checks-parser/email_parser/emailparser.cpp

108 lines
4.5 KiB
C++

#include "utils/utils.h"
#include "utils/base64.h"
#include <email_parser/emailparser.h>
#include <opencv2/opencv.hpp>
#include <opencv2/imgcodecs.hpp>
#include <opencv2/objdetect.hpp>
#include <opencv2/core/core.hpp>
#include <opencv2/videoio.hpp>
#include <utils/utils.h>
#include <check/check.h>
#include <boost/regex.hpp>
#include <boost/algorithm/string/regex.hpp>
#include <iostream>
#include <fstream>
#if __GNUC__ < 8 && __clang_major__ < 17
# include <experimental/filesystem>
using namespace std::experimental::filesystem;
#else
# include <filesystem>
using namespace std::filesystem;
#endif
EmailParser::EmailParser() {
}
std::map<std::string, std::string> EmailParser::parse(std::string &email_content) {
//1. Search "Content-Type: image/.*" in the .eml file.
// 1.1 If found 0, go to [2]
// 1.2 If found 1, try decoding it, if it's not a QR code, go to [2]
// 1.3 Loop through every found entry. If not found in any, go to [2]
//2. Try decoding content of the e-mail
//3. Search "t=\d{8}T\d{4,6}&s=\d{1,6}\.\d{1,2}&fn=\d{10,16}&i=\d{6}&fp=\d{10}&n=\d". Note that in some emails = and & signs could be replaced with its code in HTTP requests: %3D, %26
// 3.1 If not found, notify the user that we could not parse the .eml file
/* Find content-types */
Check c;
std::vector<std::pair<int, int>> content_types = {};
boost::regex content_type_regex("Content-Type: image/(gif|png|jpg)");
boost::regex part_end_regex("--.{5,48}");
for (boost::sregex_iterator it{email_content.begin(), email_content.end(), content_type_regex}, end{}; it != end; it ++) {
unsigned int start_position = it->position(), end_position = -1;
for (boost::sregex_iterator it2{email_content.begin() + start_position, email_content.end(), part_end_regex}, end2{}; it2 != end2; it2++) {
end_position = it2->position();
break;
}
content_types.push_back(std::pair<int, int>(start_position, end_position));
}
/* iterate through found content-types and try searching qr codes, decode them and see if it's the needed data */
for (unsigned int i = 0; i < content_types.size(); i ++) {
boost::regex to_erase("(Content.{5,64}\\r\\n)+");
boost::regex to_erase_two("--.{5,48}");
std::string part = email_content.substr(content_types[i].first, content_types[i].second);
boost::erase_regex(part, to_erase);
boost::erase_regex(part, to_erase_two);
part.erase(std::remove(part.begin(), part.end(), '\r'), part.end());
part.erase(std::remove(part.begin(), part.end(), '\n'), part.end());
std::string decoded = base64_decode(part);
cv::Mat image;
if (decoded.substr(1, 3) == "PNG" || decoded.substr(1, 3) == "JPG") {
std::vector<uchar> data(decoded.begin(), decoded.end());
image = cv::imdecode(cv::Mat(data), 1);
} else if (decoded.substr(0, 3) == "GIF") {
std::string gif_file_path = get_application_home_path() + "/temp.gif";
std::ofstream gif_output(gif_file_path, std::ios::binary);
gif_output << decoded;
gif_output.close();
cv::VideoCapture gif(gif_file_path, cv::CAP_FFMPEG);
gif.read(image);
}
cv::QRCodeDetector qrDecoder = cv::QRCodeDetector();
std::string decoded_qr_params = qrDecoder.detectAndDecode(image);
boost::regex check_data_content("t=\\d+T\\d+&s=\\d+\\.\\d+&fn=\\d{16}&i=\\d{4,5}&fp=\\d{10}&n=\\d");
if (boost::regex_match(decoded_qr_params, check_data_content)) {
std::map<std::string, std::string> paramsMap = get_params_from_string(decoded_qr_params);
return paramsMap;
}
}
/* If the E-Mail has no QR code in it as a separate part, there's posibilly a QR code inserted using html's tag <img> with base64-encoded image. Try searching it */
/* If there's no such case, the last chance is <img src="..."> which will have a link with needed parameters or the qr code that should be downloaded and decoded */
/* If the code has reached this part and found nothing, it's most likely that there are no QR codes at all. */
// return Check();
}
std::map<std::string, std::string> EmailParser::parse_file(std::string path) {
std::ifstream ifile(path, std::ios::in | std::ios::binary);
const unsigned int size = std::filesystem::file_size(path);
std::string content(size, '\0');
ifile.read(content.data(), size);
return parse(content);
return std::map<std::string, std::string>();
}