password-analyzer/stat_date.cpp

#include <algorithm>
#include <execution>
#include <map>
#include <numeric>
#include <regex>
#include <string_view>
#include <vector>
#include "defs.hpp"

template <class T> using vec = std::vector<T>;
using dmap                   = std::map<std::string, size_t>;
using smap                   = std::map<size_t, size_t>;
using pss                    = std::pair<std::string, size_t>;
using pvec                   = vec<pss>;

const std::regex date_match[] = {
    // YYYY MM DD
    std::regex(R"((?:19|20)\d{2}(?:0[1-9]|1[0-2])(?:0[1-9]|[1-2][0-9]|3[0-1]))"),
    std::regex(R"((?:19|20)\d{2}\.(?:0[1-9]|1[0-2])\.(?:0[1-9]|[1-2][0-9]|3[0-1]))"),
    std::regex(R"((?:19|20)\d{2}-(?:0[1-9]|1[0-2])-(?:0[1-9]|[1-2][0-9]|3[0-1]))"),
    std::regex(R"((?:19|20)\d{2}/(?:0[1-9]|1[0-2])/(?:0[1-9]|[1-2][0-9]|3[0-1]))"),
    // MM DD YYYY
    std::regex(R"((?:0[1-9]|1[0-2])(?:0[1-9]|[1-2][0-9]|3[0-1])(?:19|20)\d{2})"),
    std::regex(R"((?:0[1-9]|1[0-2])\.(?:0[1-9]|[1-2][0-9]|3[0-1])\.(?:19|20)\d{2})"),
    std::regex(R"((?:0[1-9]|1[0-2])-(?:0[1-9]|[1-2][0-9]|3[0-1])-(?:19|20)\d{2})"),
    std::regex(R"((?:0[1-9]|1[0-2])/(?:0[1-9]|[1-2][0-9]|3[0-1])/(?:19|20)\d{2})"),
    // YY MM DD
    std::regex(R"(\d{2}(?:0[1-9]|1[0-2])(?:0[1-9]|[1-2][0-9]|3[0-1]))"),
    std::regex(R"(\d{2}\.(?:0[1-9]|1[0-2])\.(?:0[1-9]|[1-2][0-9]|3[0-1]))"),
    std::regex(R"(\d{2}-(?:0[1-9]|1[0-2])-(?:0[1-9]|[1-2][0-9]|3[0-1]))"),
    std::regex(R"(\d{2}/(?:0[1-9]|1[0-2])/(?:0[1-9]|[1-2][0-9]|3[0-1]))"),
    // MM DD YY
    std::regex(R"((?:0[1-9]|1[0-2])(?:0[1-9]|[1-2][0-9]|3[0-1])\d{2})"),
    std::regex(R"((?:0[1-9]|1[0-2])\.(?:0[1-9]|[1-2][0-9]|3[0-1])\.\d{2})"),
    std::regex(R"((?:0[1-9]|1[0-2])-(?:0[1-9]|[1-2][0-9]|3[0-1])-\d{2})"),
    std::regex(R"((?:0[1-9]|1[0-2])/(?:0[1-9]|[1-2][0-9]|3[0-1])/\d{2})"),
    // MM DD
    std::regex(R"((?:0[1-9]|1[0-2])(?:0[1-9]|[1-2][0-9]|3[0-1]))"),
    std::regex(R"((?:0[1-9]|1[0-2])\.(?:0[1-9]|[1-2][0-9]|3[0-1]))"),
    std::regex(R"((?:0[1-9]|1[0-2])-(?:0[1-9]|[1-2][0-9]|3[0-1]))"),
    std::regex(R"((?:0[1-9]|1[0-2])/(?:0[1-9]|[1-2][0-9]|3[0-1]))"),
};

void stat_date(const DataSource &source) {
    timeit(fmt::format("stat_date({})", magic_enum::enum_name(source)));

    // 35s
    vec<std::string> passwords_vec;
    vec<pvec>        partial;
    dmap             date_map;
    smap             type_map;

    auto eval = [](auto const &str) {
        vec<pss> ret;
        for (int i = 0; i < sizeof(date_match) / sizeof(date_match[0]); i++) {
            const auto &match = date_match[i];
            auto        bgn   = std::sregex_iterator(str.begin(), str.end(), match);
            auto        end   = std::sregex_iterator();

            if (bgn != end) {
                auto date = bgn->str();
                ret.emplace_back(std::make_pair(date, i));
                break;
            }
        }
        return ret;
    };

    std::ranges::copy(passwords(source), std::back_inserter(passwords_vec));
    partial.resize(passwords_vec.size());

    {
        timeit("regex find date");
        std::transform(std::execution::par, passwords_vec.begin(), passwords_vec.end(), partial.begin(), eval);
    }

    {
        timeit("count result");
        for (auto const &p : partial) {
            for (auto const &r : p) {
                date_map[r.first]++;
                type_map[r.second]++;
            }
        }
    }

    std::vector<std::pair<std::string, size_t>> date_vec(date_map.begin(), date_map.end());
    std::vector<std::pair<size_t, size_t>>      type_vec(type_map.begin(), type_map.end());
    std::sort(date_vec.begin(), date_vec.end(), [](auto const &a, auto const &b) { return a.second > b.second; });
    std::sort(type_vec.begin(), type_vec.end(), [](auto const &a, auto const &b) { return a.second > b.second; });

    for (auto &&[date, count] : date_vec | std::views::take(10)) {
        spdlog::info("{}: {}", date, count);
    }

    for (auto &&[type, count] : type_vec | std::views::take(10)) {
        spdlog::info("{}: {}", type, count);
    }
}