I stumbled upon a very strange behaviour while working on some T-SQL Code.
I am working on a SQL Server 2008 R2 SP2 (build nr.: 10.50.4000).
My question to you guys is if anybody has seen such a behaviour before or if anybody might be able to explain it to me.
So,
What's the situation?
We have a table, which looks like that:
product_number | id_object | position_in_product
---------------------------------------------------
1 | 101 | 1
1 | 102 | 1
1 | 103 | 1
2 | 201 | 1
2 | 202 | 1
2 | 203 | 1
Multiple object ids are allocated to one product number. The order should be defined by the position_in_product column. The funny part lies exactly in establishing that order.
Of course, after doing that the table should look like this:
product_number | id_object | position_in_product
---------------------------------------------------
1 | 101 | 1
1 | 102 | 2
1 | 103 | 3
2 | 201 | 1
2 | 202 | 2
2 | 203 | 3
What's going on?
To update the order column we create a cursor with the following statement:
DECLARE
table_runner CURSOR LOCAL FORWARD_ONLY FOR
SELECT id_object, product_number
FROM table
WHERE ident = #ident
ORDER BY product_number
By using this cursor and counting the rows with the same product_number we should be able to update the position_in_product column. (This has worked in every installation until now)
To move the cursor to the next row we use this:
FETCH next from table_runner
INTO #table_runner$id_object, #table_runner$product_number
The whole function looks like this:
OPEN table_runner
FETCH next from table_runner
INTO #table_runner$id_object, #table_runner$product_number
while ##FETCH_STATUS = 0
BEGIN
/* update_logic */
FETCH next from table_runner
INTO #table_runner$id_object, #table_runner$product_number
END
CLOSE table_runner
And that is the part, that does not work as expected.
The fetch will not give me the next row. I am getting always the same result row.
The while loop does never end, the fetch_status is always 0, but the result stays the same.
The Workaround
After searching the web for quite a while without any results i decided to try a more pragmatical way and put another FETCH statement in.
I know that the id_object variable is unique and has to change in every loop cycle,
so i remembered the last fetched id and put this under the loop fetch statement:
if #id_object_memory = #table_runner$id_object
begin
FETCH next from table_runner
INTO #table_runner$id_object, #table_runner$product_number
set #id_object_memory = #table_runner$id_object
end
else
set #id_object_memory = #table_runner$id_object
With that the loop works as expected, the column in question is updated as it should and the cursor will reach the end of the result set.
The big ?
Has anyone any explanation for that?
There are more cursor defined in the same procedure and they all work as expected.
I have absolute no clue how to explain this.
So, thanks for reading ;)
I can't help with the cursor issue, I've never seen this before, but should point out you don't need a cursor at all to do this update. You can simply use:
WITH CTE AS
( SELECT Product_Number,
ID_Object,
Position_in_Product,
RowNumber = ROW_NUMBER() OVER(PARTITION BY Product_Number
ORDER BY id_object)
FROM T
WHERE ident = #ident
)
UPDATE CTE
SET Position_in_Product = RowNumber;
Example on SQL Fiddle
You possibly don't even need to store this column, and can just use ROW_NUMBER in a query where the position_in_product is required.
Cursors are so 2000 ;-)
Seriously though; avoid cursors at all costs. Set-based operations > looping.
Just create a view with the following:
CREATE VIEW your_view
AS
SELECT product_number
, id_object
, Row_Number() OVER (PARTITION BY product_number ORDER BY id_object) As position_in_product
FROM your_table
;
No need to ever perform the update; the row numbers will "automatically" recalculate.
Related
I have two tables, one contains user data and the other contains user ranking information (points needed for the promotion)
Let's say that the user table looks like this:
login | ArticlePoints | PhotoPoints | StageId
and the user ranking information table looks like this:
StageId | StageName | MinimumPoints
and the user information table might contain data like this:
1 | Beginner | 100
2 | Advanced | 200
3 | Expert | 300
What I would like to have is a procedure which does add user points and check whether it is enough for the ranking promotion. Right now I do it like this:
I do have a function which does check "manually" whether the user points is between 100 and 200 and then it does set the user stage = 2, id it's more it check whether it's between 200 and 300 etc.
Stored procedure which does update users set stage = MYFUNCTION from the point 1.
The thing is that it's not a good solution, right now it is not ready for the easy updates(I can't just add Super Expert with minimum 400 points, I'd need to edit the function).
I am trying to prepare a better solution for this problem but I have no idea how to "connect" both tables.
Write an UPDATE query that returns the StageID for the calculated values, something like:
UPDATE t1
SET t1.StageID =
(SELECT TOP 1 StageID
FROM [RANKING_TABLE] t2
WHERE t1.ArticlePoints + t1.PhotoPoints >= t2.MinimumPoints
ORDER BY t2.MinimumPoints DESC)
FROM [USER_TABLE] t1
So if the USER has 250 points in total, Beginner and Advanced would be achieved, using the TOP 1 and the ORDER BY t2.MinimumPoints DESC, would select the highest Stage.
I'm a little stuck here. I'm trying to modify a returned View based on a condition. I'm fairly green on SQL and am having a bit of difficultly with the returned result. Heres a partial component of the view I wrote:
WITH A AS (
SELECT
ROW_NUMBER() OVER (PARTITION BY fkidContract,fkidTemplateItem ORDER BY bStdActive DESC, dtdateplanned ASC) AS RANK,
tblWorkItems.fkidContract AS ContractNo,
....
FROM tblWorkItems
WHERE fkidTemplateItem IN
(2895,2905,2915,2907,2908,
2909,3047,2930,2923,2969,
2968,2919,2935,2936,2927,
2970,2979)
AND ...
)
SELECT * FROM A WHERE RANK = 1
The return result is similar to the following:
ContractNo| ItemNumber | Planned | Complete
001 | 100 | 01/01/1900 | 02/01/1900
001 | 101 | 03/04/1900 | 02/01/1901
001 | 102 | 03/06/1901 | 02/08/1900
002 | 100 | 01/03/1911 | 02/08/1913
This gives me the results I expect, but due a nightmare crystal report I need to alter this view slightly. I want to take this returned result set and modify an existing column with a value pulled from the same table and the same Contract relationship, something like the following:
UPDATE A
SET A.Completed = ( SELECT R.Completed
FROM myTable R
INNER JOIN A
ON A.ContractNo = R.ContractNo
WHERE A.ItemNumber = 100 AND R.ItemNumber = 101
)
What I'm trying to do is modify the "Completed Date" of one task and make it the complete date of another task if they both share the same ContractNo field value.
I'm not sure about the ItemNumber relationships between A and R (perhaps it was just for testing...), but it seems like you don't really want to UPDATE anything, but you want to use a different value under some circumstances. So, maybe you just want to change the non-cte part of your query to something like:
SELECT A.ContractNo, A.ItemNumber, A.Planned,
COALESCE(R.Completed,A.Completed) as Completed
FROM A
LEFT OUTER JOIN myTable R
ON A.ContractNo = R.ContractNo
AND A.ItemNumber = 100 AND R.ItemNumber = 101 -- I'm not sure about this part
WHERE A.Rank = 1
So it turns out that actually reading the vendor documentation helps :)
SELECT
column1,
column2 =
case
when date > 1999 then 'some value'
when date < 1999 then 'other value'
else 'back to the future'
end
FROM ....
For reference, the total query did a triple inner join over ~5 million records and this case statement was surprisingly performant.
I suggest that this gets closed as a duplicate.
I have a table that looks like this:
--------------------------------------------
| Date | House# | Subscription |
--------------------------------------------
| 3/02/10 | x | Monthly |
--------------------------------------------
| 3/03/10 | y | Weekly |
--------------------------------------------
| 3/04/10 | z | Daily |
--------------------------------------------
I need a command that will take a column name and an int and shift the values in those columns up so many levels. So (house, 1) would put z where y is, y where x is, and z would go to 0/Null. Whereas (house, 2) would put z where x is and y and z would go to 0/null.
I understand that SQL does not actually extract ables row by row, so is this possible?
Thanks ahead of time!
You can do this in a stored procedure using cursors.
You should use PL/SQL, here is an example (not for this particular example):
DECLARE
CURSOR cpaises
IS
SELECT CO_PAIS, DESCRIPCION, CONTINENTE
FROM PAISES;
co_pais VARCHAR2(3);
descripcion VARCHAR2(50);
continente VARCHAR2(25);
BEGIN
OPEN cpaises;
LOOP
FETCH cpaises INTO co_pais,descripcion,continente;
EXIT WHEN cpaises%NOTFOUND;
dbms_output.put_line(descripcion);
END LOOP;
CLOSE cpaises;
END;
I think you could use a variable to indicate which column to select and to update, and inside a loop, you can have an array, with the last n values.
You can use PL/SQL routine .Take the column name and number as input and then implement the logic as you want. Cursors as suggested above is one of the options that you have.
I would think adding a column that contains a value to use as a sort order you could then update that column as needed and then ordered by that column. If it is not possible to change that table perhaps you could create a new table to hold the sort column and join the two
I'm in a very, very tight situation here. I have an SQL query running on SQL Server 2005:
SELECT col1,col2,col3 FROM myTable
Which of course gives:
col1 | col2 | col3
------------------
1 | a | i
2 | b | ii
etc
I need to, if possible, add a COUNT query so that it will return the number of records returned. I cannot use GROUP BY or an aggregate function (It's a very edge case on some very inflexible software).
Ideally, something like this:
SELECT col1,col2,col3,COUNT(NumberOfRows) as NumRows FROM myTable
col1 | col2 | col3| NumRows
---------------------------
1 | a | i | 2
2 | b | ii | 2
I realise that this is bad. And inefficient. And against all good practices. But I'm in a corner with software whose architecture was frozen in stone in 1991!
Uuh so it turns out my collegue came back with an answer 30 seconds after asking the question.
The correct syntax is:
SELECT col1,col2,col3,##ROWCOUNT as NumRows FROM myTable
Looks like using ##ROWCOUNT will return the number of rows processed by the previous query, so I'm not sure that this is a valid solution. I think this is because ##ROWCOUNT is internally set after the query is run, so it is best used after the query has already completed. Therefore, it won't return the number of rows processed by the query in which it is placed.
let's assume i have a self referencing hierarchical table build the classical way like this one:
CREATE TABLE test
(name text,id serial primary key,parent_id integer
references test);
insert into test (name,id,parent_id) values
('root1',1,NULL),('root2',2,NULL),('root1sub1',3,1),('root1sub2',4,1),('root
2sub1',5,2),('root2sub2',6,2);
testdb=# select * from test;
name | id | parent_id
-----------+----+-----------
root1 | 1 |
root2 | 2 |
root1sub1 | 3 | 1
root1sub2 | 4 | 1
root2sub1 | 5 | 2
root2sub2 | 6 | 2
What i need now is a function (preferrably in plain sql) that would take the id of a test record and
clone all attached records (including the given one). The cloned records need to have new ids of course. The desired result
would like this for example:
Select * from cloningfunction(2);
name | id | parent_id
-----------+----+-----------
root2 | 7 |
root2sub1 | 8 | 7
root2sub2 | 9 | 7
Any pointers? Im using PostgreSQL 8.3.
Pulling this result in recursively is tricky (although possible). However, it's typically not very efficient and there is a much better way to solve this problem.
Basically, you augment the table with an extra column which traces the tree to the top - I'll call it the "Upchain". It's just a long string that looks something like this:
name | id | parent_id | upchain
root1 | 1 | NULL | 1:
root2 | 2 | NULL | 2:
root1sub1 | 3 | 1 | 1:3:
root1sub2 | 4 | 1 | 1:4:
root2sub1 | 5 | 2 | 2:5:
root2sub2 | 6 | 2 | 2:6:
root1sub1sub1 | 7 | 3 | 1:3:7:
It's very easy to keep this field updated by using a trigger on the table. (Apologies for terminology but I have always done this with SQL Server). Every time you add or delete a record, or update the parent_id field, you just need to update the upchain field on that part of the tree. That's a trivial job because you just take the upchain of the parent record and append the id of the current record. All child records are easily identified using LIKE to check for records with the starting string in their upchain.
What you're doing effectively is trading a bit of extra write activity for a big saving when you come to read the data.
When you want to select a complete branch in the tree it's trivial. Suppose you want the branch under node 1. Node 1 has an upchain '1:' so you know that any node in the branch of the tree under that node must have an upchain starting '1:...'. So you just do this:
SELECT *
FROM table
WHERE upchain LIKE '1:%'
This is extremely fast (index the upchain field of course). As a bonus it also makes a lot of activities extremely simple, such as finding partial trees, level within the tree, etc.
I've used this in applications that track large employee reporting hierarchies but you can use it for pretty much any tree structure (parts breakdown, etc.)
Notes (for anyone who's interested):
I haven't given a step-by-step of the SQL code but once you get the principle, it's pretty simple to implement. I'm not a great programmer so I'm speaking from experience.
If you already have data in the table you need to do a one time update to get the upchains synchronised initially. Again, this isn't difficult as the code is very similar to the UPDATE code in the triggers.
This technique is also a good way to identify circular references which can otherwise be tricky to spot.
The Joe Celko's method which is similar to the njreed's answer but is more generic can be found here:
Nested-Set Model of Trees (at the middle of the article)
Nested-Set Model of Trees, part 2
Trees in SQL -- Part III
#Maximilian: You are right, we forgot your actual requirement. How about a recursive stored procedure? I am not sure if this is possible in PostgreSQL, but here is a working SQL Server version:
CREATE PROCEDURE CloneNode
#to_clone_id int, #parent_id int
AS
SET NOCOUNT ON
DECLARE #new_node_id int, #child_id int
INSERT INTO test (name, parent_id)
SELECT name, #parent_id FROM test WHERE id = #to_clone_id
SET #new_node_id = ##IDENTITY
DECLARE #children_cursor CURSOR
SET #children_cursor = CURSOR FOR
SELECT id FROM test WHERE parent_id = #to_clone_id
OPEN #children_cursor
FETCH NEXT FROM #children_cursor INTO #child_id
WHILE ##FETCH_STATUS = 0
BEGIN
EXECUTE CloneNode #child_id, #new_node_id
FETCH NEXT FROM #children_cursor INTO #child_id
END
CLOSE #children_cursor
DEALLOCATE #children_cursor
Your example is accomplished by EXECUTE CloneNode 2, null (the second parameter is the new parent node).
This sounds like an exercise from "SQL For Smarties" by Joe Celko...
I don't have my copy handy, but I think it's a book that'll help you quite a bit if this is the kind of problems you need to solve.