This is an automated email from the ASF dual-hosted git repository.
wesm pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/arrow.git
The following commit(s) were added to refs/heads/master by this push:
new 09466ce ARROW-3550: [C++] use kUnknownNullCount for the default
null_count argument
09466ce is described below
commit 09466ce0c143c7676c140b4e5cc4c1151056875e
Author: Benjamin Kietzman <[email protected]>
AuthorDate: Thu Mar 7 08:54:20 2019 -0600
ARROW-3550: [C++] use kUnknownNullCount for the default null_count argument
Previous default for this parameter in Array constructors was 0, which is
not correct for all bitmaps the array might've been constructed with.
Sidecar: corrected `TypeTraits<NullType>` to reflect that it is parameter
free and has a type_singleton.
Author: Benjamin Kietzman <[email protected]>
Closes #3781 from bkietz/ARROW-3550-array-constructor-kUnknownNullCount and
squashes the following commits:
feaa3d36d <Benjamin Kietzman> add simple test for default null_count
fbe491572 <Benjamin Kietzman> revert deep copy of ArrayData
6ebaf37fa <Benjamin Kietzman> use kUnknownNullCount for the default
null_count argument
---
cpp/src/arrow/array-test.cc | 4 ++++
cpp/src/arrow/array.h | 28 ++++++++++++++--------------
cpp/src/arrow/type_traits.h | 5 ++++-
3 files changed, 22 insertions(+), 15 deletions(-)
diff --git a/cpp/src/arrow/array-test.cc b/cpp/src/arrow/array-test.cc
index dead8de..f6f66d2 100644
--- a/cpp/src/arrow/array-test.cc
+++ b/cpp/src/arrow/array-test.cc
@@ -73,6 +73,10 @@ TEST_F(TestArray, TestNullCount) {
std::unique_ptr<Int32Array> arr_no_nulls(new Int32Array(100, data));
ASSERT_EQ(0, arr_no_nulls->null_count());
+
+ std::unique_ptr<Int32Array> arr_default_null_count(
+ new Int32Array(100, data, null_bitmap));
+ ASSERT_EQ(kUnknownNullCount, arr_default_null_count->data()->null_count);
}
TEST_F(TestArray, TestLength) {
diff --git a/cpp/src/arrow/array.h b/cpp/src/arrow/array.h
index 35c7bd8..bee133c 100644
--- a/cpp/src/arrow/array.h
+++ b/cpp/src/arrow/array.h
@@ -378,7 +378,7 @@ class ARROW_EXPORT PrimitiveArray : public FlatArray {
PrimitiveArray(const std::shared_ptr<DataType>& type, int64_t length,
const std::shared_ptr<Buffer>& data,
const std::shared_ptr<Buffer>& null_bitmap = NULLPTR,
- int64_t null_count = 0, int64_t offset = 0);
+ int64_t null_count = kUnknownNullCount, int64_t offset = 0);
/// Does not account for any slice offset
std::shared_ptr<Buffer> values() const { return data_->buffers[1]; }
@@ -414,8 +414,8 @@ class NumericArray : public PrimitiveArray {
NumericArray(
typename std::enable_if<TypeTraits<T1>::is_parameter_free,
int64_t>::type length,
const std::shared_ptr<Buffer>& data,
- const std::shared_ptr<Buffer>& null_bitmap = NULLPTR, int64_t null_count
= 0,
- int64_t offset = 0)
+ const std::shared_ptr<Buffer>& null_bitmap = NULLPTR,
+ int64_t null_count = kUnknownNullCount, int64_t offset = 0)
: PrimitiveArray(TypeTraits<T1>::type_singleton(), length, data,
null_bitmap,
null_count, offset) {}
@@ -441,7 +441,7 @@ class ARROW_EXPORT BooleanArray : public PrimitiveArray {
BooleanArray(int64_t length, const std::shared_ptr<Buffer>& data,
const std::shared_ptr<Buffer>& null_bitmap = NULLPTR,
- int64_t null_count = 0, int64_t offset = 0);
+ int64_t null_count = kUnknownNullCount, int64_t offset = 0);
bool Value(int64_t i) const {
return BitUtil::GetBit(reinterpret_cast<const uint8_t*>(raw_values_),
@@ -467,8 +467,8 @@ class ARROW_EXPORT ListArray : public Array {
ListArray(const std::shared_ptr<DataType>& type, int64_t length,
const std::shared_ptr<Buffer>& value_offsets,
const std::shared_ptr<Array>& values,
- const std::shared_ptr<Buffer>& null_bitmap = NULLPTR, int64_t
null_count = 0,
- int64_t offset = 0);
+ const std::shared_ptr<Buffer>& null_bitmap = NULLPTR,
+ int64_t null_count = kUnknownNullCount, int64_t offset = 0);
/// \brief Construct ListArray from array of offsets and child value array
///
@@ -527,7 +527,7 @@ class ARROW_EXPORT BinaryArray : public FlatArray {
BinaryArray(int64_t length, const std::shared_ptr<Buffer>& value_offsets,
const std::shared_ptr<Buffer>& data,
const std::shared_ptr<Buffer>& null_bitmap = NULLPTR,
- int64_t null_count = 0, int64_t offset = 0);
+ int64_t null_count = kUnknownNullCount, int64_t offset = 0);
/// Return the pointer to the given elements bytes
// XXX should GetValue(int64_t i) return a string_view?
@@ -585,7 +585,7 @@ class ARROW_EXPORT BinaryArray : public FlatArray {
const std::shared_ptr<Buffer>& value_offsets,
const std::shared_ptr<Buffer>& data,
const std::shared_ptr<Buffer>& null_bitmap = NULLPTR,
- int64_t null_count = 0, int64_t offset = 0);
+ int64_t null_count = kUnknownNullCount, int64_t offset = 0);
const int32_t* raw_value_offsets_;
const uint8_t* raw_data_;
@@ -601,7 +601,7 @@ class ARROW_EXPORT StringArray : public BinaryArray {
StringArray(int64_t length, const std::shared_ptr<Buffer>& value_offsets,
const std::shared_ptr<Buffer>& data,
const std::shared_ptr<Buffer>& null_bitmap = NULLPTR,
- int64_t null_count = 0, int64_t offset = 0);
+ int64_t null_count = kUnknownNullCount, int64_t offset = 0);
};
// ----------------------------------------------------------------------
@@ -617,7 +617,7 @@ class ARROW_EXPORT FixedSizeBinaryArray : public
PrimitiveArray {
FixedSizeBinaryArray(const std::shared_ptr<DataType>& type, int64_t length,
const std::shared_ptr<Buffer>& data,
const std::shared_ptr<Buffer>& null_bitmap = NULLPTR,
- int64_t null_count = 0, int64_t offset = 0);
+ int64_t null_count = kUnknownNullCount, int64_t offset
= 0);
const uint8_t* GetValue(int64_t i) const;
const uint8_t* Value(int64_t i) const { return GetValue(i); }
@@ -673,8 +673,8 @@ class ARROW_EXPORT StructArray : public Array {
StructArray(const std::shared_ptr<DataType>& type, int64_t length,
const std::vector<std::shared_ptr<Array>>& children,
- std::shared_ptr<Buffer> null_bitmap = NULLPTR, int64_t
null_count = 0,
- int64_t offset = 0);
+ std::shared_ptr<Buffer> null_bitmap = NULLPTR,
+ int64_t null_count = kUnknownNullCount, int64_t offset = 0);
const StructType* struct_type() const;
@@ -712,8 +712,8 @@ class ARROW_EXPORT UnionArray : public Array {
const std::vector<std::shared_ptr<Array>>& children,
const std::shared_ptr<Buffer>& type_ids,
const std::shared_ptr<Buffer>& value_offsets = NULLPTR,
- const std::shared_ptr<Buffer>& null_bitmap = NULLPTR, int64_t
null_count = 0,
- int64_t offset = 0);
+ const std::shared_ptr<Buffer>& null_bitmap = NULLPTR,
+ int64_t null_count = kUnknownNullCount, int64_t offset = 0);
/// \brief Construct Dense UnionArray from types_ids, value_offsets and
children
///
diff --git a/cpp/src/arrow/type_traits.h b/cpp/src/arrow/type_traits.h
index 5d3fed3..0cea584 100644
--- a/cpp/src/arrow/type_traits.h
+++ b/cpp/src/arrow/type_traits.h
@@ -43,7 +43,10 @@ struct TypeTraits<NullType> {
using ArrayType = NullArray;
using BuilderType = NullBuilder;
using ScalarType = NullScalar;
- constexpr static bool is_parameter_free = false;
+
+ static constexpr int64_t bytes_required(int64_t) { return 0; }
+ constexpr static bool is_parameter_free = true;
+ static inline std::shared_ptr<DataType> type_singleton() { return null(); }
};
template <>
