initial stab at arrow::table(data.frame)

romainfrancois · romainfrancois · commit 79205fbaa7d5 · 2018-09-24T20:15:02.000+02:00
diff --git a/r/NAMESPACE b/r/NAMESPACE
@@ -24,6 +24,7 @@ export(list_of)
 export(null)
 export(schema)
 export(struct)
+export(table)
 export(time32)
 export(time64)
 export(timestamp)
diff --git a/r/R/RcppExports.R b/r/R/RcppExports.R
@@ -85,6 +85,22 @@ RecordBatch_schema <- function(x) {
     .Call(`_arrow_RecordBatch_schema`, x)
 }
 
+dataframe_to_Table <- function(tbl) {
+    .Call(`_arrow_dataframe_to_Table`, tbl)
+}
+
+Table_num_columns <- function(x) {
+    .Call(`_arrow_Table_num_columns`, x)
+}
+
+Table_num_rows <- function(x) {
+    .Call(`_arrow_Table_num_rows`, x)
+}
+
+Table_schema <- function(x) {
+    .Call(`_arrow_Table_schema`, x)
+}
+
 Field_initialize <- function(name, type, nullable = TRUE) {
     .Call(`_arrow_Field_initialize`, name, type, nullable)
 }
diff --git a/r/R/array.R b/r/R/array.R
@@ -109,3 +109,23 @@ array <- function(...){
 record_batch <- function(.data){
   `arrow::RecordBatch`$new(.data)
 }
+
+`arrow::Table` <- R6Class("arrow::Table", inherit = `arrow::Object`,
+  public = list(
+    initialize = function(.data){
+      self$set_pointer(dataframe_to_Table(.data))
+    },
+    num_columns = function() Table_num_columns(self),
+    num_rows = function() Table_num_rows(self),
+    schema = function() schema(.xp = Table_schema(self))
+  )
+)
+
+#' Create an arrow::Table from a data frame
+#'
+#' @param .data a data frame
+#'
+#' @export
+table <- function(.data){
+  `arrow::Table`$new(.data)
+}
diff --git a/r/man/table.Rd b/r/man/table.Rd
diff --git a/r/src/RcppExports.cpp b/r/src/RcppExports.cpp
@@ -244,6 +244,50 @@ BEGIN_RCPP
     return rcpp_result_gen;
 END_RCPP
 }
+// dataframe_to_Table
+std::shared_ptr<arrow::Table> dataframe_to_Table(DataFrame tbl);
+RcppExport SEXP _arrow_dataframe_to_Table(SEXP tblSEXP) {
+BEGIN_RCPP
+    Rcpp::RObject rcpp_result_gen;
+    Rcpp::RNGScope rcpp_rngScope_gen;
+    Rcpp::traits::input_parameter< DataFrame >::type tbl(tblSEXP);
+    rcpp_result_gen = Rcpp::wrap(dataframe_to_Table(tbl));
+    return rcpp_result_gen;
+END_RCPP
+}
+// Table_num_columns
+int Table_num_columns(const std::shared_ptr<arrow::Table>& x);
+RcppExport SEXP _arrow_Table_num_columns(SEXP xSEXP) {
+BEGIN_RCPP
+    Rcpp::RObject rcpp_result_gen;
+    Rcpp::RNGScope rcpp_rngScope_gen;
+    Rcpp::traits::input_parameter< const std::shared_ptr<arrow::Table>& >::type x(xSEXP);
+    rcpp_result_gen = Rcpp::wrap(Table_num_columns(x));
+    return rcpp_result_gen;
+END_RCPP
+}
+// Table_num_rows
+int Table_num_rows(const std::shared_ptr<arrow::Table>& x);
+RcppExport SEXP _arrow_Table_num_rows(SEXP xSEXP) {
+BEGIN_RCPP
+    Rcpp::RObject rcpp_result_gen;
+    Rcpp::RNGScope rcpp_rngScope_gen;
+    Rcpp::traits::input_parameter< const std::shared_ptr<arrow::Table>& >::type x(xSEXP);
+    rcpp_result_gen = Rcpp::wrap(Table_num_rows(x));
+    return rcpp_result_gen;
+END_RCPP
+}
+// Table_schema
+std::shared_ptr<arrow::Schema> Table_schema(const std::shared_ptr<arrow::Table>& x);
+RcppExport SEXP _arrow_Table_schema(SEXP xSEXP) {
+BEGIN_RCPP
+    Rcpp::RObject rcpp_result_gen;
+    Rcpp::RNGScope rcpp_rngScope_gen;
+    Rcpp::traits::input_parameter< const std::shared_ptr<arrow::Table>& >::type x(xSEXP);
+    rcpp_result_gen = Rcpp::wrap(Table_schema(x));
+    return rcpp_result_gen;
+END_RCPP
+}
 // Field_initialize
 std::shared_ptr<arrow::Field> Field_initialize(const std::string& name, const std::shared_ptr<arrow::DataType>& type, bool nullable);
 RcppExport SEXP _arrow_Field_initialize(SEXP nameSEXP, SEXP typeSEXP, SEXP nullableSEXP) {
@@ -1550,6 +1594,10 @@ static const R_CallMethodDef CallEntries[] = {
     {"_arrow_RecordBatch_num_columns", (DL_FUNC) &_arrow_RecordBatch_num_columns, 1},
     {"_arrow_RecordBatch_num_rows", (DL_FUNC) &_arrow_RecordBatch_num_rows, 1},
     {"_arrow_RecordBatch_schema", (DL_FUNC) &_arrow_RecordBatch_schema, 1},
+    {"_arrow_dataframe_to_Table", (DL_FUNC) &_arrow_dataframe_to_Table, 1},
+    {"_arrow_Table_num_columns", (DL_FUNC) &_arrow_Table_num_columns, 1},
+    {"_arrow_Table_num_rows", (DL_FUNC) &_arrow_Table_num_rows, 1},
+    {"_arrow_Table_schema", (DL_FUNC) &_arrow_Table_schema, 1},
     {"_arrow_Field_initialize", (DL_FUNC) &_arrow_Field_initialize, 3},
     {"_arrow_Field_ToString", (DL_FUNC) &_arrow_Field_ToString, 1},
     {"_arrow_Field_name", (DL_FUNC) &_arrow_Field_name, 1},
diff --git a/r/src/buffer.cpp b/r/src/buffer.cpp
@@ -111,3 +111,27 @@ int RecordBatch_num_rows(const std::shared_ptr<arrow::RecordBatch>& x){
 std::shared_ptr<arrow::Schema> RecordBatch_schema(const std::shared_ptr<arrow::RecordBatch>& x){
   return x->schema();
 }
+
+// [[Rcpp::export]]
+std::shared_ptr<arrow::Table> dataframe_to_Table(DataFrame tbl){
+  auto rb = dataframe_to_RecordBatch(tbl);
+
+  std::shared_ptr<arrow::Table> out;
+  auto status = arrow::Table::FromRecordBatches({ std::move(rb) }, &out);
+  return out;
+}
+
+// [[Rcpp::export]]
+int Table_num_columns(const std::shared_ptr<arrow::Table>& x){
+  return x->num_columns();
+}
+
+// [[Rcpp::export]]
+int Table_num_rows(const std::shared_ptr<arrow::Table>& x){
+  return x->num_rows();
+}
+
+// [[Rcpp::export]]
+std::shared_ptr<arrow::Schema> Table_schema(const std::shared_ptr<arrow::Table>& x){
+  return x->schema();
+}
