I have implemented an audit trail framework based on the information provided by the first answer to the following post:
SQL Server history table - populate through SP or Trigger?
Ultimately, the framework that I have implemented uses three triggers per table that insert audit information based on changes to the tables.
My insert and delete audit triggers are fairly simple. However, the update triggers are far more complex because the trigger has to check to determine whether or not each column is under audit control and then perform an insert based on whether or not the column values in the Inserted and Deleted columns are equal or not since I don't want to write unnecessary audit records. Ultimately, I want to know if there is a way to write a stored procedure that will reduce the amount of code in my trigger by allowing me to dynamically perform the insert statement below. Basically, I envision the trigger firing the sproc with each column name that is under audit control and then the stored procedure will used the column name to perform the code snippet below. Currently, I have the code below for every column under audit control which unfortunately results in lots of redundant code.
Revised Trigger After Suggested Changes
CREATE TRIGGER [dbo].[Audit_Customers_Update] ON [dbo].[Customers]
FOR UPDATE AS
select FirstName,LastName into #deleted from deleted;
declare /*const*/ @TABLE_NAME sysname = '[table name]';
declare f cursor
local
forward_only
read_only
for
select c.name, quotename(c.name, '[')
from
sys.columns c
inner join sys.types t on c.system_type_id = t.system_type_id
where
c.object_id = object_id(@TABLE_NAME)
and c.is_computed = 0
and c.is_identity = 0
and t.name not in ('text', 'image', 'timestamp', 'xml')
and (substring(COLUMNS_UPDATED(), ((c.column_id - 1) / 8) + 1, 1) & power(2, (c.column_id - 1) % 8)) > 0
;
declare @field_name sysname, @field_name_sanitised sysname;
create table #results (row_id int not null,
field_name sysname not null,
oldval nvarchar(150) null,
newval nvarchar(150) null);
-- For each changed field, insert what exactly changed into #results
open f;
fetch next from f into @field_name, @field_name_sanitised;
while @@fetch_status = 0
begin
declare @query nvarchar(4000);
set @query = N'insert into #results(row_id, field_name, oldval, newval)
select d.row_id, @field_name, d.' + @field_name_sanitised + N', i.' + @field_name_sanitised + N'
from
#deleted d inner join ' + @TABLE_NAME + N' i on d.row_id = i.row_id
where
(d.' + @field_name_sanitised + N' <> i.' + @field_name_sanitised + N')
or
(case when d.' + @field_name_sanitised + N' is null then 1 else 0 end <> case when i.' + @field_name_sanitised + N' is null then 1 else 0 end);'
;
exec sp_executesql
@stmt = @query,
@params = N'@field_name sysname',
@field_name = @field_name
;
fetch next from f into @field_name, @field_name_sanitised;
end;
close f;
deallocate f;
-- Do something meaningful to #results here
How do I access #results? Do I have to use a cursor?
We've solved that problem in the following way.
select <list of tracked columns here> into #deleted from deleted;
declare /*const*/ @TABLE_NAME sysname = '[table name]';
declare f cursor
local
forward_only
read_only
for
select c.name, quotename(c.name, '[')
from
sys.columns c
inner join sys.types t on c.system_type_id = t.system_type_id
where
c.object_id = object_id(@TABLE_NAME)
and c.is_computed = 0
and c.is_identity = 0
and t.name not in ('text', 'image', 'timestamp', 'xml')
and (substring(COLUMNS_UPDATED(), ((c.column_id - 1) / 8) + 1, 1) & power(2, (c.column_id - 1) % 8)) > 0
;
declare @field_name sysname, @field_name_sanitised sysname;
create table #results (row_id int not null, field_name sysname not null, oldval nvarchar(150) null, newval nvarchar(150) null);
-- For each changed field, insert what exactly changed into #results
open f;
fetch next from f into @field_name, @field_name_sanitised;
while @@fetch_status = 0
begin
declare @query nvarchar(4000);
set @query = N'insert into #results(row_id, field_name, oldval, newval)
select d.row_id, @field_name, d.' + @field_name_sanitised + N', i.' + @field_name_sanitised + N'
from
#deleted d inner join ' + @TABLE_NAME + N' i on d.row_id = i.row_id
where
(d.' + @field_name_sanitised + N' <> i.' + @field_name_sanitised + N')
or
(case when d.' + @field_name_sanitised + N' is null then 1 else 0 end <> case when i.' + @field_name_sanitised + N' is null then 1 else 0 end);'
;
exec sp_executesql
@stmt = @query,
@params = N'@field_name sysname',
@field_name = @field_name
;
fetch next from f into @field_name, @field_name_sanitised;
end;
close f;
deallocate f;
-- Do something meaningful to #results here
Related reading:
Ran into a similar problem... figured it out this way... may not be the most elegant solution but works for the compliance guys... So here goes...
Based on the solution given here
The xml is extracted with FOR XML from the trigger that updated the table... The "OldValues" come from the DELETED table and the "NewValues" from the INSERTED table... so the final xml looks like this...
DECLARE @x XML= '<FieldData>
<UpdatedColumns>
<trType>OldValues</trType>
<ID>5</ID>
<def_label>TEST_TIE</def_label>
<def_code />
</UpdatedColumns>
<UpdatedColumns>
<trType>OldValues</trType>
<ID>4</ID>
<def_label>RP_TIE</def_label>
<def_code />
</UpdatedColumns>
<UpdatedColumns>
<trType>OldValues</trType>
<ID>3</ID>
<def_label>ERR_TIE</def_label>
<def_code />
</UpdatedColumns><UpdatedColumns>
<trType>NewValues</trType>
<ID>5</ID>
<def_label>TEST_TIE</def_label>
<def_code>A</def_code>
</UpdatedColumns>
<UpdatedColumns>
<trType>NewValues</trType>
<ID>4</ID>
<def_label>RP_TIE</def_label>
<def_code>A</def_code>
</UpdatedColumns>
<UpdatedColumns>
<trType>NewValues</trType>
<ID>3</ID>
<def_label>ERR_TIE</def_label>
<def_code>A</def_code>
</UpdatedColumns>
</FieldData>'
declare @timestamp datetime2= SYSDATETIME()
select
ID = identity(int,1,1),
T.N.value('local-name(.)', 'nvarchar(100)') as NodeName,
T.N.value('../ID[1]','nvarchar(100)') AS table_ID,
T.N.value('.', 'nvarchar(100)') as OldValue
INTO #old
from @x.nodes('//UpdatedColumns/*') as T(N)
WHERE T.N.value('../trType[1]', 'nvarchar(100)') ='OldValues'
select
ID = identity(int,1,1),
T.N.value('local-name(.)', 'nvarchar(100)') as NodeName,
T.N.value('../ID[1]','nvarchar(100)') AS Table_ID,
T.N.value('.', 'nvarchar(100)') as NewValue
into #new
from @x.nodes('//UpdatedColumns/*') as T(N)
WHERE T.N.value('../trType[1]', 'nvarchar(100)') ='NewValues'
SELECT n.table_ID, n.NodeName, o.OldValue, n.NewValue,@timestamp as transation_time FROM #new n
left outer JOIN #old o ON n.NodeName = o.NodeName AND n.ID = o.ID
WHERE isnull(o.[OldValue],'') <> isnull(n.[newValue],'') AND n.NodeName <> 'trType'
DROP TABLE #new,#old
GO
来源：https://stackoverflow.com/questions/8873335/how-to-create-generic-sql-server-stored-procedure-to-perform-inserts-into-audit