Function split_text_chunks

src/framework/text/chunking.cpp:390–394

        utf8_codepoint_count(parenthetical->first, "tag-aware leading text chunk tag"));
    const int64_t body_budget = std::max<int64_t>(1, codepoint_budget - prefix_codepoints - 1);
    auto chunks = split_tag_aware_body(parenthetical->second, body_budget);
    for (auto & chunk : chunks) {
        chunk = parenthetical->first + " " + chunk;
    }
    return chunks;
}

std::vector<std::string> split_text_chunks_japanese(