Skip to content

Specialised data types in R #22

@nbenn

Description

@nbenn

DBI should offer some type of plugin system that other packages can build upon by offering implementations for representing some of the more exotic data types in R. Throughout the DBI packages, there are many open issues surrounding this problem. A selection:

r-dbi/DBI#199 enum types
r-dbi/DBI#179 uuid types
r-dbi/RPostgres#132 geometry types, sf
r-dbi/RPostgres#114 geometry types, sf
r-dbi/RPostgres#86 geometry types
r-dbi/bigrquery#145 array/struct types
r-dbi/RMariaDB#42 json

other types that fall into this area include arbitrary precision floats (Rmpfr, gmp), xml types (xml2), monetary types, etc.

Now if geometry types are implemented for Postgres, this is great. But they are also available in MySQL/MariaDB. It therefore might be useful to consider these issues in a more general fashion. Furthermore, approaching this in a type by type fashion might not be sufficient. How could a user map a Postgres composite type, if there is not some inherent extensibility?

Unfortunately, I have no idea how to tackle such an issue. Maybe a pragmatic approach, where things such as composite types are simply not considered, is the best we can do. I just was hoping to get a discussion started on this topic.

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions