ARROW-8403: [C++] Add ToString() to ChunkedArray, Table and RecordBatch

Closes #6900 from kou/cpp-to-string Authored-by: Sutou Kouhei <kou@clear-code.com> Signed-off-by: Sutou Kouhei <kou@clear-code.com>
apache · Apr 13, 2020 · d0f56ed · d0f56ed
1 parent c6f194a
commit d0f56ed
Show file tree

Hide file tree

Showing 11 changed files with 40 additions and 37 deletions.
diff --git a/c_glib/arrow-glib/basic-array.cpp b/c_glib/arrow-glib/basic-array.cpp
@@ -519,13 +519,7 @@ gchar *
 garrow_array_to_string(GArrowArray *array, GError **error)
 {
   const auto arrow_array = garrow_array_get_raw(array);
-  std::stringstream sink;
-  auto status = arrow::PrettyPrint(*arrow_array, 0, &sink);
-  if (garrow_error_check(error, status, "[array][to-string]")) {
-    return g_strdup(sink.str().c_str());
-  } else {
-    return NULL;
-  }
+  return g_strdup(arrow_array->ToString().c_str());
 }
 
 /**

diff --git a/c_glib/arrow-glib/chunked-array.cpp b/c_glib/arrow-glib/chunked-array.cpp
@@ -328,13 +328,7 @@ gchar *
 garrow_chunked_array_to_string(GArrowChunkedArray *chunked_array, GError **error)
 {
   const auto arrow_chunked_array = garrow_chunked_array_get_raw(chunked_array);
-  std::stringstream sink;
-  auto status = arrow::PrettyPrint(*arrow_chunked_array, 0, &sink);
-  if (garrow_error_check(error, status, "[chunked-array][to-string]")) {
-    return g_strdup(sink.str().c_str());
-  } else {
-    return NULL;
-  }
+  return g_strdup(arrow_chunked_array->ToString().c_str());
 }
 
 G_END_DECLS

diff --git a/c_glib/arrow-glib/decimal128.cpp b/c_glib/arrow-glib/decimal128.cpp
@@ -266,9 +266,8 @@ garrow_decimal128_greater_than_or_equal(GArrowDecimal128 *decimal,
 gchar *
 garrow_decimal128_to_string_scale(GArrowDecimal128 *decimal, gint32 scale)
 {
-  auto arrow_decimal = garrow_decimal128_get_raw(decimal);
-  auto string =  arrow_decimal->ToString(scale);
-  return g_strndup(string.data(), string.size());
+  const auto arrow_decimal = garrow_decimal128_get_raw(decimal);
+  return g_strdup(arrow_decimal->ToString(scale).c_str());
 }
 
 /**
@@ -284,9 +283,8 @@ garrow_decimal128_to_string_scale(GArrowDecimal128 *decimal, gint32 scale)
 gchar *
 garrow_decimal128_to_string(GArrowDecimal128 *decimal)
 {
-  auto arrow_decimal = garrow_decimal128_get_raw(decimal);
-  auto string =  arrow_decimal->ToIntegerString();
-  return g_strndup(string.data(), string.size());
+  const auto arrow_decimal = garrow_decimal128_get_raw(decimal);
+  return g_strdup(arrow_decimal->ToIntegerString().c_str());
 }
 
 /**

diff --git a/c_glib/arrow-glib/file-system.cpp b/c_glib/arrow-glib/file-system.cpp
@@ -366,8 +366,7 @@ gchar *
 garrow_file_info_to_string(GArrowFileInfo *file_info)
 {
   const auto arrow_file_info = garrow_file_info_get_raw(file_info);
-  auto string = arrow_file_info->ToString();
-  return g_strndup(string.data(), string.size());
+  return g_strdup(arrow_file_info->ToString().c_str());
 }
 
 /* arrow::fs::FileSelector */

diff --git a/c_glib/arrow-glib/record-batch.cpp b/c_glib/arrow-glib/record-batch.cpp
@@ -341,13 +341,7 @@ gchar *
 garrow_record_batch_to_string(GArrowRecordBatch *record_batch, GError **error)
 {
   const auto arrow_record_batch = garrow_record_batch_get_raw(record_batch);
-  std::stringstream sink;
-  auto status = arrow::PrettyPrint(*arrow_record_batch, 0, &sink);
-  if (garrow_error_check(error, status, "[record-batch][to-string]")) {
-    return g_strdup(sink.str().c_str());
-  } else {
-    return NULL;
-  }
+  return g_strdup(arrow_record_batch->ToString().c_str());
 }
 
 /**

diff --git a/c_glib/arrow-glib/table.cpp b/c_glib/arrow-glib/table.cpp
@@ -541,13 +541,7 @@ gchar *
 garrow_table_to_string(GArrowTable *table, GError **error)
 {
   const auto arrow_table = garrow_table_get_raw(table);
-  std::stringstream sink;
-  auto status = arrow::PrettyPrint(*arrow_table, 0, &sink);
-  if (garrow_error_check(error, status, "[table][to-string]")) {
-    return g_strdup(sink.str().c_str());
-  } else {
-    return NULL;
-  }
+  return g_strdup(arrow_table->ToString().c_str());
 }
 
 /**

diff --git a/c_glib/test/test-record-batch.rb b/c_glib/test/test-record-batch.rb
@@ -76,7 +76,7 @@ def setup
           Arrow::Field.new("visible", Arrow::BooleanDataType.new),
           Arrow::Field.new("valid", Arrow::BooleanDataType.new),
         ]
-        @schema = Arrow::Schema.new(fields)
+        @schema = Arrow::Schema.new(@fields)
         @columns = [
           build_boolean_array([true, false, true, false, true]),
           build_boolean_array([false, true, false, true, false]),

diff --git a/cpp/src/arrow/record_batch.cc b/cpp/src/arrow/record_batch.cc
@@ -21,11 +21,13 @@
 #include <atomic>
 #include <cstdlib>
 #include <memory>
+#include <sstream>
 #include <string>
 #include <utility>
 
 #include "arrow/array.h"
 #include "arrow/array/validate.h"
+#include "arrow/pretty_print.h"
 #include "arrow/status.h"
 #include "arrow/table.h"
 #include "arrow/type.h"
@@ -246,6 +248,12 @@ std::shared_ptr<RecordBatch> RecordBatch::Slice(int64_t offset) const {
   return Slice(offset, this->num_rows() - offset);
 }
 
+std::string RecordBatch::ToString() const {
+  std::stringstream ss;
+  ARROW_CHECK_OK(PrettyPrint(*this, 0, &ss));
+  return ss.str();
+}
+
 Status RecordBatch::Validate() const {
   for (int i = 0; i < num_columns(); ++i) {
     const auto& array = *this->column(i);

diff --git a/cpp/src/arrow/record_batch.h b/cpp/src/arrow/record_batch.h
@@ -177,6 +177,9 @@ class ARROW_EXPORT RecordBatch {
   /// \return new record batch
   virtual std::shared_ptr<RecordBatch> Slice(int64_t offset, int64_t length) const = 0;
 
+  /// \return PrettyPrint representation suitable for debugging
+  std::string ToString() const;
+
   /// \brief Perform cheap validation checks to determine obvious inconsistencies
   /// within the record batch's schema and internal data.
   ///

diff --git a/cpp/src/arrow/table.cc b/cpp/src/arrow/table.cc
@@ -27,6 +27,7 @@
 #include "arrow/array.h"
 #include "arrow/array/concatenate.h"
 #include "arrow/array/validate.h"
+#include "arrow/pretty_print.h"
 #include "arrow/record_batch.h"
 #include "arrow/status.h"
 #include "arrow/type.h"
@@ -176,6 +177,12 @@ Status ChunkedArray::View(const std::shared_ptr<DataType>& type,
   return View(type).Value(out);
 }
 
+std::string ChunkedArray::ToString() const {
+  std::stringstream ss;
+  ARROW_CHECK_OK(PrettyPrint(*this, 0, &ss));
+  return ss.str();
+}
+
 Status ChunkedArray::Validate() const {
   if (chunks_.size() == 0) {
     return Status::OK();
@@ -599,6 +606,12 @@ Status Table::Flatten(MemoryPool* pool, std::shared_ptr<Table>* out) const {
   return Flatten(pool).Value(out);
 }
 
+std::string Table::ToString() const {
+  std::stringstream ss;
+  ARROW_CHECK_OK(PrettyPrint(*this, 0, &ss));
+  return ss.str();
+}
+
 Result<std::shared_ptr<Table>> ConcatenateTables(
     const std::vector<std::shared_ptr<Table>>& tables,
     const ConcatenateTablesOptions options, MemoryPool* memory_pool) {

diff --git a/cpp/src/arrow/table.h b/cpp/src/arrow/table.h
@@ -107,6 +107,9 @@ class ARROW_EXPORT ChunkedArray {
   /// \brief Determine if two chunked arrays are equal.
   bool Equals(const std::shared_ptr<ChunkedArray>& other) const;
 
+  /// \return PrettyPrint representation suitable for debugging
+  std::string ToString() const;
+
   /// \brief Perform cheap validation checks to determine obvious inconsistencies
   /// within the chunk array's internal data.
   ///
@@ -366,6 +369,9 @@ class ARROW_EXPORT Table {
   ARROW_DEPRECATED("Use Result-returning version")
   Status Flatten(MemoryPool* pool, std::shared_ptr<Table>* out) const;
 
+  /// \return PrettyPrint representation suitable for debugging
+  std::string ToString() const;
+
   /// \brief Perform cheap validation checks to determine obvious inconsistencies
   /// within the table's schema and internal data.
   ///