Question

I’ve created a few table types in my database to be used as stored procedure parameters. These correspond to real database tables, so if they’re out of sync there’ll be a problem. I’d like to add a unit test that looks at the two and fails if they are different, but I’m not sure where to start.

I don’t know if there’s a recommended way to do this – I was going to try to somehow pull out the column information, loop through it and fail the test if they’re different, but it seems a bit fiddly.

Is there a better way?

Answer 1

For SQL Server 2008, take a look at the sys.tables, sys.table_types and sys.columns system tables.

In one of my databases I have a table type called candidateRoutes and a physical (real) table called RouteArea

The following two queries:

select sys.columns.* from sys.table_types join sys.columns on sys.columns.object_id = sys.table_types.type_table_object_id where sys.table_types.name = 'candidateRoutes'
select sys.columns.* from sys.tables join sys.columns on sys.columns.object_id = sys.tables.object_id where sys.tables.name = 'RouteArea'

return:

object_id  name     column_id  system_type_id  user_type_id  max_length  precision  scale  collation_name  is_nullable  is_ansi_padded    is_rowguidcol  is_identity  is_computed  is_filestream  is_replicated  is_non_sql_subscribed  is_merge_published  is_dts_replicated  is_xml_document  xml_collection_id  default_object_id  rule_object_id  is_sparse  is_column_set
215671816  RouteId  1           56              56            4          10         0      NULL            0            0                 0              0            0            0              0              0                      0                   0                  0                0                  0                  0               0          0
215671816  Area     2          240             130           -1           0         0      NULL            0            0                 0              0            0            0              0              0                      0                   0                  0                0                  0                  0               0          0

and

object_id   name       column_id  system_type_id  user_type_id  max_length  precision  scale  collation_name  is_nullable  is_ansi_padded  is_rowguidcol  is_identity  is_computed  is_filestream  is_replicated  is_non_sql_subscribed  is_merge_published  is_dts_replicated  is_xml_document  xml_collection_id  default_object_id  rule_object_id  is_sparse  is_column_set
1675153013  RouteId    1          127             127            8          19         0      NULL            0            0               0              0            0            0              0              0                      0                   0                  0                0                  0                  0               0          0
1675153013  ValidFrom  2           61              61            8          23         3      NULL            0            0               0              0            0            0              0              0                      0                   0                  0                0                  0                  0               0          0
1675153013  ValidTo    3           61              61            8          23         3      NULL            1            0               0              0            0            0              0              0                      0                   0                  0                0                  0                  0               0          0
1675153013  Line       4          240             130           -1           0         0      NULL            0            0               0              0            0            0              0              0                      0                   0                  0                0                  0                  0               0          0
1675153013  Area       5          240             130           -1           0         0      NULL            1            0               0              0            0            0              0              0                      0                   0                  0                0                  0                  0               0          0

so you could perhaps do something like this:

with
    TableType as
        (select name, user_type_id, max_length, precision from sys.columns where object_id = (select type_table_object_id from sys.table_types where name = 'candidateRoutes')),
    PhysicalTable as
        (select name, user_type_id, max_length, precision from sys.columns where object_id = (select object_id from sys.tables where name = 'RouteArea'))
    select * from TableType full join PhysicalTable
        on TableType.name = PhysicalTable.name
    where TableType.name is null
       or PhysicalTable.name is null
       or TableType.user_type_id <> PhysicalTable.user_type_id
       or TableType.max_length   <> PhysicalTable.max_length
       or TableType.precision    <> PhysicalTable.precision

but including scale, collation_name, is_nullable, etc., to find all columns that do not match. In my case, I get:

name     user_type_id  max_length  precision  name       user_type_id  max_length  precision
RouteId  56            4           10         RouteId    127           8           19
NULL     NULL          NULL        NULL       ValidFrom  61            8           23
NULL     NULL          NULL        NULL       ValidTo    61            8           23
NULL     NULL          NULL        NULL       Line       130           -1          0

If no rows are returned, the type and the table are the same.