Multiples minors changes (cgi env, comment, ...)

srcs/Client.cpp

@@ -75,10 +75,12 @@ Client &	Client::operator=( Client const & rhs )
  * PUBLIC MEMBER FUNCTIONS
  *********************************************/
 
-// http headers :
+/* 	HTTP Headers :
-// https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers
+	https://www.iana.org/assignments/http-fields/http-fields.xhtml
-// https://www.ibm.com/docs/en/cics-ts/5.3?topic=protocol-http-requests
+	https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers
-// https://www.tutorialspoint.com/http/http_requests.htm
+	https://www.ibm.com/docs/en/cics-ts/5.3?topic=protocol-http-requests
+	https://www.tutorialspoint.com/http/http_requests.htm
+*/
 void	Client::parse_request_headers(std::vector<ServerConfig> &servers)
 {
 	if (raw_request.find(CRLF CRLF) == NPOS)
@@ -96,8 +98,8 @@ void	Client::parse_request_headers(std::vector<ServerConfig> &servers)
 
 	if (status)
 		return;
-	assigned_server = ::_determine_process_server(this, servers);
+	assigned_server = _determine_process_server(this, servers);
-	assigned_location = ::_determine_location(*assigned_server, _request.abs_path);
+	assigned_location = _determine_location(*assigned_server, _request.abs_path);
 	_check_request_errors();
 	if (status)
 		return;
@@ -389,6 +391,7 @@ void	Client::_parse_request_fields()
 	::str_map_key_tolower(_request.headers);
 }
 
+// TODO : I think its now useless. Probably to delete.
 void	Client::_parse_port_hostname(std::string host)
 {
 	size_t	pos;
@@ -453,6 +456,98 @@ void	Client::_check_request_errors()
 	return;
 }
 
+ServerConfig	*Client::_determine_process_server(Client *client, std::vector<ServerConfig> &servers)
+{
+/*
+	Behavior like this :
+	http://nginx.org/en/docs/http/request_processing.html
+*/
+
+	std::string	server_name = client->get_rq_headers("Host");
+	std::cerr << "server_name = " << server_name << "\n";
+	size_t	pos = server_name.rfind(':');
+	if (pos != NPOS)
+		server_name.erase(pos);
+	std::cerr << "server_name = " << server_name << "\n";
+
+	std::vector<ServerConfig>::iterator	it = servers.begin();
+	std::vector<ServerConfig>::iterator	default_server = servers.end();
+
+	while (it != servers.end())
+	{
+		if (it->host == client->get_cl_lsocket()->host
+		&&	it->port == client->get_cl_lsocket()->port)
+		{
+			if ( std::find(it->server_name.begin(), it->server_name.end(), server_name) != it->server_name.end() )
+				break;
+			else if (default_server == servers.end())
+				default_server = it;
+		}
+		++it;
+	}
+	if (it != servers.end())
+		return (&(*it));
+	else
+		return (&(*default_server));
+}
+
+// const?
+const LocationConfig	*Client::_determine_location(const ServerConfig &server, const std::string &path)
+{
+/* RULES ***
+
+If a path coresponds exactly to a location, use that one
+if no path coresponds then use the most correct one
+	most correct means the most precise branch that is still above
+	the point we are aiming for
+
+New Rule for location paths, they never end in /
+Sooo
+If we get a url that ends in / ignore the last /
+
+*/
+
+	std::string	uri = path;
+	if (uri[uri.size() - 1] == '/' && uri.size() != 1)
+		uri.erase(uri.size() - 1);
+
+
+	for (std::vector<LocationConfig>::const_iterator it = server.locations.begin(); it != server.locations.end(); it++)
+	{
+//		std::cout << it->path << "  --  ";
+		if (it->path.size() > uri.size())
+			continue ;
+
+		if (uri.compare(0, it->path.size(), it->path) == 0)
+		{
+			if (it->path.size() == uri.size())
+				return (&(*it));
+			else if (uri[it->path.size()] == '/')
+				return (&(*it));
+	// this works cuz only ever looking for a / burried in a longer path
+		}
+	}
+	return (&(server.locations.back()));
+
+
+//	/test/mdr
+//	/test/mdr/
+//	/test/mdrBST
+
+/////// More stuff to check this still works with
+
+// /test/test_
+// /test/test_/
+// /test/test_deeper
+// /test/test_deeper/
+// /test/test_deepei
+// /test/test_deepei/
+// /test/test_deeperi
+// /test/test_deeper/super_deep/
+// /test/aaaaaaaaaaa/super_deep/
+
+}
+
 /*********************************************
  * OVERLOAD
  *********************************************/

srcs/Client.hpp

@@ -105,15 +105,15 @@ class Client
 		void	_parse_chunked_body(size_t pos);
 		void	_parse_multipart_body(size_t pos);
 		void	_check_request_errors();
+
+		ServerConfig*
+			_determine_process_server(Client *client, std::vector<ServerConfig> &servers);
+		const LocationConfig*
+			_determine_location(const ServerConfig &server, const std::string &path);
 };
 
 bool	operator==(const Client& lhs, const Client& rhs);
 bool	operator==(const Client& lhs, int fd);
 bool	operator==(int fd, const Client& rhs);
 
-// Temporary Global Scope. Probably move to Client in the future.
-ServerConfig	*_determine_process_server(Client *client, std::vector<ServerConfig> &servers);
-const LocationConfig	*_determine_location(const ServerConfig &server, const std::string &path);
-
 #endif
-

srcs/webserv/Webserv.hpp

@@ -83,40 +83,21 @@ class Webserv
 		void		_append_base_headers(Client *client);
 		void		_construct_response(Client *client);
 		void		_process_method(Client *client, std::string &path);
+		std::string	_replace_url_root(Client *client, std::string path);
 		void		_insert_status_line(Client *client);
 		void		_error_html_response(Client *client);
 		void		_append_body(Client *client, const std::string &body, const std::string &file_extension = "");
-		// ServerConfig			*_determine_process_server(Client *client); // cant be const cause of error_pages.operator[]
-		// const LocationConfig	*_determine_location(const ServerConfig &server, const std::string &path) const;
-		std::string				_determine_file_extension(const std::string &path) const;
 	// method_get.cpp
-
-// move later
-		std::string	_replace_url_root(Client *client, std::string path);
-
 		void		_get(Client *client, std::string &path);
 		void		_get_file(Client *client, const std::string &path);
 		void		_autoindex(Client *client, const std::string &path);
+		std::string	_determine_file_extension(const std::string &path) const;
 	// method_post.cpp
 		void		_post(Client *client, const std::string &path);
 		void		_upload_files(Client *client);
 	// method_delete.cpp
 		void		_delete(Client *client, const std::string &path);
 		void		_delete_file(Client *client, const std::string &path);
-	// cgi_script.cpp
-		bool		_is_cgi(Client *client, std::string path);
-		size_t		_cgi_pos(Client *client, std::string &path, size_t pos);
-		std::string	_exec_cgi(Client *client);
-		void		_set_env_vector(Client *client, std::vector<std::string> &env_vector);
-		void		_set_env_cstr(char *env_cstr[], std::vector<std::string> &env_vector);
-		std::string	_dup_env(std::string var, std::string val);
-		std::string	_dup_env(std::string var, int i);
-		std::string	_exec_script(Client *client, char *env[]);
-		void		_check_script_output(Client *client, std::string & output);
-		void		_check_script_status(Client *client, std::string & output);
-		void		_check_script_fields(Client *client, std::string & output);
-		void		_add_script_body_length_header(std::string & output);
-		void		_remove_body_leading_empty_lines(std::string & output);
 	// epoll_update.cpp
 		int			_epoll_update(int fd, uint32_t events, int op);
 		int			_epoll_update(int fd, uint32_t events, int op, void *ptr);
@@ -136,7 +117,32 @@ class Webserv
 		void		_init_mime_types_map();
 	// timeout.cpp
 		void		_timeout();
+	// cgi.cpp
+		bool		_is_cgi(Client *client, std::string path);
+		size_t		_cgi_pos(Client *client, std::string &path, size_t pos);
+		std::string	_exec_cgi(Client *client);
+		void		_set_env_vector(Client *client, std::vector<std::string> &env_vector);
+		void		_set_env_cstr(char *env_cstr[], std::vector<std::string> &env_vector);
+		std::string	_dup_env(std::string var, std::string val);
+		std::string	_dup_env(std::string var, int i);
+		std::string	_exec_script(Client *client, char *env[]);
+		void		_check_script_output(Client *client, std::string & output);
+		void		_check_script_status(Client *client, std::string & output);
+		void		_check_script_fields(Client *client, std::string & output);
+		void		_add_script_body_length_header(std::string & output);
+		void		_remove_body_leading_empty_lines(std::string & output);
 };
 
 #endif
 
+/*
+	HTTP Semantics:
+		https://www.rfc-editor.org/rfc/rfc9110.html
+		https://www.bortzmeyer.org/9110.html
+		https://www.bortzmeyer.org/cours-http-cnam.html
+	HTTP/1.1:
+		https://www.rfc-editor.org/rfc/rfc9112.html
+		https://www.bortzmeyer.org/9112.html
+	CGI:
+		https://www.rfc-editor.org/rfc/rfc3875.html
+*/

srcs/webserv/accept.cpp

@@ -21,8 +21,8 @@ void	Webserv::_accept_connection(listen_socket &lsocket)
 	::fcntl(accepted_fd, F_SETFL, O_NONBLOCK);
 
 	infos = _extract_infos(addr);
-	Client	client(accepted_fd, &lsocket, infos["port"], infos["ip"]);
+	Client	new_client(accepted_fd, &lsocket, infos["port"], infos["ip"]);
-	_clients.push_back(client);
+	_clients.push_back(new_client);
 	_epoll_update(accepted_fd, EPOLLIN, EPOLL_CTL_ADD);
 }
 

srcs/webserv/cgi.cpp

@@ -1,5 +1,9 @@
 
 #include "Webserv.hpp"
+/*
+	CGI RFC:
+	https://www.rfc-editor.org/rfc/rfc3875.html
+*/
 
 bool	Webserv::_is_cgi(Client *client, std::string path)
 {
@@ -99,23 +103,26 @@ std::string	Webserv::_dup_env(std::string var, int i)
 }
 
 // TODO : verifier que les variables sont corrects
+/*
+	https://www.rfc-editor.org/rfc/rfc3875#section-4.1
+*/
 void	Webserv::_set_env_vector(Client *client, std::vector<std::string> &env_vector)
 {
 	env_vector.push_back(_dup_env("AUTH_TYPE"));		// authentification not supporte
 	env_vector.push_back(_dup_env("CONTENT_LENGTH"		, client->get_rq_body().size()));
 	env_vector.push_back(_dup_env("CONTENT_TYPE"		, client->get_rq_headers("Content-Type")));
-	env_vector.push_back(_dup_env("GATEWAY_INTERFACE"	, "CGI/1.1")); // https://www.rfc-editor.org/rfc/rfc387)
+	env_vector.push_back(_dup_env("GATEWAY_INTERFACE"	, "CGI/1.1")); // https://www.rfc-editor.org/rfc/rfc3875#section-4.1.4
-	env_vector.push_back(_dup_env("PATH_INFO"			, client->get_rq_script_info()));
+	env_vector.push_back(_dup_env("PATH_INFO"			, client->get_rq_script_info())); // LUKE: To Check
-	env_vector.push_back(_dup_env("PATH_TRANSLATED"));	// not supported
+	env_vector.push_back(_dup_env("PATH_TRANSLATED"));	// not supported // LUKE: Why not supported ?
 	env_vector.push_back(_dup_env("QUERY_STRING"		, client->get_rq_query()));
 	env_vector.push_back(_dup_env("REMOTE_ADDR"			, client->get_cl_ip()));
-	env_vector.push_back(_dup_env("REMOTE_HOST"			, client->get_rq_headers("Host"))); // just tes
+	env_vector.push_back(_dup_env("REMOTE_HOST"			, client->get_cl_ip())); // equal to REMOTE_ADDR or empty
-	env_vector.push_back(_dup_env("REMOTE_IDENT"));		// authentification not supporte	
+	env_vector.push_back(_dup_env("REMOTE_IDENT"));		// authentification not supported	
-	env_vector.push_back(_dup_env("REMOTE_USER"));		// authentification not supporte
+	env_vector.push_back(_dup_env("REMOTE_USER"));		// authentification not supported
 	env_vector.push_back(_dup_env("REQUEST_METHOD"		, client->get_rq_method_str()));
-	env_vector.push_back(_dup_env("SCRIPT_NAME"			, client->get_rq_script_path()));
+	env_vector.push_back(_dup_env("SCRIPT_NAME"			, client->get_rq_script_path())); // LUKE: To Check
-	env_vector.push_back(_dup_env("SERVER_NAME"			, client->get_rq_hostname()));
+	env_vector.push_back(_dup_env("SERVER_NAME"			, client->get_cl_lsocket()->host));
-	env_vector.push_back(_dup_env("SERVER_PORT"			, client->get_rq_port()));
+	env_vector.push_back(_dup_env("SERVER_PORT"			, client->get_cl_lsocket()->port));
 	env_vector.push_back(_dup_env("SERVER_PROTOCOL"		, "HTTP/1.1"));
 	env_vector.push_back(_dup_env("SERVER_SOFTWARE"		, "Webserv/0.1"));
 	env_vector.push_back(_dup_env("REDIRECT_STATUS"		, "200"));

srcs/webserv/http_status.hpp

@@ -2,7 +2,9 @@
 #ifndef HTTP_STATUS_HPP
 # define HTTP_STATUS_HPP
 
-// https://www.iana.org/assignments/http-status-codes/http-status-codes.xhtml
+/*
+	https://www.iana.org/assignments/http-status-codes/http-status-codes.xhtml
+*/
 
 /*
 	First version of macro HTML_ERROR(STATUS) dont work with call like this :

srcs/webserv/method_get.cpp

@@ -1,24 +1,13 @@
 
 #include "Webserv.hpp"
-
+#define MAX_FILESIZE 1 * MB // unused
-std::string	Webserv::_replace_url_root(Client *client, std::string path)
-{
-	std::cerr << "assigned_location->path =	" << client->assigned_location->path << "\n"; // debug
-	std::cerr << "path before =			" << path << "\n"; // DEBUG
-	if (client->assigned_location->path == "/")
-		path.insert(0, client->assigned_location->root);
-	else
-		path.replace(0, client->assigned_location->path.size(), client->assigned_location->root);
-	std::cerr << "path after =			" << path << "\n"; // DEBUG
-	return path;
-}
 
 // const?
-void	Webserv::_get(Client *client, std::string &path)
-{
 /*
 	https://www.rfc-editor.org/rfc/rfc9110.html#name-get
 */
+void	Webserv::_get(Client *client, std::string &path)
+{
 	std::cout << "_get()\n";
 	if (eval_file_type(path) == IS_DIR)
 	{
@@ -42,14 +31,13 @@ void	Webserv::_get(Client *client, std::string &path)
 		_get_file(client, path);
 }
 
-# define MAX_FILESIZE 1 * MB // unused
-void	Webserv::_get_file(Client *client, const std::string &path)
-{
 /*
 	std::ios::binary
 	https://gcc.gnu.org/onlinedocs/libstdc++/manual/fstreams.html#std.io.filestreams.binary
 	tldr : its seems to not be so simple to do read/write of binary file in a portable way.
 */
+void	Webserv::_get_file(Client *client, const std::string &path)
+{
 	std::ifstream		ifd; // For chunk, ifstream directly in struct CLient for multiples read without close() ?
 	std::stringstream	buf;
 
@@ -155,3 +143,11 @@ void	Webserv::_autoindex(Client *client, const std::string &path)
 		return ;
 	}
 }
+
+std::string	Webserv::_determine_file_extension(const std::string &path) const
+{
+	size_t	dot_pos = path.rfind(".");
+	if (dot_pos != NPOS && dot_pos + 1 < path.size())
+		return ( path.substr(dot_pos + 1) );
+	return (std::string(""));
+}

srcs/webserv/response.cpp

@@ -119,6 +119,21 @@ void	Webserv::_process_method(Client *client, std::string &path)
 	}
 }
 
+std::string	Webserv::_replace_url_root(Client *client, std::string path)
+{
+	std::cerr << "assigned_location->path =	" << client->assigned_location->path << "\n"; // debug
+	std::cerr << "path before =			" << path << "\n"; // DEBUG
+	if (client->assigned_location->path == "/")
+		path.insert(0, client->assigned_location->root);
+	else
+		path.replace(0, client->assigned_location->path.size(), client->assigned_location->root);
+	std::cerr << "path after =			" << path << "\n"; // DEBUG
+	return path;
+}
+
+/*
+	https://www.rfc-editor.org/rfc/rfc9112.html#name-status-line
+*/
 void	Webserv::_insert_status_line(Client *client)
 {
 	std::string	status_line;
@@ -176,111 +191,3 @@ void	Webserv::_append_body(Client *client, const std::string &body, const std::s
 	client->response.append(CRLF);
 	client->response.append(body);
 }
-
-// Temporary Global Scope. Probably move to Client in the future.
-ServerConfig	*_determine_process_server(Client *client, std::vector<ServerConfig> &servers)
-{
-/*
-	Behavior like this :
-	http://nginx.org/en/docs/http/request_processing.html
-*/
-
-	std::string	server_name = client->get_rq_headers("Host");
-	std::cerr << "server_name = " << server_name << "\n";
-	size_t	pos = server_name.rfind(':');
-	if (pos != NPOS)
-		server_name.erase(pos);
-	std::cerr << "server_name = " << server_name << "\n";
-
-	std::vector<ServerConfig>::iterator	it = servers.begin();
-	std::vector<ServerConfig>::iterator	default_server = servers.end();
-
-	while (it != servers.end())
-	{
-		if (it->host == client->get_cl_lsocket()->host
-		&&	it->port == client->get_cl_lsocket()->port)
-		{
-			if ( std::find(it->server_name.begin(), it->server_name.end(), server_name) != it->server_name.end() )
-				break;
-			else if (default_server == servers.end())
-				default_server = it;
-		}
-		++it;
-	}
-	if (it != servers.end())
-		return (&(*it));
-	else
-		return (&(*default_server));
-}
-
-// const?
-// Temporary Global Scope. Probably move to Client in the future.
-// is it still TMP Global Scope?
-const LocationConfig	*_determine_location(const ServerConfig &server, const std::string &path)
-{
-/* RULES ***
-
-If a path coresponds exactly to a location, use that one
-if no path coresponds then use the most correct one
-	most correct means the most precise branch that is still above
-	the point we are aiming for
-
-New Rule for location paths, they never end in /
-Sooo
-If we get a url that ends in / ignore the last /
-
-*/
-
-	std::string	uri = path;
-	if (uri[uri.size() - 1] == '/' && uri.size() != 1)
-		uri.erase(uri.size() - 1);
-
-
-	for (std::vector<LocationConfig>::const_iterator it = server.locations.begin(); it != server.locations.end(); it++)
-	{
-//		std::cout << it->path << "  --  ";
-		if (it->path.size() > uri.size())
-			continue ;
-
-		if (uri.compare(0, it->path.size(), it->path) == 0)
-		{
-			if (it->path.size() == uri.size())
-				return (&(*it));
-			else if (uri[it->path.size()] == '/')
-				return (&(*it));
-	// this works cuz only ever looking for a / burried in a longer path
-		}
-	}
-	return (&(server.locations.back()));
-
-
-//	/test/mdr
-//	/test/mdr/
-//	/test/mdrBST
-
-/* More stuff to check this still works with ***
-
-/test/test_
-/test/test_/
-/test/test_deeper
-/test/test_deeper/
-/test/test_deepei
-/test/test_deepei/
-/test/test_deeperi
-/test/test_deeper/super_deep/
-/test/aaaaaaaaaaa/super_deep/
-
-*/
-
-}
-
-
-
-
-std::string	Webserv::_determine_file_extension(const std::string &path) const
-{
-	size_t	dot_pos = path.rfind(".");
-	if (dot_pos != NPOS && dot_pos + 1 < path.size())
-		return ( path.substr(dot_pos + 1) );
-	return (std::string(""));
-}