T-SQL Update table columns using function

问题

I have the following table:

RecordID 
Name
Col1
Col2
....
ColN

The RecordID is BIGINT PRIMARY KEY CLUSTERED IDENTITY(1,1) and RecordID and Name are initialized. The other columns are NULLs.

I have a function which returns information about the other columns by Name.

To initialized my table I use the following algorithm:

Create a LOOP
Get a row, select its Name value
Execute the function using the selected name, and store its result in temp variables
Insert the temp variables in the table
Move to the next record

Is there a way to do this without looping?

回答1:

Cross apply was basically built for this

SELECT D.deptid, D.deptname, D.deptmgrid
    ,ST.empid, ST.empname, ST.mgrid
FROM Departments AS D
    CROSS APPLY fn_getsubtree(D.deptmgrid) AS ST;

Using APPLY

UPDATE some_table
SET some_row = another_row,
    some_row2 = another_row/2
FROM some_table st
  CROSS APPLY
    (SELECT TOP 1 another_row FROM another_table at WHERE at.shared_id=st.shared_id)
WHERE ...

using cross apply in an update statement

回答2:

You can simply say the following if you already have the records in the table.

UPDATE MyTable
SET 
    col1 = dbo.col1Method(Name),
    col2 = dbo.col2Method(Name),
    ...

While inserting new records, assuming RecordID is auto-generated, you can say

INSERT INTO MyTable(Name, Col1, Col2, ...)
VALUES(@Name, dbo.col1Method(@Name), dbo.col2Method(@name), ...)

where @Name contains the value for the Name column.

来源：https://stackoverflow.com/questions/12583043/t-sql-update-table-columns-using-function

标签

tsql

function

loops

updates