Method PutBinaryArray

cpp/src/parquet/encoder.cc:1534–1575

  template <typename ArrayType>
  void PutBinaryArray(const ArrayType& array) {
    auto previous_len = static_cast<uint32_t>(last_value_.length());
    std::string_view last_value_view = last_value_;

    PARQUET_THROW_NOT_OK(::arrow::VisitArraySpanInline<typename ArrayType::TypeClass>(
        *array.data(),
        [&](::std::string_view view) {
          if (ARROW_PREDICT_FALSE(view.size() >= kMaxByteArraySize)) {
            return Status::Invalid(
                "Parquet cannot store strings with size 2GB or more, got: ", view.size());
          }
          const ByteArray src{view};

          uint32_t common_prefix_length = 0;
          const uint32_t len = src.len;
          const uint32_t maximum_common_prefix_length = std::min(previous_len, len);
          while (common_prefix_length < maximum_common_prefix_length) {
            if (last_value_view[common_prefix_length] != view[common_prefix_length]) {
              break;
            }
            common_prefix_length++;
          }
          previous_len = len;
          prefix_length_encoder_.Put({static_cast<int32_t>(common_prefix_length)}, 1);

          last_value_view = view;
          const auto suffix_length = static_cast<uint32_t>(len - common_prefix_length);
          if (suffix_length == 0) {
            suffix_encoder_.Put(&empty_, 1);
            return Status::OK();
          }
          const uint8_t* suffix_ptr = src.ptr + common_prefix_length;
          // Convert to ByteArray, so it can be passed to the suffix_encoder_.
          const ByteArray suffix(suffix_length, suffix_ptr);
          suffix_encoder_.Put(&suffix, 1);

          unencoded_byte_array_data_bytes_ += len;
          return Status::OK();
        },
        []() { return Status::OK(); }));
    last_value_ = last_value_view;
  }

  ::arrow::BufferBuilder sink_;
  DeltaBitPackEncoder<Int32Type> prefix_length_encoder_;