#include "../include/DownloadUtils.hpp" #include "../include/Program.hpp" #include "../include/Storage.hpp" #include "../external/cppcodec/base64_url.hpp" #include #include #include static const bool debug_download = false; static int accumulate_string(char *data, int size, void *userdata) { std::string *str = (std::string*)userdata; if(str->size() + size > 1024 * 1024 * 100) // 100mb sane limit, TODO: make configurable return 1; str->append(data, size); return 0; } static const char *useragent_str = "user-agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/81.0.4044.138 Safari/537.36"; namespace QuickMedia { DownloadResult download_head_to_string(const std::string &url, std::string &result, bool use_browser_useragent, bool fail_on_error) { sf::Clock timer; std::vector args; args.insert(args.end(), { "curl", "-I", "-H", "Accept-Language: en-US,en;q=0.5", "-H", "Connection: keep-alive", "--compressed", "-s" }); if(fail_on_error) args.push_back("-f"); if(use_browser_useragent) { args.push_back("-H"); args.push_back(useragent_str); } args.push_back("--"); args.push_back(url.c_str()); args.push_back(nullptr); if(debug_download) { for(const char *arg : args) { if(arg) fprintf(stderr, "'%s' ", arg); } fprintf(stderr, "\n"); } if(exec_program(args.data(), accumulate_string, &result) != 0) return DownloadResult::NET_ERR; fprintf(stderr, "Download duration for %s: %d ms\n", url.c_str(), timer.getElapsedTime().asMilliseconds()); return DownloadResult::OK; } // TODO: Add timeout DownloadResult download_to_string(const std::string &url, std::string &result, const std::vector &additional_args, bool use_browser_useragent, bool fail_on_error) { sf::Clock timer; std::vector args; args.insert(args.end(), { "curl", "-H", "Accept-Language: en-US,en;q=0.5", "-H", "Connection: keep-alive", "--compressed", "-s", "-L" }); if(fail_on_error) args.push_back("-f"); for(const CommandArg &arg : additional_args) { args.push_back(arg.option.c_str()); args.push_back(arg.value.c_str()); } if(use_browser_useragent) { args.push_back("-H"); args.push_back(useragent_str); } args.push_back("--"); args.push_back(url.c_str()); args.push_back(nullptr); if(debug_download) { for(const char *arg : args) { if(arg) fprintf(stderr, "'%s' ", arg); } fprintf(stderr, "\n"); } if(exec_program(args.data(), accumulate_string, &result) != 0) return DownloadResult::NET_ERR; fprintf(stderr, "Download duration for %s: %d ms\n", url.c_str(), timer.getElapsedTime().asMilliseconds()); return DownloadResult::OK; } DownloadResult download_to_string_cache(const std::string &url, std::string &result, const std::vector &additional_args, bool use_browser_useragent, DownloadErrorHandler error_handler, Path cache_path) { Path media_dir; Path media_file_path; if(cache_path.data.empty()) { media_dir = get_cache_dir().join("media"); media_file_path = Path(media_dir).join(cppcodec::base64_url::encode(url)); } else { media_dir = cache_path.parent(); media_file_path = std::move(cache_path); } if(get_file_type(media_file_path) == FileType::REGULAR) { if(file_get_content(media_file_path, result) == 0) { fprintf(stderr, "Loaded %s from cache\n", url.c_str()); return DownloadResult::OK; } else { fprintf(stderr, "Failed to get content of cached media file: %s\n", media_file_path.data.c_str()); return DownloadResult::ERR; } } else { DownloadResult download_result = download_to_string(url, result, additional_args, use_browser_useragent, error_handler ? false : true); if(download_result == DownloadResult::OK && error_handler) download_result = error_handler(result) ? DownloadResult::OK : DownloadResult::ERR; if(download_result == DownloadResult::OK) { Path media_file_path_tmp(media_file_path.data + ".tmp"); if(create_directory_recursive(media_dir) == 0 && file_overwrite(media_file_path_tmp, result) == 0) { if(rename_atomic(media_file_path_tmp.data.c_str(), media_file_path.data.c_str()) != 0) { perror("rename"); download_result = DownloadResult::ERR; } } else { download_result = DownloadResult::ERR; } } return download_result; } } DownloadResult download_to_file(const std::string &url, const std::string &destination_filepath, const std::vector &additional_args, bool use_browser_useragent) { Path tmp_filepath = destination_filepath; tmp_filepath.append(".tmp"); // TODO: Optimize with temporary '\0' size_t dir_end = tmp_filepath.data.rfind('/'); if(dir_end != std::string::npos && create_directory_recursive(tmp_filepath.data.substr(0, dir_end)) != 0) return DownloadResult::ERR; std::vector args = additional_args; args.push_back({ "-o", tmp_filepath.data.c_str() }); std::string dummy; DownloadResult res = download_to_string(url, dummy, std::move(args), use_browser_useragent); if(res != DownloadResult::OK) return res; if(rename_atomic(tmp_filepath.data.c_str(), destination_filepath.c_str()) != 0) { perror("rename"); return DownloadResult::ERR; } return DownloadResult::OK; } // TODO: Add timeout DownloadResult download_to_json(const std::string &url, rapidjson::Document &result, const std::vector &additional_args, bool use_browser_useragent, bool fail_on_error) { sf::Clock timer; std::vector args; args.insert(args.end(), { "curl", "-H", "Accept-Language: en-US,en;q=0.5", "-H", "Connection: keep-alive", "--compressed", "-s", "-L" }); if(fail_on_error) args.push_back("-f"); for(const CommandArg &arg : additional_args) { args.push_back(arg.option.c_str()); args.push_back(arg.value.c_str()); } if(use_browser_useragent) { args.push_back("-H"); args.push_back(useragent_str); } args.push_back("--"); args.push_back(url.c_str()); args.push_back(nullptr); if(debug_download) { for(const char *arg : args) { if(arg) fprintf(stderr, "'%s' ", arg); } fprintf(stderr, "\n"); } ReadProgram read_program; if(exec_program_pipe(args.data(), &read_program) != 0) return DownloadResult::NET_ERR; FILE *file = fdopen(read_program.read_fd, "rb"); if(!file) { program_clear_current_thread(); wait_program(read_program.pid); return DownloadResult::ERR; } char read_buffer[8192]; rapidjson::FileReadStream is(file, read_buffer, sizeof(read_buffer)); rapidjson::ParseResult parse_result = result.ParseStream(is); program_clear_current_thread(); fclose(file); wait_program(read_program.pid); fprintf(stderr, "Download duration for %s: %d ms\n", url.c_str(), timer.getElapsedTime().asMilliseconds()); return parse_result.IsError() ? DownloadResult::ERR : DownloadResult::OK; } }