Cannot write to table with computed column

[chenfeiw]

I have a table like this:
Create table Test
(
Id int,
Year nvarchar(4),
Month nvarch(2),
Date As (Year + '-' + Month)
)

Because Date is a computed column, my dataframe doesn't have this column, I get exception 'Spark Dataframe and SQL Server table have different numbers of columns'.

[shivsood]

does this work with default JDBC connector?
Here's what needs to be done to support this.

1. Support option to not do a strict check
2. Utilize SqlBulkCopyColumnMapping during bluk copy to map specific columns.

[chenfeiw]

Yes, it works with default JDBC.

[shivsood]

Fix would be required to support a non-strict option. Also Utilize SqlBulkCopyColumnMapping during bulk copy to map specific columns.

[shivsood]

Looks same as #14

[pmooij]

does this work with default JDBC connector?
Here's what needs to be done to support this.

1. Support option to not do a strict check
2. Utilize SqlBulkCopyColumnMapping during bluk copy to map specific columns.

indeed, these 2 features are needed to deal with Identity, Computed & Defaults constraints.
quite cumbersome to do the workarrounds now

[rajmera3]

Solved in #52