Adds initial aliasing support. #346
Conversation
The highlights: - Moving to use 'Qualified ColumnName' in many/most places. This allows for the aliasing to be used naturally in those places. - Add an alias type and function to create a 'Qualified ColumnName' with it. This signals that 'Alias' is in fact different than 'ColumnName' or 'TableName', and reusing them for aliasing is not always appropriate. - Adding a 'MarshallAlias' constructor to the 'SqlMarshaller' so that an alias can be attached to all of the natural fields in a marshaller in a single function call from users. Synthetic fields could be refactored with this facility, but that was left out here to keep the already big changeset smaller. - Many of the fieldDefinition helpers now have a version to support generating various subexpressions with an alias along with the fieldDefinition. This is a choice to allow the simple case to not deal with the complexity of the alias, but make it easier for when aliases are needed. - A small number of helpers are added to make generation of bits of SQL, such as a 'TableReferenceList' that includes the 'AS alias' piece.
| 'TableDefinition'. This is the safest way to build a 'Select', because table | ||
| name and columns are all read from the 'TableDefinition'. If the table is | ||
| being managed with Orville auto-migrations, this will match the schema in the | ||
| database. |
There was a problem hiding this comment.
This should probably say something about how the alias is used
| @@ -49,7 +49,7 @@ newtype ColumnDefinition | |||
| -} | |||
| columnDefinition :: | |||
| -- | The name the resulting column should have. | |||
| ColumnName -> | |||
| Qualified ColumnName -> | |||
There was a problem hiding this comment.
It seems like there are a number of replace like this one where ColumnName got replaced with Qualified ColumnName in DDL related expressions. In a good number of these I don't think qualified columns names are actually permitted by PostgreSQL. For instance PostgreSQL rejects the following:
create table qxjit (qxjit.foo int);I don't think that our Haskell interface should use Qualified ColumnName in places where PostgreSQL will reject qualified syntax.
| fieldNotEqualsWithAlias :: Maybe Expr.Alias -> FieldDefinition nullability a -> a -> Expr.BooleanExpr | ||
| fieldNotEqualsWithAlias mbAlias = | ||
| whereColumnComparisonWithAlias mbAlias Expr.notEquals | ||
|
|
There was a problem hiding this comment.
I think we should revisit adding WithAlias versions of all these functions. I don't have an immediate suggestion, but having 2 versions of each function seems like a pretty inconvenient API for users
Removing the Qualified in a handful of places where it really should have never been in the first place.
|
Closing this in favor of #352. |
The highlights:
Moving to use 'Qualified ColumnName' in many/most places. This allows for the aliasing to be used naturally in those places.
Add an alias type and function to create a 'Qualified ColumnName' with it. This signals that 'Alias' is in fact different than 'ColumnName' or 'TableName', and reusing them for aliasing is not always appropriate.
Adding a 'MarshallAlias' constructor to the 'SqlMarshaller' so that an alias can be attached to all of the natural fields in a marshaller in a single function call from users. Synthetic fields could be refactored with this facility, but that was left out here to keep the already big changeset smaller.
Many of the fieldDefinition helpers now have a version to support generating various subexpressions with an alias along with the fieldDefinition. This is a choice to allow the simple case to not deal with the complexity of the alias, but make it easier for when aliases are needed.
A small number of helpers are added to make generation of bits of SQL, such as a 'TableReferenceList' that includes the 'AS alias' piece.