Simply put, code similar to this should compile: rawDf.select { “str”<Int>() }.str. I.e. after a column is references by a String API, it's implied such column should exist => plugin updates the schema accordingly.
Goal here is to provide workflow alternative to defining full @DataSchema for AnyFrame. Instead, sometimes information from a few "String API" calls can be enough to migrate to typesafe API. It enables "incremental migration".
Related change in the library:
#1666, #1659, #1667
It all needs to be documented with its own page: an example of a moderately sized pipeline, few shorter examples with CS DSL operators that are recommended to use
Simply put, code similar to this should compile:
rawDf.select { “str”<Int>() }.str. I.e. after a column is references by a String API, it's implied such column should exist => plugin updates the schema accordingly.Goal here is to provide workflow alternative to defining full
@DataSchemaforAnyFrame. Instead, sometimes information from a few "String API" calls can be enough to migrate to typesafe API. It enables "incremental migration".Related change in the library:
#1666, #1659, #1667
It all needs to be documented with its own page: an example of a moderately sized pipeline, few shorter examples with CS DSL operators that are recommended to use