feat: data frame with lazy relation AltrepDataFrameRelation

[Antonov548]

Closes #949

[krlmlr]

I expect the test output to be unchanged? Please also add rel_filter2() .

[krlmlr]

The rapi_() functions are external, this coupling is a bit too tight. Perhaps extract an internal function?

[Antonov548]

Just temporarily

[krlmlr]

Can we move these to a separate .cpp file and keep them in the same order as here?

[krlmlr]

This is still valid. Can you please follow up?

[krlmlr]

I like the idea of rel_from_any_df(), please proceed.

[Antonov548]

@krlmlr seems it works. Maybe it also have to checked with more examples/tests

[krlmlr]

Current state:

drv <- duckdb::duckdb()
con <- DBI::dbConnect(drv)
df1 <- tibble::tibble(a = 1)

"mutate"
#> [1] "mutate"
df2 <- duckdb:::rel_project2(
  df1,
  con,
  list(
    {
      tmp_expr <- duckdb:::expr_reference("a")
      duckdb:::expr_set_alias(tmp_expr, "a")
      tmp_expr
    },
    {
      tmp_expr <- duckdb:::expr_constant(2)
      duckdb:::expr_set_alias(tmp_expr, "b")
      tmp_expr
    }
  )
)
"filter"
#> [1] "filter"
df3 <- duckdb:::rel_filter2(
  df2,
  con,
  list(
    duckdb:::expr_comparison(
      "==",
      list(
        duckdb:::expr_reference("b"),
        duckdb:::expr_constant(2)
      )
    )
  )
)

duckdb:::rel_from_altrep_df(df3)
#> DuckDB Relation: 
#> ---------------------
#> --- Relation Tree ---
#> ---------------------
#> Filter [(b = 2.0)]
#>   Projection [a as a, 2.0 as b]
#>     r_dataframe_scan(0x1349a2ba8)
#> 
#> ---------------------
#> -- Result Columns  --
#> ---------------------
#> - a (DOUBLE)
#> - b (DOUBLE)

# This materializes the data frame
df2
#>   a b
#> 1 1 2

# Expecting this to use a data frame scan only, without a projection
rel2 <- duckdb:::rel_from_altrep_df(df2)
rel2
#> Error in `print()`:
#> ! RebuildRelationException

# Expecting this to use filter only, with a data frame scan based on rel2
rel3 <- duckdb:::rel_from_altrep_df(df3)
rel3
#> Error in `print()`:
#> ! RebuildRelationException

^{Created on 2025-01-10 with reprex v2.1.1}

[krlmlr]

Would that be 255 here and below?

Suggested change

	return StringUtil::EnumToString(GetRelationTypeValues(), 29, "RelationType", static_cast<uint32_t>(value));
	return StringUtil::EnumToString(GetRelationTypeValues(), 255, "RelationType", static_cast<uint32_t>(value));

[krlmlr]

The branch has conflicts. Can you please git merge main -X theirs and then run cpp11::cpp_register() ?