Remove TypedArrayStatistics

kou · Jul 12, 2024 · cb2cb97 · cb2cb97
1 parent 8b02c51
commit cb2cb97
Show file tree

Hide file tree

Showing 5 changed files with 13 additions and 49 deletions.
diff --git a/cpp/src/arrow/array/array_primitive.h b/cpp/src/arrow/array/array_primitive.h
@@ -70,11 +70,6 @@ class ARROW_EXPORT BooleanArray : public PrimitiveArray {
 
   IteratorType end() const { return IteratorType(*this, length()); }
 
-  /// \brief Return the statistics for boolean.
-  std::shared_ptr<BooleanArrayStatistics> statistics() const {
-    return std::static_pointer_cast<BooleanArrayStatistics>(statistics_);
-  }
-
  protected:
   using PrimitiveArray::PrimitiveArray;
 
@@ -130,11 +125,6 @@ class NumericArray : public PrimitiveArray {
 
   IteratorType end() const { return IteratorType(*this, length()); }
 
-  /// \brief Return the typed statistics.
-  std::shared_ptr<TypedArrayStatistics<TYPE>> statistics() const {
-    return std::static_pointer_cast<TypedArrayStatistics<TYPE>>(Array::statistics());
-  }
-
  protected:
   using PrimitiveArray::PrimitiveArray;
 };

diff --git a/cpp/src/arrow/array/statistics.h b/cpp/src/arrow/array/statistics.h
@@ -63,29 +63,4 @@ struct ARROW_EXPORT ArrayStatistics {
   }
 };
 
-/// \brief A typed implementation of ArrayStatistics
-template <typename TypeClass>
-class TypedArrayStatistics : public ArrayStatistics {
- public:
-  using ElementType = typename TypeClass::c_type;
-
-  /// \brief The current minimum value, may not be set
-  std::optional<ElementType> min() const {
-    if (min_buffer && std::holds_alternative<ElementType>(*min_buffer)) {
-      return std::get<ElementType>(*min_buffer);
-    } else {
-      return std::nullopt;
-    }
-  }
-
-  /// \brief The current maximum value, may not be set
-  std::optional<ElementType> max() const {
-    if (max_buffer && std::holds_alternative<ElementType>(*max_buffer)) {
-      return std::get<ElementType>(*max_buffer);
-    } else {
-      return std::nullopt;
-    }
-  }
-};
-
 }  // namespace arrow
diff --git a/cpp/src/arrow/type_fwd.h b/cpp/src/arrow/type_fwd.h
@@ -89,9 +89,6 @@ using ChunkedArrayVector = std::vector<std::shared_ptr<ChunkedArray>>;
 using RecordBatchVector = std::vector<std::shared_ptr<RecordBatch>>;
 using RecordBatchIterator = Iterator<std::shared_ptr<RecordBatch>>;
 
-template <typename TypeClass>
-class TypedArrayStatistics;
-
 class DictionaryType;
 class DictionaryArray;
 struct DictionaryScalar;
@@ -105,7 +102,6 @@ class FixedWidthType;
 
 class BooleanType;
 class BooleanArray;
-using BooleanArrayStatistics = TypedArrayStatistics<BooleanType>;
 class BooleanBuilder;
 struct BooleanScalar;
 
@@ -222,7 +218,6 @@ class NumericTensor;
 #define _NUMERIC_TYPE_DECL(KLASS)                                   \
   class KLASS##Type;                                                \
   using KLASS##Array = NumericArray<KLASS##Type>;                   \
-  using KLASS##ArrayStatistics = TypedArrayStatistics<KLASS##Type>; \
   using KLASS##Builder = NumericBuilder<KLASS##Type>;               \
   struct KLASS##Scalar;                                             \
   using KLASS##Tensor = NumericTensor<KLASS##Type>;

diff --git a/cpp/src/parquet/arrow/arrow_statistics_test.cc b/cpp/src/parquet/arrow/arrow_statistics_test.cc
@@ -190,12 +190,14 @@ TEST(TestStatisticsRead, Boolean) {
   ASSERT_EQ(true, statistics->null_count.has_value());
   ASSERT_EQ(1, statistics->null_count.value());
   ASSERT_EQ(false, statistics->distinct_count.has_value());
-  ASSERT_EQ(true, statistics->min().has_value());
-  ASSERT_EQ(true, statistics->min().value());
+  ASSERT_EQ(true, statistics->min_buffer.has_value());
+  ASSERT_EQ(true, std::holds_alternative<bool>(*statistics->min_buffer));
+  ASSERT_EQ(true, std::get<bool>(*statistics->min_buffer));
   ASSERT_EQ(true, statistics->is_min_exact.has_value());
   ASSERT_EQ(true, statistics->is_min_exact.value());
-  ASSERT_EQ(true, statistics->max().has_value());
-  ASSERT_EQ(true, statistics->max().value());
+  ASSERT_EQ(true, statistics->max_buffer.has_value());
+  ASSERT_EQ(true, std::holds_alternative<bool>(*statistics->max_buffer));
+  ASSERT_EQ(true, std::get<bool>(*statistics->max_buffer));
   ASSERT_EQ(true, statistics->is_min_exact.has_value());
   ASSERT_EQ(true, statistics->is_min_exact.value());
 }
@@ -208,12 +210,14 @@ TEST(TestStatisticsRead, Int8) {
   ASSERT_EQ(true, statistics->null_count.has_value());
   ASSERT_EQ(1, statistics->null_count.value());
   ASSERT_EQ(false, statistics->distinct_count.has_value());
-  ASSERT_EQ(true, statistics->min().has_value());
-  ASSERT_EQ(-1, statistics->min().value());
+  ASSERT_EQ(true, statistics->min_buffer.has_value());
+  ASSERT_EQ(true, std::holds_alternative<int8_t>(*statistics->min_buffer));
+  ASSERT_EQ(-1, std::get<int8_t>(*statistics->min_buffer));
   ASSERT_EQ(true, statistics->is_min_exact.has_value());
   ASSERT_EQ(true, statistics->is_min_exact.value());
-  ASSERT_EQ(true, statistics->max().has_value());
-  ASSERT_EQ(1, statistics->max().value());
+  ASSERT_EQ(true, statistics->max_buffer.has_value());
+  ASSERT_EQ(true, std::holds_alternative<int8_t>(*statistics->max_buffer));
+  ASSERT_EQ(1, std::get<int8_t>(*statistics->max_buffer));
   ASSERT_EQ(true, statistics->is_min_exact.has_value());
   ASSERT_EQ(true, statistics->is_min_exact.value());
 }

diff --git a/cpp/src/parquet/arrow/reader_internal.cc b/cpp/src/parquet/arrow/reader_internal.cc
@@ -404,7 +404,7 @@ Status TransferBool(RecordReader* reader,
   }
   auto array_data = ::arrow::ArrayData::Make(::arrow::boolean(), length,
                                              std::move(buffers), null_count);
-  auto array_statistics = std::make_shared<::arrow::BooleanArrayStatistics>();
+  auto array_statistics = std::make_shared<::arrow::ArrayStatistics>();
   array_statistics->null_count = null_count;
   auto statistics = metadata->statistics().get();
   if (statistics) {