Failing test on sql server datatypes

[Peterdha]

This issue seems similar to [https://github.com//issues/662] but for sqlserver
Using the product table from sql server Adventureworks sample database I generated DDL.
Already there needed to do some edits as it didn't recognize the 'money' datatype.

CREATE TABLE [SalesLT].[Product]( [ProductID] [int] IDENTITY(1,1) NOT NULL, [Name] [nvarchar](50) NOT NULL, [ProductNumber] [nvarchar](25) NOT NULL, [Color] [nvarchar](15) NULL, [StandardCost] [decimal](8, 2) NOT NULL, [ListPrice] [decimal](8, 2) NOT NULL, [Size] [nvarchar](5) NULL, [Weight] [decimal](8, 2) NULL, [ProductCategoryID] [int] NULL, [ProductModelID] [int] NULL, [SellStartDate] [datetime] NOT NULL, [SellEndDate] [datetime] NULL, [DiscontinuedDate] [datetime] NULL, [ThumbNailPhoto] [varbinary](max) NULL, [ThumbnailPhotoFileName] [nvarchar](50) NULL, [rowguid] [uniqueidentifier] NOT NULL, [ModifiedDate] [datetime] NOT NULL )

Then imported the CREATE TABLE statement and wrote out the datacontract:
tst_dc = DataContract().import_from_source(format='sql',source='product_DDL.sql',dialect='tsql')

Testing the generated contract against the same table yields below failures:

[GabrielEisenberg]

We have faced a similar issue when trying to use import on SQL Server DDLs whereby the money and datetime2(7) both get converted to variant. We've then found the problems when running data contract lint to check the validity of the configs. In our case, datetime2(7) should probably be converted to timestamp and money should be converted to decimal with precision 5 and scale 3.

Is it perhaps possible to add custom conversion rules within the import functionality or to better handle these mappings internal to the tool?

Thanks so much!

[dmaresma]

Hi I refactor the sql import, I add the money data type translation to decimal, and the datatime2 is related to the timestamp_ntz (no time zone) it's fixe in the PR #790