Extract multipart messages from raw HTTP captures into individual files

extract-multipart.cc

/**
 * extract-multipart.cc
 *
 * This program extracts multipart messages from raw HTTP captures into
 * individual files.
 *
 * Usage: a.out <file ...>
 */
#include <iostream>
#include <fstream>
#include <string>
#include <sstream>
#include <iomanip>
#include <cassert>

using namespace std;

enum { kMaxLength = 70 };

const string hyphens = "--";
const string match("boundary=");

int main(int argc, char *argv[]) {
  int count = 0;

  for (int i = 1; i < argc; i++) {
    for (ifstream ifs(argv[i], ifstream::in); ifs; ifs.close()) {
      string boundary;

      for (string line; std::getline(ifs, line);) {
        if (line.find("Content-Type") != string::npos) {
          size_t pos = line.find(match);
          if (pos != string::npos) {
            pos += match.size();  // skip match pattern
            size_t end = line.find_first_of(" \r\n", pos);
            if ((end - pos) > 0) {
              boundary = line.substr(pos, end - pos);
              assert(boundary.size() <= kMaxLength);
            }
          }
        }

        if (line.size() > hyphens.size() &&
            line.substr(0, hyphens.size()) == hyphens &&
            line.substr(hyphens.size(), boundary.size()) == boundary) {
          string type;
          string lenInStr;
          ifs >> type >> type;
          ifs >> lenInStr >> lenInStr;

          int len = std::stoi(lenInStr);
#ifdef DEBUG
          cout << "Content-Type: " << type << ", "
               << "Content-Length: " << len << endl;
#endif

          // eat two newline characters
          std::getline(ifs, line);
          std::getline(ifs, line);

          string content(len, '\0');
          ifs.read(&content[0], len);

          // add leading 0's
          std::ostringstream ss;
          ss << std::setw(4) << std::setfill('0') << count++;

          std::ofstream out(boundary + "_" + string(ss.str()) + ".dat");
          out << content;
          out.close();
        }
      }
    }
  }

  return 0;
}