Method GenerateTestCase

cpp/src/arrow/json/reader_test.cc:537–586

Creates a test case from valid JSON objects with a human-readable index field and a struct field of random data. `block_size_multiplier` is applied to the largest generated row length to determine the target block_size. i.e - higher multiplier means fewer batches

  // generated row length to determine the target block_size. i.e - higher multiplier
  // means fewer batches
  static TestCase GenerateTestCase(int num_rows, double block_size_multiplier = 3.0) {
    FieldVector data_fields = {field("s", utf8()), field("f", float64()),
                               field("b", boolean())};
    FieldVector fields = {field("i", int64()), field("d", struct_({data_fields}))};
    TestCase out;
    out.schema = schema(fields);
    out.num_rows = num_rows;

    constexpr int kSeed = 0x432432;
    std::default_random_engine engine(kSeed);
    std::vector<std::string> rows(num_rows);
    size_t max_row_size = 1;

    auto options = GenerateOptions::Defaults();
    options.null_probability = 0;
    for (int i = 0; i < num_rows; ++i) {
      StringBuffer string_buffer;
      Writer writer(string_buffer);
      ABORT_NOT_OK(Generate(data_fields, engine, &writer, options));
      std::string json = string_buffer.GetString();
      rows[i] = Join({"{\"i\":", std::to_string(i), ",\"d\":", json, "}\n"});
      max_row_size = std::max(max_row_size, rows[i].size());
    }

    auto block_size = static_cast<size_t>(max_row_size * block_size_multiplier);
    // Deduce the expected record batches from the target block size.
    std::vector<std::string> batch_rows;
    size_t pos = 0;
    for (const auto& row : rows) {
      pos += row.size();
      if (pos > block_size) {
        out.batches.push_back(
            RecordBatchFromJSON(out.schema, Join({"[", Join(batch_rows, ","), "]"})));
        batch_rows.clear();
        pos -= block_size;
      }
      batch_rows.push_back(row);
      out.json += row;
    }
    if (!batch_rows.empty()) {
      out.batches.push_back(
          RecordBatchFromJSON(out.schema, Join({"[", Join(batch_rows, ","), "]"})));
    }

    out.json_size = static_cast<int>(out.json.size());
    out.block_size = static_cast<int>(block_size);
    out.num_batches = static_cast<int>(out.batches.size());

    return out;
  }

  static std::string Join(const std::vector<std::string>& strings,
                          const std::string& delim = "", bool trailing_delim = false) {