Skip to content

Allow alias/materialized columns to be exported #1323

@arthurpassos

Description

@arthurpassos

A bit of context:

Some people have array columns as indexes in ClickHouse.

For instance:

companyIds Array(UInt64)
...
ORDER BY (..., companyIds[5])

Apache Iceberg does not support indexes using arrays like this. Therefore, we need to create an alias, something like:

companyId UInt64 ALIAS companyIds[5]

The problem is that S3 tables don't support alias / materialized and other types of special columns. I don't quite recall the reason. We could investigate, understand why and possibly implement it. But that would be slightly harder.

Instead, we decided to loose up the compatibility check between table structures so that a source table with an alias column is considered compatible to a table that has that column as a physical one.

Metadata

Metadata

Assignees

No one assigned

    Labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions