rok commented on code in PR #14341:
URL: https://github.com/apache/arrow/pull/14341#discussion_r1157748887


##########
cpp/src/parquet/encoding_test.cc:
##########
@@ -1955,5 +1955,285 @@ TEST(DeltaLengthByteArrayEncodingAdHoc, ArrowDirectPut) 
{
   CheckDecode(encoded, ::arrow::ArrayFromJSON(::arrow::large_binary(), 
values));
 }
 
+// ----------------------------------------------------------------------
+// DELTA_BYTE_ARRAY encode/decode tests.
+
+template <typename Type>
+class TestDeltaByteArrayEncoding : public TestEncodingBase<Type> {
+ public:
+  using c_type = typename Type::c_type;
+  static constexpr int TYPE = Type::type_num;
+
+  virtual void CheckRoundtrip() {
+    auto encoder =
+        MakeTypedEncoder<Type>(Encoding::DELTA_BYTE_ARRAY, false, 
descr_.get());
+    auto decoder = MakeTypedDecoder<Type>(Encoding::DELTA_BYTE_ARRAY, 
descr_.get());
+
+    encoder->Put(draws_, num_values_);
+    encode_buffer_ = encoder->FlushValues();
+
+    decoder->SetData(num_values_, encode_buffer_->data(),
+                     static_cast<int>(encode_buffer_->size()));
+    int values_decoded = decoder->Decode(decode_buf_, num_values_);
+    ASSERT_EQ(num_values_, values_decoded);
+    ASSERT_NO_FATAL_FAILURE(VerifyResults<c_type>(decode_buf_, draws_, 
num_values_));
+  }
+
+  void CheckRoundtripSpaced(const uint8_t* valid_bits, int64_t 
valid_bits_offset) {
+    auto encoder =
+        MakeTypedEncoder<Type>(Encoding::DELTA_BYTE_ARRAY, false, 
descr_.get());
+    auto decoder = MakeTypedDecoder<Type>(Encoding::DELTA_BYTE_ARRAY, 
descr_.get());
+    int null_count = 0;
+    for (auto i = 0; i < num_values_; i++) {
+      if (!bit_util::GetBit(valid_bits, valid_bits_offset + i)) {
+        null_count++;
+      }
+    }
+
+    encoder->PutSpaced(draws_, num_values_, valid_bits, valid_bits_offset);
+    encode_buffer_ = encoder->FlushValues();
+    decoder->SetData(num_values_ - null_count, encode_buffer_->data(),
+                     static_cast<int>(encode_buffer_->size()));
+    auto values_decoded = decoder->DecodeSpaced(decode_buf_, num_values_, 
null_count,
+                                                valid_bits, valid_bits_offset);
+    ASSERT_EQ(num_values_, values_decoded);
+    ASSERT_NO_FATAL_FAILURE(VerifyResultsSpaced<c_type>(decode_buf_, draws_, 
num_values_,
+                                                        valid_bits, 
valid_bits_offset));
+  }
+
+ protected:
+  USING_BASE_MEMBERS();
+};
+
+typedef ::testing::Types<ByteArrayType> TestDeltaByteArrayEncodingTypes;
+TYPED_TEST_SUITE(TestDeltaByteArrayEncoding, TestDeltaByteArrayEncodingTypes);
+
+// TODO: add FLBAType and Decimal type tests
+
+TYPED_TEST(TestDeltaByteArrayEncoding, BasicRoundTrip) {
+  ASSERT_NO_FATAL_FAILURE(this->Execute(0, 0));
+  ASSERT_NO_FATAL_FAILURE(this->Execute(250, 2));
+  ASSERT_NO_FATAL_FAILURE(this->ExecuteSpaced(
+      /*nvalues*/ 1234, /*repeats*/ 1, /*valid_bits_offset*/ 64, /*null_prob*/ 
0));
+
+  ASSERT_NO_FATAL_FAILURE(this->Execute(2000, 200));
+  ASSERT_NO_FATAL_FAILURE(this->ExecuteSpaced(
+      /*nvalues*/ 1234, /*repeats*/ 10, /*valid_bits_offset*/ 64,
+      /*null_probability*/ 0.1));
+}
+
+TEST(DeltaByteArrayEncodingAdHoc, ArrowBinaryDirectPut) {
+  const int64_t size = 50;
+  const int32_t min_length = 0;
+  const int32_t max_length = 10;
+  const int32_t num_unique = 10;
+  const double null_probability = 0.25;
+  auto encoder = MakeTypedEncoder<ByteArrayType>(Encoding::DELTA_BYTE_ARRAY);
+  auto decoder = MakeTypedDecoder<ByteArrayType>(Encoding::DELTA_BYTE_ARRAY);
+
+  auto CheckSeed = [&](std::shared_ptr<::arrow::Array> values) {
+    ASSERT_NO_THROW(encoder->Put(*values));
+    auto buf = encoder->FlushValues();
+
+    int num_values = static_cast<int>(values->length() - values->null_count());
+    decoder->SetData(num_values, buf->data(), static_cast<int>(buf->size()));
+
+    typename EncodingTraits<ByteArrayType>::Accumulator acc;
+    if (::arrow::is_string(values->type()->id())) {
+      acc.builder = std::make_unique<::arrow::StringBuilder>();
+    } else {
+      acc.builder = std::make_unique<::arrow::BinaryBuilder>();
+    }
+    ASSERT_EQ(num_values,
+              decoder->DecodeArrow(static_cast<int>(values->length()),
+                                   static_cast<int>(values->null_count()),
+                                   values->null_bitmap_data(), 
values->offset(), &acc));
+
+    std::shared_ptr<::arrow::Array> result;
+    ASSERT_OK(acc.builder->Finish(&result));
+    ASSERT_EQ(values->length(), result->length());
+    ASSERT_OK(result->ValidateFull());
+
+    auto upcast_result = CastBinaryTypesHelper(result, values->type());
+    ::arrow::AssertArraysEqual(*values, *result);
+  };
+
+  ::arrow::random::RandomArrayGenerator rag(42);
+  auto values = rag.String(0, min_length, max_length, null_probability);
+  CheckSeed(values);
+  for (auto seed : {0, 1, 2, 3, 4, 5, 6, 7, 8, 9}) {
+    rag = ::arrow::random::RandomArrayGenerator(seed);
+
+    values = rag.String(size, min_length, max_length, null_probability);
+    CheckSeed(values);
+
+    values =
+        rag.BinaryWithRepeats(size, num_unique, min_length, max_length, 
null_probability);
+    CheckSeed(values);
+  }
+}
+
+TEST(DeltaByteArrayEncodingAdHoc, ArrowBinaryDirectPutFixedLength) {
+  const int64_t size = 50;
+  const double null_probability = 0.25;
+  ::arrow::random::RandomArrayGenerator rag(0);
+  auto encoder = MakeTypedEncoder<FLBAType>(Encoding::DELTA_BYTE_ARRAY);
+  auto decoder = MakeTypedDecoder<FLBAType>(Encoding::DELTA_BYTE_ARRAY);
+
+  auto CheckSeed = [&](std::shared_ptr<::arrow::Array> values) {
+    ASSERT_NO_THROW(encoder->Put(*values));
+    auto buf = encoder->FlushValues();
+
+    int num_values = static_cast<int>(values->length() - values->null_count());
+    decoder->SetData(num_values, buf->data(), static_cast<int>(buf->size()));
+
+    typename EncodingTraits<FLBAType>::Accumulator acc(values->type());
+    ASSERT_EQ(num_values,
+              decoder->DecodeArrow(static_cast<int>(values->length()),
+                                   static_cast<int>(values->null_count()),
+                                   values->null_bitmap_data(), 
values->offset(), &acc));
+
+    std::shared_ptr<::arrow::Array> result;
+    ASSERT_OK(acc.Finish(&result));
+    ASSERT_EQ(values->length(), result->length());
+    ASSERT_OK(result->ValidateFull());
+    ::arrow::AssertArraysEqual(*values, *result);
+  };
+
+  for (auto seed : {0, 1, 2, 3, 4, 5, 6, 7, 8, 9}) {
+    for (auto length : {0, 10, 100, 1000}) {
+      rag = ::arrow::random::RandomArrayGenerator(seed);
+      auto values = rag.FixedSizeBinary(size, length, null_probability);
+      CheckSeed(values);
+    }
+  }
+}
+
+TEST(DeltaByteArrayEncodingAdHoc, ArrowDirectPut) {
+  auto CheckEncode = [](std::shared_ptr<::arrow::Array> values,
+                        std::shared_ptr<::arrow::Array> prefix_lengths,
+                        std::shared_ptr<::arrow::Array> suffix_lengths,
+                        std::string_view value) {
+    auto encoder = MakeTypedEncoder<ByteArrayType>(Encoding::DELTA_BYTE_ARRAY);
+    ASSERT_NO_THROW(encoder->Put(*values));
+    auto buf = encoder->FlushValues();
+
+    auto prefix_lengths_encoder =
+        MakeTypedEncoder<Int32Type>(Encoding::DELTA_BINARY_PACKED);
+    ASSERT_NO_THROW(prefix_lengths_encoder->Put(*prefix_lengths));
+    auto prefix_lengths_buf = prefix_lengths_encoder->FlushValues();
+
+    auto encoded_prefix_lengths_buf = SliceBuffer(buf, 0, 
prefix_lengths_buf->size());
+
+    auto suffix_lengths_encoder =
+        MakeTypedEncoder<Int32Type>(Encoding::DELTA_BINARY_PACKED);
+    ASSERT_NO_THROW(suffix_lengths_encoder->Put(*suffix_lengths));
+    auto suffix_lengths_buf = suffix_lengths_encoder->FlushValues();
+    auto encoded_values_buf =
+        SliceBuffer(buf, prefix_lengths_buf->size() + 
suffix_lengths_buf->size());
+
+    auto encoded_prefix_length_buf = SliceBuffer(buf, 0, 
prefix_lengths_buf->size());
+    EXPECT_TRUE(prefix_lengths_buf->Equals(*encoded_prefix_length_buf));
+    auto encoded_suffix_length_buf =
+        SliceBuffer(buf, prefix_lengths_buf->size(), 
suffix_lengths_buf->size());
+    EXPECT_TRUE(suffix_lengths_buf->Equals(*encoded_suffix_length_buf));
+    EXPECT_EQ(value, encoded_values_buf->ToString());
+  };
+
+  auto arrayToI32 = [](const std::shared_ptr<::arrow::Array>& lengths) {
+    std::vector<int32_t> arrays;
+    auto data_ptr = checked_cast<::arrow::Int32Array*>(lengths.get());
+    for (int i = 0; i < lengths->length(); ++i) {
+      arrays.push_back(data_ptr->GetView(i));
+    }
+    return arrays;
+  };
+
+  auto CheckDecode = [](std::shared_ptr<Buffer> buf,
+                        std::shared_ptr<::arrow::Array> values) {
+    int num_values = static_cast<int>(values->length());
+    auto decoder = MakeTypedDecoder<ByteArrayType>(Encoding::DELTA_BYTE_ARRAY);
+    decoder->SetData(num_values, buf->data(), static_cast<int>(buf->size()));
+
+    typename EncodingTraits<ByteArrayType>::Accumulator acc;
+    if (::arrow::is_string(values->type()->id())) {
+      acc.builder = std::make_unique<::arrow::StringBuilder>();
+    } else {
+      acc.builder = std::make_unique<::arrow::BinaryBuilder>();
+    }
+
+    ASSERT_EQ(num_values,
+              decoder->DecodeArrow(static_cast<int>(values->length()),
+                                   static_cast<int>(values->null_count()),
+                                   values->null_bitmap_data(), 
values->offset(), &acc));
+
+    std::shared_ptr<::arrow::Array> result;
+    ASSERT_OK(acc.builder->Finish(&result));
+    ASSERT_EQ(num_values, result->length());
+    ASSERT_OK(result->ValidateFull());
+
+    auto upcast_result = CastBinaryTypesHelper(result, values->type());
+    ::arrow::AssertArraysEqual(*values, *upcast_result);
+  };
+
+  auto checkEncodeDecode = [&](std::string_view values,
+                               std::shared_ptr<::arrow::Array> prefix_lengths,
+                               std::shared_ptr<::arrow::Array> suffix_lengths,
+                               std::string_view suffix_data) {
+    CheckEncode(::arrow::ArrayFromJSON(::arrow::utf8(), values), 
prefix_lengths,
+                suffix_lengths, suffix_data);
+    CheckEncode(::arrow::ArrayFromJSON(::arrow::large_utf8(), values), 
prefix_lengths,
+                suffix_lengths, suffix_data);
+    CheckEncode(::arrow::ArrayFromJSON(::arrow::binary(), values), 
prefix_lengths,
+                suffix_lengths, suffix_data);
+    CheckEncode(::arrow::ArrayFromJSON(::arrow::large_binary(), values), 
prefix_lengths,
+                suffix_lengths, suffix_data);
+
+    auto encoded = 
::arrow::ConcatenateBuffers({DeltaEncode(arrayToI32(prefix_lengths)),
+                                                
DeltaEncode(arrayToI32(suffix_lengths)),
+                                                
std::make_shared<Buffer>(suffix_data)})
+                       .ValueOrDie();
+
+    CheckDecode(encoded, ::arrow::ArrayFromJSON(::arrow::utf8(), values));
+    CheckDecode(encoded, ::arrow::ArrayFromJSON(::arrow::large_utf8(), 
values));
+    CheckDecode(encoded, ::arrow::ArrayFromJSON(::arrow::binary(), values));
+    CheckDecode(encoded, ::arrow::ArrayFromJSON(::arrow::large_binary(), 
values));
+  };
+
+  {
+    auto values = R"(["axis", "axle", "babble", "babyhood"])";

Review Comment:
   I'm trying this example but encoded result currently doesn't match up:
   ```cpp
     {
       auto values = R"(["καλημέρα", "καμηλιέρη", "καμηλιέρη", "καλημέρα"])";
       auto prefix_lengths = ::arrow::ArrayFromJSON(::arrow::int32(), R"([0, 2, 
9, 2])");
       auto suffix_lengths = ::arrow::ArrayFromJSON(::arrow::int32(), R"([8, 7, 
0, 6])");
       constexpr std::string_view suffix_data = "καλημέραμηλιέρηλημέρα";
       CheckEncodeDecode(values, prefix_lengths, suffix_lengths, suffix_data);
     }
   ```
   
   ```
   Value of: prefix_lengths_buf->Equals(*encoded_prefix_length_buf)
     Actual: false
   Expected: true
   
   Value of: suffix_lengths_buf->Equals(*encoded_suffix_length_buf)
     Actual: false
   Expected: true
   
   Expected equality of these values:
     value
       Which is: 
"\xCE\xBA\xCE\xB1\xCE\xBB\xCE\xB7\xCE\xBC\xCE\xAD\xCF\x81\xCE\xB1\xCE\xBC\xCE\xB7\xCE\xBB\xCE\xB9\xCE\xAD\xCF\x81\xCE\xB7\xCE\xBB\xCE\xB7\xCE\xBC\xCE\xAD\xCF\x81\xCE\xB1"
       As Text: "καλημέραμηλιέρηλημέρα"
     encoded_values_buf->ToString()
       Which is: 
"\xCE\xBA\xCE\xB1\xCE\xBB\xCE\xB7\xCE\xBC\xCE\xAD\xCF\x81\xCE\xB1\xBC\xCE\xB7\xCE\xBB\xCE\xB9\xCE\xAD\xCF\x81\xCE\xB7\xBB\xCE\xB7\xCE\xBC\xCE\xAD\xCF\x81\xCE\xB1"
   ```



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to