Method LoadBatch

cpp/src/parquet/arrow/reader.cc:491–546

  bool IsOrHasRepeatedChild() const final { return false; }

  Status LoadBatch(int64_t records_to_read) final {
    BEGIN_PARQUET_CATCH_EXCEPTIONS
    out_ = nullptr;
    record_reader_->Reset();
    // Pre-allocation gives much better performance for flat columns
    record_reader_->Reserve(records_to_read);
    const bool should_load_statistics = ctx_->reader_properties->should_load_statistics();
    int64_t num_target_row_groups = 0;
    while (records_to_read > 0) {
      if (!record_reader_->HasMoreData()) {
        break;
      }
      int64_t records_read = record_reader_->ReadRecords(records_to_read);
      records_to_read -= records_read;
      if (records_read == 0) {
        NextRowGroup();
      } else {
        num_target_row_groups++;
        // We can't mix multiple row groups when we load statistics
        // because statistics are associated with a row group. If we
        // want to mix multiple row groups and keep valid statistics,
        // we need to implement a statistics merge logic.
        if (should_load_statistics) {
          break;
        }
      }
    }
    RETURN_NOT_OK(TransferColumnData(
        record_reader_.get(),
        num_target_row_groups == 1 ? input_->column_chunk_metadata() : nullptr, field_,
        descr_, ctx_.get(), &out_));
    return Status::OK();
    END_PARQUET_CATCH_EXCEPTIONS
  }

  ::arrow::Status BuildArray(int64_t length_upper_bound,
                             std::shared_ptr<::arrow::ChunkedArray>* out) final {
    *out = out_;
    return Status::OK();
  }

  const std::shared_ptr<Field> field() override { return field_; }

 private:
  std::shared_ptr<ChunkedArray> out_;
  void NextRowGroup() {
    std::unique_ptr<PageReader> page_reader = input_->NextChunk();
    record_reader_->SetPageReader(std::move(page_reader));
  }

  std::shared_ptr<ReaderContext> ctx_;
  std::shared_ptr<Field> field_;
  std::unique_ptr<FileColumnIterator> input_;
  const ColumnDescriptor* descr_;
  std::shared_ptr<RecordReader> record_reader_;
};

// Column reader for extension arrays