Schema DDL: separate redshift-specific code from standard SQL (Closes #372) #373
Conversation
| @@ -78,7 +80,7 @@ case class AddColumn( | |||
| columnName: String, | |||
| columnType: DataType, | |||
| default: Option[Default], | |||
| encode: Option[CompressionEncoding], | |||
| encode: Option[ColumnAttribute], | |||
There was a problem hiding this comment.
Because regular sql dbs do not have compression encoding, but may have other attributes.
szareiangm
force-pushed
the
master-sql-package
branch
from
ddba45b to
7fd6e27
Compare
szareiangm
force-pushed
the
master-sql-package
branch
from
7fd6e27 to
86fcf7d
Compare
|
@szareiangm has signed the Software Grant and Corporate Contributor License Agreement |
szareiangm
force-pushed
the
master-sql-package
branch
from
86fcf7d to
3089571
Compare
|
@chuwy I changed the code to use upper bound inheritance to keep the Ddl consistent as per your suggestion in the issue discussion. Does the code now similar to what you meant? |
szareiangm
force-pushed
the
master-sql-package
branch
from
3089571 to
7479c56
Compare
|
@chuwy I was wondering if I could follow up on this PR. It has been a while? |
|
Hey @szareiangm, sorry about that. Soon, I'll be backporting another SQL-like AST (BigQuery) to Schema DDL and it should give us some insights regarding sufficient volume of abstraction in these DDLs. |
I tried to follow up with the first step for proposed plan in #371 .
There is no serious logic change. It is just moving codes around. Please support with your comments.
cc: @Nafid @zakipatel