otherwise, use SELECT * to list all. Me be if you are using a data reader object or any other ado.net methods to get the data from database you can do something like this, but this will be based on the column names list listed in your SQL statement.
However, you can do something like this dynamically, by reading columns' metadata ordinal_position from information_schema.columns as explained in the following answer:
Is it possible to select sql server data using column ordinal position?
But, you can do this in the ORDER BY clause. You can ORDER BY column number:
SELECT *
FROM TableName
ORDER BY 2; -- for col2
But this is not recommended to use in ORDER BY or in the SELECT (if any). Furthermore, columns order is not significant in the relational model.
Update: If you want to select at least 3 columns from any table parameter passed to your stored procedure. Try this as follows:
Your stored procedure supposed to receive a parameter #tableNameParam. The folowing code should return the first three columns from the #tablenameParam passed to the stored procedure:
DECLARE #col1 AS VARCHAR(100);
DECLARE #col2 AS VARCHAR(100);
DECLARE #col3 AS VARCHAR(100);
DECLARE #tableNameParam AS VARCHAR(50) = 'Tablename';
DECLARE #sql AS VARCHAR(MAX) ;
SELECT #col1 = column_name FROM information_schema.columns
WHERE table_name = #tableNameParam
AND ordinal_position = 1;
SELECT #col2 = column_name FROM information_schema.columns
WHERE table_name = #tableNameParam;
AND ordinal_position = 2;
SELECT #col3 = column_name FROM information_schema.columns
WHERE table_name = #tableNameParam;
AND ordinal_position = 3;
SET #sql = 'SELECT ' + col1 + ',' + col2 ' + 'col3 ' + FROM ' + #tablename;
you always can do
select * from Product
I'd like to share the following code as a solution to CRUD processing on Ordinal Position within a table. I had this problem today and it took me quite a long time to research and find a working solution. Many of the posted answers indicated that it was not possible to interact with the tables columns on an Ordinal bases but as indicated in the post above using the information_schema table will allow using the column position.
My situation was interacting with a table populated through the use of a pivot view so the columns are always changing based on the data, which is fine in a view result but when the dataset is stored into a table the columns are dynamic. The column names are a Year-Month combination such as 201801, 201802 with an Item Number as a primary key. This pivot table is to indicate manufacturing quantities by Year-Month on a rolling 12 month period so each month the column names with change/shift which changes their ordinal position when the table is rebuilt each month.
The Pivot view is used to build the Staging table, The Staging table is used to build the
Target table so the ordinal position of the staging and target tables are lined up with the same ordinal position.
Declare #colname Varchar(55) -- Column Name
Declare #ordpos INT -- Ordinal Position
Declare #Item Varchar(99) -- PK
Declare #i INT -- Counter
Declare #cnt INT -- Count
Declare #ids table(idx int identity(1,1), Item Varchar(25))
-- Item List
Insert INTO #ids Select Item From DBName.Schema.TableName
select #i = min(idx) - 1, #cnt = max(idx) from #ids
-- Row Loop
While #i < #cnt
Begin
Select #i = #i + 1
Set #ordpos=3
Set #Item = (select Item from #ids where idx = #i)
-- Column Loop
While #ordpos < 27
Begin
Select #colname =column_name From INFORMATION_SCHEMA.Columns Where table_name='TargetTable' and ordinal_position=#ordpos
Exec ('Update TargetTable set ['+#colname+']= (Select ['+#colname+'] From StagingTable Where Item='''+#Item+''') where Item='''+#Item+'''')
Set #ordpos=#ordpos + 1
End -- End Column Loop
End -- End Row Loop
The code here will loop through the Item matrix by rows and by columns and uses Dynamic SQL to build the action, in this case the action is an update but it could just as easily be a select. Each column is processed through the While Loop and then loops through the next row. This allows updates to a specific cell in the matrix by (Item X YearMonth) without actually knowing what the column name at a given position.
The one concern is that depending on the size of the data in this matrix it can be SLOW. I just wanted to show this as a way to use unknown column names in an ordinal position.
I have a MySQL table Page with 2 columns: PageID and OrderByMethod.
I also then have a Data table with lots of columns including PageID (the Page the data is on), DataName, and DataDate.
I want OrderByMethod to have one of three entries: Most Recent Data First, Most Recent Data Last, and Alphabetically.
Is there a way for me to tack an "ORDER BY" clause to the end of this query that will vary its ordering method based on the contents of the "OrderByMethod" column? For example, in this query, I would want to have the ORDER BY clause contain whatever ordering rule is stored in Page 1's OrderByMethod column.
GET * FROM `Data` WHERE `Data`.`PageID`=1 ORDER BY xxxxxx;
Maybe a SELECT clause in the ORDER BY clause? I'm not sure how that would work though.
Thanks!
select Data.*
from Data
inner join Page on (Data.PageID=Page.PageID)
where Data.PageID=1
order by
if(Page.OrderByMethod='Most Recent Data First', now()-DataDate,
if(Page.OrderByMethod='Most Recent Data Last', DataDate-now(), DataName)
);
You can probably do this with the IF syntax to generate a column that you can then order by.
SELECT *, IF(Page.OrderBy = 'Alphabetically', Data.DataName, IF(Page.OrderBy = 'Most Recent Data First', NOW() - Data.DataDate, Data.DataDate - NOW())) AS OrderColumn
FROM Data
INNER JOIN Page ON Data.PageID = Page.PageID
WHERE Page.PageID = 1
ORDER BY OrderColumn
The direction of the ordering is determined in the calculation of the data instead of specifying a direction in the ORDER BY
Can you just append the order by clause to the select statement and rebind the table on postback?
If you want to use the content of the column in Page table as an expression in ORDER BY you have to do it using prepared statements. Let say, you store in OrderByMethod something like "field1 DESC, field2 ASC" and you want this string to be used as it is:
SET #order_by =(SELECT OrderByMethod FROM Page WHERE id = [value]);
SET #qr = CONCAT(your original query,' ORDER BY ', #order_by);
PREPARE stmt FROM #qr;
EXECUTE stmt;
DEALLOCATE PREPARE stmt;
If you want the result set to be sorted based on the value of OrderByMethod , you can use IF as it was already mentioned by others, or CASE :
...
ORDER BY
CASE OrderByMethod
WHEN 'val1' THEN field_name1
WHEN 'val2' THEN field_name2
....etc
END