dict_optimze.cpp

#include <iostream>
#include <cstdio>

#include <arrow/api.h>
#include <arrow/io/memory.h>

inline constexpr auto arrow_throw_not_ok = [](auto status) {
    if (!status.ok())
    {
        throw std::runtime_error(status.ToString());
    }
};

int main()
{
    const std::vector<std::string> list = {"old lady", "sudens", "camp nou"};

    // initialize builder
    std::unique_ptr<arrow::ArrayBuilder> builder;
    arrow_throw_not_ok(arrow::MakeBuilder(arrow::default_memory_pool(), arrow::dictionary(arrow::int32(), arrow::utf8()),
                                          &builder));

    // append/finish
    std::shared_ptr<arrow::Array> value_array;
    auto typed_builder = static_cast<arrow::StringDictionaryBuilder *>(builder.get());
    for (const auto &str : list)
    {
        arrow_throw_not_ok(typed_builder->Append(str));
    }
    arrow_throw_not_ok(builder->Finish(&value_array));
    std::cout << value_array->type()->ToString() << std::endl;
}