add date analyzer

This commit is contained in:
Paul Pan 2023-11-03 22:17:53 +08:00
parent beea7f66c8
commit e64ee43824
4 changed files with 72 additions and 0 deletions

View File

@ -58,6 +58,7 @@ include_directories(password-analyzer PRIVATE ${CMAKE_BINARY_DIR}/external)
set(COMMON_SOURCES set(COMMON_SOURCES
defs.hpp defs.hpp
utils.cpp utils.cpp
stat_date.cpp
stat_keystroke.cpp stat_keystroke.cpp
stat_length.cpp stat_length.cpp
stat_struct.cpp stat_struct.cpp

View File

@ -10,5 +10,8 @@ int main() {
stat_keystroke(DataSource::YAHOO); stat_keystroke(DataSource::YAHOO);
stat_keystroke(DataSource::CSDN); stat_keystroke(DataSource::CSDN);
stat_date(DataSource::YAHOO);
stat_date(DataSource::CSDN);
return 0; return 0;
} }

View File

@ -26,6 +26,7 @@ bool is_alpha(char c);
std::generator<std::string> passwords(const DataSource &source); std::generator<std::string> passwords(const DataSource &source);
void stat_date(const DataSource &source);
void stat_keystroke(const DataSource &source); void stat_keystroke(const DataSource &source);
void stat_length(const DataSource &source); void stat_length(const DataSource &source);
void stat_struct(const DataSource &source); void stat_struct(const DataSource &source);

67
stat_date.cpp Normal file
View File

@ -0,0 +1,67 @@
#include <map>
#include <regex>
#include <vector>
#include "defs.hpp"
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) {
spdlog::info("stat_date({})", magic_enum::enum_name(source));
std::map<std::string, size_t> date_map;
std::map<size_t, size_t> type_map;
for (auto const &password : passwords(source)) {
for (int i = 0; i < sizeof(date_match) / sizeof(date_match[0]); i++) {
const auto &match = date_match[i];
auto bgn = std::sregex_iterator(password.begin(), password.end(), match);
auto end = std::sregex_iterator();
if (bgn != end) {
auto date = bgn->str();
date_map[date]++;
type_map[i]++;
break;
}
}
}
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);
}
}