我创建了以下procedure：
if object_id('tool.create_insert', 'P') is null
begin
exec('create procedure tool.create_insert as');
end;
go
alter procedure tool.create_insert(@schema varchar(200) = 'dbo',
@table varchar(200),
@where varchar(max) = null,
@top int = null,
@insert varchar(max) output)
as
begin
declare @insert_fields varchar(max),
@select varchar(max),
@error varchar(500),
@query varchar(max);
declare @values table(description varchar(max));
set nocount on;
-- Get columns
select @insert_fields = isnull(@insert_fields + ', ', '') + c.name,
@select = case type_name(c.system_type_id)
when 'varchar' then isnull(@select + ' + '', '' + ', '') + ' isnull('''''''' + cast(' + c.name + ' as varchar) + '''''''', ''null'')'
when 'datetime' then isnull(@select + ' + '', '' + ', '') + ' isnull('''''''' + convert(varchar, ' + c.name + ', 121) + '''''''', ''null'')'
else isnull(@select + ' + '', '' + ', '') + 'isnull(cast(' + c.name + ' as varchar), ''null'')'
end
from sys.columns c with(nolock)
inner join sys.tables t with(nolock) on t.object_id = c.object_id
inner join sys.schemas s with(nolock) on s.schema_id = t.schema_id
where s.name = @schema
and t.name = @table;
-- If there's no columns...
if @insert_fields is null or @select is null
begin
set @error = 'There''s no ' + @schema + '.' + @table + ' inside the target database.';
raiserror(@error, 16, 1);
return;
end;
set @insert_fields = 'insert into ' + @schema + '.' + @table + '(' + @insert_fields + ')';
if isnull(@where, '') <> '' and charindex('where', ltrim(rtrim(@where))) < 1
begin
set @where = 'where ' + @where;
end
else
begin
set @where = '';
end;
set @query = 'select ' + isnull('top(' + cast(@top as varchar) + ')', '') + @select + ' from ' + @schema + '.' + @table + ' with (nolock) ' + @where;
insert into @values(description)
exec(@query);
set @insert = isnull(@insert + char(10), '') + '--' + upper(@schema + '.' + @table);
select @insert = @insert + char(10) + @insert_fields + char(10) + 'values(' + v.description + ');' + char(10) + 'go' + char(10)
from @values v
where isnull(v.description, '') <> '';
end;
go
那么你可以这样使用它：
declare @insert varchar(max),
@part varchar(max),
@start int,
@end int;
set @start = 1;
exec tool.create_insert @schema = 'dbo',
@table = 'customer',
@where = 'id = 1',
@insert = @insert output;
-- Print one line to avoid the maximum 8000 characters problem
while len(@insert) > 0
begin
set @end = charindex(char(10), @insert);
if @end = 0
begin
set @end = len(@insert) + 1;
end;
print substring(@insert, @start, @end - 1);
set @insert = substring(@insert, @end + 1, len(@insert) - @end + 1);
end;
输出会是这样的：
--DBO.CUSTOMER
insert into dbo.customer(id, name, type)
values(1, 'CUSTOMER NAME', 'F');
go
如果您只想获取一系列行，请使用@top 参数，如下所示：
declare @insert varchar(max),
@part varchar(max),
@start int,
@end int;
set @start = 1;
exec tool.create_insert @schema = 'dbo',
@table = 'customer',
@top = 100,
@insert = @insert output;
-- Print one line to avoid the maximum 8000 characters problem
while len(@insert) > 0
begin
set @end = charindex(char(10), @insert);
if @end = 0
begin
set @end = len(@insert) + 1;
end;
print substring(@insert, @start, @end - 1);
set @insert = substring(@insert, @end + 1, len(@insert) - @end + 1);
end;