added text inside youpibanane files

+ renamed client's public function parse_request() into parse_request_headers()

.gitignore

@@ -20,3 +20,5 @@ ubuntu_cgi_tester
 webserv
 !**/webserv/
 *.log
+
+large.jpg

Makefile

@@ -1,11 +1,11 @@
 
 NAME		 =	webserv
-CXX			 =	c++
+CXX			 =	clang++
 
 CXXFLAGS	 =	-Wall -Wextra #-Werror
 CXXFLAGS	+=	$(HEADERS_D:%=-I%)
 CXXFLAGS	+=	-std=c++98
-CXXFLAGS	+=	-g
+CXXFLAGS	+=	-g3
 CXXFLAGS	+=	-MMD -MP #header dependencie
 #CXXFLAGS	+=	-O3
 
@@ -22,10 +22,11 @@ SRCS_D		 =	srcs \
 SRCS		 =	main.cpp \
 				base.cpp init.cpp close.cpp epoll_update.cpp signal.cpp \
 				accept.cpp request.cpp response.cpp \
+				method_get.cpp method_post.cpp method_delete.cpp \
 				run_loop.cpp \
-				ConfigParser.cpp \
-				ConfigParserUtils.cpp \
-				ConfigParserPost.cpp \
+				parser.cpp \
+				extraConfig.cpp \
+				postProcessing.cpp \
 				utils.cpp \
 				cgi_script.cpp \
 				Client.cpp \

README.md

@@ -1,25 +1,15 @@
 
 ## work together
 
-#### output cgi script :
-!	TODO : change all the '\n' by '\r\n'
-!	TODO : there is at least one header field followed by '\r\n\r\n' :
-			- "Content-Type"
-			- "Location"
-			- "Status"
-->	TODO : there is no field duplicate (resolve conflicts)
-->	TODO : if status field, change server status for this one
-!	TODO : there is no space between filed name and ":"
-!	TODO : if no Location field && no Status field -> status code = 200
-!	TODO?: handle Location field, either :
-			- local : start with '/' --> rerun the request with new uri
-			- client : start with '<scheme>:' --> send back status code 302
-
-#### what cgi ?
-- a basic form with "name" and "something", that return a html page with that
-- a script called by a file extension in URI
+#### TODO hugo
+- `_is_cgi()` and `_fill_cgi_path()`
+- two cgi tests :
+?	- a basic form with "name" and "something", that return a html page with that
+?		- for GET and POST
+?	- a script called by a file extension in URI
 
 #### questions
+- in client, fd is in privat, so accesible by getter, is it ok ?
 - in client.cpp i fill the port, is there a default one in case it's not in the request ?
 - timeout server but still works ?
 - path contains double "//" from `Webserv::_get()` in response.cpp
@@ -35,42 +25,6 @@
 - I don't save the STDIN and STDOUT before dup2 in child process, is it wrong ?
 - the response page is received long after the cgi-script is done, why ?
 
-#### notifications
-- i changed the Client getters in two categories :
-	- getters for requests infos : `get_rq_<info>`
-	- getters for client sides infos : `get_cl_<info>` (such as ip of client)
-- i changed the variables in request struct in Client :
-	- `path` become `uri` (ex. `/path/to/file?var=val`)
-	-  add `abs_path`     (ex. `/path/to/file`        )
-	-  add `query`        (ex.               `var=val`)
-- the header fields names, as key in map, are stored in lowercase, and getters are case-insensitives
-
-
-respsonse.cpp
-```
-_response()
-{
-    _determine_process_server()
-    _send_response()
-    {
-        _append_base_headers()
-        _construct_response()
-        {
-            _process_method()
-            {
-                _get()
-				{
-					_exec_cgi()
-				}
-            }
-        _insert_status_line()
-        ::send(headers)
-        ::send(body)
-        }
-    }
-}
-```
-
 ---
 ## man
 
@@ -149,6 +103,20 @@ _response()
 ## cgi rfc
 [rfc 3875](https://www.rfc-editor.org/rfc/rfc3875)
 
+#### output cgi script :
+!	TODO : change all the '\n' by '\r\n'
+!	TODO : there is at least one header field followed by '\r\n\r\n' :
+			- "Content-Type"
+			- "Location"
+			- "Status"
+!	TODO : there is no space between filed name and ":"
+!	TODO?: handle Location field, either :
+			- local : start with '/' --> rerun the request with new uri
+			- client : start with '<scheme>:' --> send back status code 302
+->	TODO : there is no field duplicate (resolve conflicts)
+->	TODO : if status field, change server status for this one
+->	TODO : if no Location field && no Status field -> status code = 200
+
 #### summary :
 - the cgi-script will send back at least one header field followed by an empty line
 - this header field will be one of three :
@@ -264,6 +232,45 @@ _response()
 [7 and 8: usefull informations about implementation and security](https://www.rfc-editor.org/rfc/rfc3875#section-7)
 
 
+---
+## http errors
+
+#### HTTP Client Errors
+- 400 	Bad Request 							This error code indicates that the request cannot be processed because of wrong syntax usage by the client.
+- 401 	Unauthorized 							This error code indicates that the client is not authorized to receive the requested data, without authentication. A login name and password based authentication might be required to access the requested data.
+- 403 	Forbidden 								There is no way you can access the requested data. A 403 error announces that the data is off limits.
+- 404 	Not Found 								This error indicates that the resources requested by the client are currently unavailable.
+- 405 	Method Not Allowed 						This error indicates wrong usage of request method. Depending on the kind of data requested, the appropriate request method must be chosen.
+- 406 	Not Acceptable 							When the data provided by a web server does not match the specifications made in ‘Accept’ header of the client HTTP request, this error is the result.
+- 407 	Proxy Authentication Required 			This error clearly indicates that an authentication from the proxy server is required to gain access to requested resources.
+- 408 	Request Timeout 						This type of error indicates that the client was delayed in making a request, within the specified time allocated to it, by the server.
+- 409 	Conflict 								This error code is displayed when the server perceives a conflict between two requests made simultaneously by different clients, for the same resource.
+- 410 	Gone 									This error code indicates that the requested data is no longer hosted on the server and therefore further requests made for it, would be futile.
+- 411 	Length Required 						If the request made by the client does not include information about the length of the requested data or resource, this error code is displayed.
+- 412 	Precondition Failed 					Some requests made by clients come attached with conditions that need to be satisfied by the server, before data transaction may happen. If these conditions are not met, error 412 results.
+- 413 	Request Entity Too Large 				When a client makes a request which is too overwhelming for the server’s resources to handle, it presents this error code.
+- 414 	Requested URI Too Long 					A Uniform Resource Identifier (URI) is a character string used to describe a data stream or resource on a server. Error 414 occurs when the server is unable to process the URI, because of limited resources and long string length.
+- 415 	Unsupported Media Type 					A server may be designed to allow only certain formats for media files. When error 415 is displayed, it indicates that the format of file being uploaded through a client request, does not match the requisite format.
+- 416 	Request Range Not Satisfiable 			Sometimes, a client may request for only a small part of a file, instead of asking for the entire file. If this request is not specified properly and the part of the file requested does not exist, this error is displayed.
+- 417 	Expectation Failed 						This error code is displayed when the server cannot meet the specifications provided in the request.
+- 422 	Unprocessable Entity 					This error is displayed when the request made, cannot be processed due to an error in semantic structure.
+- 423 	Locked 									This error is displayed when a requested piece of data or resource has been locked, making it inaccessible for a server.
+- 424 	Failed Dependency			 			A server may process a succession of requests from a client with the fulfillment of each, dependent on the one made before. This error is displayed when a request made before is not fulfilled, due to which the current request cannot be processed.
+- 426 	Upgrade Required 						This error signifies that the client may need to switch over to a secure protocol like TLS to get the request processed.
+- 444 	No Response								This error signifies that the server has simply rejected the client request and terminated connection.
+- 449 	Retry With 								This is a request made by the server to the client, to make the request again after executing certain actions or making specific changes in request. This is an error code introduced by Microsoft.
+- 499 	Client Closed Request	 				When client terminates a connection made with the server, while its processing the associated request, this error code is displayed.
+- 450 	Blocked By Windows Parental Controls 	Another error code introduced by Microsoft, this one is displayed when a URL is blocked by parental control settings on the web browsers.
+
+#### HTTP Server Errors
+- 500 	Internal Server Error 					A generic message displayed by the server, when the problem with the request cannot be specified by any other appropriate code.
+- 501 	Not Implemented							This error indicates the inability of the server to process a request, as it hasn’t been configured to respond to the request method used.
+- 502 	Bad Gateway 							Sometimes, hosted pages on web servers are transmitted to client via proxy servers. If the proxy server (to which a client has sent a request), fails connecting with the web server (known as the upstream server), error 502 results.
+- 503 	Service Unavailable 					When the server is already overloaded with multiple requests, it will temporarily stop entertaining new requests, by displaying a 503 error code.
+- 504 	Gateway Timeout 						When the request made by a proxy server to the web server hosting a resource times out, error 504 is reported.
+- 505 	HTTP Version Not Supported 				An error code seen rarely, it is displayed when the web server does not support the protocol version of the client request.
+
+
 ---
 ## cgi env variables
 [cgi env variables](https://www.rfc-editor.org/rfc/rfc3875#section-4.1)

YoupiBanane/Yeah/not_happy.bad_extension

@@ -0,0 +1 @@
+inside YoupiBanane/Yeah/not_happy.bad_extension

YoupiBanane/nop/other.pouic

@@ -0,0 +1 @@
+inside YoupiBanane/nop/other.pouic

YoupiBanane/nop/youpi.bad_extension

@@ -0,0 +1 @@
+inside YoupiBanane/nop/youpi.bad_extension

YoupiBanane/youpi.bad_extension

@@ -0,0 +1 @@
+inside YoupiBanane/youpi.bad_extension

YoupiBanane/youpi.bla

@@ -0,0 +1 @@
+inside YoupiBanane/youpi.bla

big.config

@@ -16,6 +16,7 @@ server {
 	location /board {
 		allow_methods GET;
 		root ./www/html;
+		cgi_ext php cgi
 	}
 
 	location /board/content {

default.config

@@ -10,14 +10,63 @@ server {
 #	client_body_limit 400;
 
 	index index.html; # this is another comment
+
 	root ./www/;
 
-# If not explicitly set, ConfigParser need to genererate a location block
-# like this for path "/" (based on field "root" and "index" of the server)
-	location / {
-		root ./www/;
-		index index.html;
+	error_page 404 ./www/error_pages/error_404.html;
+
+	location /list {
+		autoindex on;
 	}
 
-	allow_methods GET;
+	location /redirect {
+		redirect 307 https://fr.wikipedia.org/wiki/Ketchup;
+#		redirect 307 https://www.youtube.com/watch?v=rG6b8gjMEkw;
+	}
+
+	location /test {
+		index index1.html subdex.html;
+	}
+
+	location /test/index1.html {
+		index index1.html subdex.html;
+	}
+
+	location /hilarious_404/ {
+		redirect 301 https://berniesanders.com/404/;
+	}
+
+	# /stylesheet/ alone doesn't work, i mean we don't have wildcards...
+	location /stylesheet/style.css {
+#		root ./www/../;
+		root ./;
+	}
+
+	location /test/something.html {
+	#	allow_methods DELETE;
+	}
+
+#	location /something/long/here {
+#	}
+
+	location /test/test_deeper/ {
+#		allow_methods
+		autoindex on;
+	}
+
+	location /test/test_deeper/super_deep {
+		autoindex on;
+	}
+
+#	location /test/test_deeper/something.html {
+#		allow_methods DELETE;
+#	}
+
+
+# ok in theory if one were to go to /test they would get the index in www
+# as opposed to the one in /website...
+#	location /test {
+#		root /www;
+#	}
+
 }

memo.txt

@@ -1,15 +1,27 @@
-
+IN 42 SUBJECT AND/OR PRIORITY :
+- CGI
+- chunked request (response not mandatory it seems)
+- fix need for index and autoindex config
+- Ecrire des tests !
+- "root" need to replace "location->path" part of "client.path"
+	replace up to the last "/" of the "location->path" part
+	(if its a folder this will be in fact the entire path)
+- handle redirection (Work, but weird behavior need deeper test)
+- upload files with config "upload_dir"
+- _determine_location() review (New version to complete and test)
+- replace atoi() with a better function to avoid overflow
+	like strtol : https://www32.cplusplus.com/reference/cstdlib/strtol/
+-----------------------------
+- 408 Request Timeout
+- gerer le champ "Accept" du client
+- gerer les ".." dans un URL (verifier que l'on ne sort pas du dossier "root")
 - do correct handling of special character in url (/rfc2119_files/errata.js.t%C3%A9l%C3%A9chargement -> /rfc2119_files/errata.js.téléchargement)
-- handle redirection
 - maybe add a "last_action_time" in Client for timeout handling
 	little global timeout on epoll, like 100ms, then find client that actualy need to timeout
 	if (actual_time - client.last_action_time > 10000ms){timeout(client)}
 - add headers "Date" and "Last-Modified" to response
 - change "std::string" to reference "std::string &" in most functions
 and add "const" if apropriate.
-- http_method en mode binary flags. "std::vector<http_method> allow_methods" -> "unsigned int allow_methods;"
-- Dans le parsing, trier les "locations" par ordre de precision.
-Compter les "/" dans le chemin, les locations avec le plus de "/" seront en premier dans le vector.
 - Il faut vérifier le path de la requête, voir si le serveur est bien censé délivrer cette ressource et si le client y a accès, avant d'appeler le CGI.
 
 

srcs/Client.cpp

@@ -1,29 +1,34 @@
 
 #include "Client.hpp"
 
-char	Client::buf[MAX_FILESIZE+1];
-
-
 /*********************************************
  * COPLIENS
  *********************************************/
 
 Client::Client()
-	: body_size(0)
-	, status(0)
-	, _fd(0)
+	: status(0),
+	header_complete(false),
+	read_body_size(0),
+	assigned_server(NULL),
+	assigned_location(NULL),
+	_fd(0),
+	_port(""),
+	_ip(""),
+	_lsocket(NULL)
 {
 	return;
 }
 
-//Client::Client(int afd, listen_socket *lsocket, struct sockaddr_in addr)
 Client::Client(int afd, listen_socket *lsocket, std::string aport, std::string aip)
-	: body_size(0)
-	, status(0)
-	, _fd(afd)
-	, _port(aport)
-	, _ip(aip)
-	, _lsocket(lsocket)
+	: status(0),
+	header_complete(false),
+	read_body_size(0),
+	assigned_server(NULL),
+	assigned_location(NULL),
+	_fd(afd),
+	_port(aport),
+	_ip(aip),
+	_lsocket(lsocket)
 {
 	return;
 }
@@ -32,24 +37,26 @@ Client::~Client() {
 	return;
 }
 
-// WIP placeholder because of const values
+/* // WIP not sure fo what is more logic here
 Client::Client( Client const & src )
-	: body_size(0)
-	, status(0)
-	, _fd		( src.get_cl_fd() )
-	, _port		( src.get_cl_port() )
-	, _ip		( src.get_cl_ip() )
-	, _lsocket	( src.get_cl_lsocket() )
+	: status			( src.status ),
+	header_complete		( src.header_complete ),
+	read_body_size		( src.read_body_size ),
+	assigned_server		( src.assigned_server ),
+	assigned_location	( src.assigned_location ),
+	_fd					( src._fd ),
+	_port				( src._port ),
+	_ip					( src._ip ),
+	_lsocket			( src._lsocket )
 {
 	raw_request = src.raw_request;
 	response = src.response;
 //	buf = strdup(src.buf); // TODO: this doesn't work
-	body_size = src.body_size;
-	status = src.status;
 	return;
 }
+ */
 
-// WIP placeholder because of const values
+/* // WIP placeholder because of const values
 Client &	Client::operator=( Client const & rhs )
 {
 	if ( this != &rhs )
@@ -58,7 +65,7 @@ Client &	Client::operator=( Client const & rhs )
 	}
 	return *this;
 }
-
+ */
 
 /*********************************************
  * PUBLIC MEMBER FUNCTIONS
@@ -68,26 +75,39 @@ Client &	Client::operator=( Client const & rhs )
 // https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers
 // 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()
+void	Client::parse_request_headers(std::vector<ServerConfig> &servers)
 {
-	std::string							sub;
-	std::vector<std::string>			list;
-	size_t								pos;
-
+	clear_request(); // not mandatory
+	_parse_request_line();
+	if (status)
+		return;
+	_parse_request_fields();
 // DEBUG
-std::cout	<< "\nREQUEST _____________________\n"
-			<< raw_request
-			<< "\n\n";
+print_client("headers");
+	if (status)
+		return;
+	assigned_server = ::_determine_process_server(this, servers);
+	assigned_location = ::_determine_location(*assigned_server, _request.abs_path);
+	_check_request_errors();
+	if (status)
+		return;
+	_parse_port_hostname(this->get_rq_headers("Host")); // use getter for headers because it works case insensitive
 
-	pos = (raw_request).find(CRLF CRLF);
-	sub = (raw_request).substr(0, pos);
-	list = split(sub, '\n');
-	_parse_request_line(*list.begin());
-	list.erase(list.begin());
-	_parse_request_headers(list);
-	_parse_request_body(pos +  4);
-	_parse_port_hostname(this->get_rq_headers("Host"));
-	raw_request.clear();
+//	dont clear raw_request, we need it for future reparsing of body
+//	see call of parse_request() in _read_request()
+	// raw_request.clear(); 
+}
+
+void	Client::parse_request_body()
+{
+	size_t	pos;
+
+	pos = raw_request.find(CRLF CRLF);
+	pos += std::string(CRLF CRLF).size();
+	_request.body = raw_request.substr(pos);
+
+	if (_request.body.size() > assigned_server->client_body_limit)
+		status = 413; // HTTP Client Errors
 }
 
 bool	Client::fill_script_path(std::string script)
@@ -111,9 +131,12 @@ bool	Client::fill_script_path(std::string script)
 void	Client::clear()
 {
 	clear_request();
+	header_complete = false;
+	read_body_size = 0;
+	assigned_server = NULL;
+	assigned_location = NULL;	
 	raw_request.clear();
 	response.clear();
-	body_size = 0;
 	status = 0;
 }
 
@@ -137,16 +160,43 @@ void	Client::clear_script()
 	_request.script.info.clear();
 }
 
+// debug
+void	Client::print_client(std::string message)
+{
+	std::map<std::string, std::string>::iterator	it;
+
+	std::cout << "\n=== DEBUG PRINT CLIENT ===\n";
+	std::cout << message << ":\n----------\n\n" << "raw_request:\n__\n";
+	::print_special(raw_request);
+	std::cout << "\n__\n"
+		<< "get_cl_fd()          : [" << get_cl_fd()			<< "]\n"
+		<< "get_cl_port()        : [" << get_cl_port()			<< "]\n"
+		<< "get_cl_ip()          : [" << get_cl_ip()			<< "]\n"
+		<< "get_rq_method_str()  : [" << get_rq_method_str()	<< "]\n"
+		<< "get_rq_uri()         : [" << get_rq_uri()			<< "]\n"
+		<< "get_rq_abs_path()    : [" << get_rq_abs_path()		<< "]\n"
+		<< "get_rq_query()       : [" << get_rq_query()			<< "]\n"
+		<< "get_rq_version()     : [" << get_rq_version()		<< "]\n"
+		<< "get_rq_body()        : [" << get_rq_body()			<< "]\n"
+		<< "get_rq_port()        : [" << get_rq_port()			<< "]\n"
+		<< "get_rq_hostname()    : [" << get_rq_hostname()		<< "]\n"
+		<< "get_rq_script_path() : [" << get_rq_script_path()	<< "]\n"
+		<< "get_rq_script_info() : [" << get_rq_script_info()	<< "]\n"
+		<< "headers              :\n";
+	for (it = _request.headers.begin(); it != _request.headers.end(); it++)
+		std::cout << "   " << it->first << ": [" << it->second << "]\n";
+	std::cout << "\n=== END PRINT CLIENT ===\n\n";
+}
 
 /*********************************************
  * GETTERS
  *********************************************/
 
 // client side
-int					 Client::get_cl_fd()		const { return _fd; }
-const std::string	&Client::get_cl_ip() 		const { return _ip; }
-const std::string	&Client::get_cl_port()		const { return _port; }
-const listen_socket *Client::get_cl_lsocket()	const { return _lsocket; }
+int					 Client::get_cl_fd()	const { return _fd; }
+const std::string	&	Client::get_cl_ip() 		const { return _ip; }
+const std::string	&	Client::get_cl_port()		const { return _port; }
+const listen_socket *	Client::get_cl_lsocket()	const { return _lsocket; }
 
 // requette
 http_method 	Client::get_rq_method()		const { return _request.method; }
@@ -176,30 +226,25 @@ std::string		Client::get_rq_headers(const std::string & key) const
  * PRIVATE MEMBER FUNCTIONS
  *********************************************/
 
-void	Client::_parse_request_line( std::string rline )
+void	Client::_parse_request_line()
 {
-	std::vector<std::string>	sline;
-	std::string					tmp;
+	std::vector<std::string>	line;
+	std::string					raw_line;
 
-	sline = split(rline, ' ');
-	if (sline.size() != 3)
+	raw_line = ::get_line(raw_request, 0, CRLF);
+	line = ::split_trim(raw_line, " ");
+	if (line.size() != 3)
 	{
-		std::cerr << "err _parse_request_line(): ";
-		throw std::runtime_error("bad request-line header");
+		std::cerr << "err _parse_first_line(): wrong number of elements (" << line.size() << " instead of 3)\n";
+		status = 400; // "bad request"
+	}
+	else
+	{
+		_request.method = str_to_http_method(line[0]);
+		_request.uri = line[1];
+		_parse_request_uri(line[1]);
+		_request.version = line[2];
 	}
-	// method
-	tmp = ::trim(sline[0], ' ');
-	tmp = ::trim(tmp, '\r');
-	_request.method = str_to_http_method(tmp);
-	// uri
-	tmp = ::trim(sline[1], ' ');
-	tmp = ::trim(tmp, '\r');
-	_request.uri = tmp;
-	_parse_request_uri( tmp );
-	// http version
-	tmp = ::trim(sline[2], ' ');
-	tmp = ::trim(tmp, '\r');
-	_request.version = tmp;
 }
 
 void	Client::_parse_request_uri( std::string uri )
@@ -214,32 +259,32 @@ void	Client::_parse_request_uri( std::string uri )
 	_request.abs_path = uri.substr(0, pos);
 }
 
-void	Client::_parse_request_headers( std::vector<std::string> list )
+void	Client::_parse_request_fields()
 {
-	std::string							key;
-	std::string							val;
-	std::vector<std::string>::iterator	it;
-	size_t								pos;
+	std::string	headers;
+	size_t		pos;
+	int			ret;
 
-	for (it = list.begin(); it != list.end(); it++)
-	{
-		pos = (*it).find(':');
-		key = (*it).substr( 0, pos );
-		key = ::trim(key, ' ');
-		key = ::trim(key, '\r');
-		key = ::str_tolower(key);
-		val = (*it).substr( pos + 1 );
-		val = ::trim(val, ' ');
-		val = ::trim(val, '\r');
-		_request.headers.insert( std::pair<std::string, std::string>(key, val) );
+	headers = raw_request;
+	// delete first line
+	pos = headers.find(CRLF);
+	if (pos != std::string::npos)
+		headers.erase(0, pos + std::string(CRLF).size());
+	// delete body part
+	pos = headers.find(CRLF CRLF);
+	if (pos != std::string::npos)
+		headers.erase(pos);
+	else {
+		std::cerr << "err _parse_request_fields(): request header doesn't end with empty line\n";
+		status = 400; // "bad request"
 	}
-}
-
-void	Client::_parse_request_body( size_t pos )
-{
-	std::string body = &raw_request[pos];
-
-	_request.body = body;
+	// copy result of parser into headers
+	ret = ::parse_http_headers(headers, _request.headers);
+	if (ret > 0) {
+		std::cerr << "err _parse_request_fields(): " << ret << " fields are bad formated\n";
+		status = 400; // "bad request"
+	}
+	::str_map_key_tolower(_request.headers);
 }
 
 void	Client::_parse_port_hostname(std::string host)
@@ -247,10 +292,7 @@ void	Client::_parse_port_hostname(std::string host)
 	size_t	pos;
 
 	if (host == "")
-	{
 		std::cerr << "no host\n";
-		throw std::runtime_error("no host in request");
-	}
 
 	pos = host.find(':');
 	// port :
@@ -264,6 +306,41 @@ void	Client::_parse_port_hostname(std::string host)
 	_request.hostname = host.substr(0, pos);
 }
 
+void	Client::_check_request_errors()
+{
+//////////////////////
+// Request line checks
+	if (_request.method == UNKNOWN)
+		status = 501; // HTTP Client Errors
+	else if (_request.version.compare(0, sizeof("HTTP/1") - 1, "HTTP/1") != 0)
+		status = 505; // HTTP Client Errors
+	else if (!(assigned_location->allow_methods & _request.method))
+	{
+		status = 405; // HTTP Client Errors
+		response.append("Allow: ");
+		response.append(::http_methods_to_str(assigned_location->allow_methods));
+		response.append(CRLF);
+	}
+	else if (assigned_location->redirect_status)
+	{ // Weird behavior. Sometimes, the web browser seems to wait for a complete response until timeout.
+		// (for codes 301, 302, 303, 307, and 308)
+		status = assigned_location->redirect_status;
+		response.append("Location: ");
+		response.append(assigned_location->redirect_uri);
+		response.append(CRLF CRLF);
+	}
+
+	if (status)
+		return;
+
+/////////////////
+// Headers checks
+	if (!this->get_rq_headers("Content-Length").empty()
+	&& ::atoi(this->get_rq_headers("Content-Length").c_str()) > (int)assigned_server->client_body_limit)
+		status = 413; // HTTP Client Errors
+	
+	return;
+}
 
 /*********************************************
  * OVERLOAD

srcs/Client.hpp

@@ -7,9 +7,10 @@
 # include <map>
 # include <vector>
 # include <string.h>		// strdup
-# include "utils.hpp"
 # include <netinet/in.h>	// sockaddr_in, struct in_addr
 # include <arpa/inet.h>		// htonl, htons, ntohl, ntohs, inet_addr, inet_ntoa
+# include "utils.hpp"
+# include "ServerConfig.hpp"
 
 struct Script
 {
@@ -31,21 +32,22 @@ struct Request
 	struct Script						script;
 };
 
-# define MAX_FILESIZE 1000000 // (1Mo)
 class Client
 {
 	public:
 		Client();
 		Client(int afd, listen_socket *lsocket, std::string aport, std::string aip);
 		~Client();
-		Client(Client const &src);
-		Client &operator=(Client const &rhs);
+		// Client(Client const &src);
+		// Client &operator=(Client const &rhs);
 
-		std::string		raw_request;
-		std::string		response;
-		static char		buf[MAX_FILESIZE+1];
-		size_t			body_size;
-		unsigned int	status;
+		std::string				raw_request;
+		std::string				response;
+		unsigned int			status;
+		bool					header_complete;
+		size_t					read_body_size;
+		ServerConfig			*assigned_server; // cant be const cause of error_pages.operator[]
+		const LocationConfig	*assigned_location;
 
 		// getters
 		int						get_cl_fd() const;
@@ -67,30 +69,37 @@ class Client
 		std::string		get_rq_script_info() const;
 		std::string		get_rq_headers(const std::string & key) const;
 
-		void	parse_request();
+		void	parse_request_headers(std::vector<ServerConfig> &servers);
+		void	parse_request_body();
 		void	clear();
 		void	clear_request();
 		void	clear_script();
 		bool	fill_script_path(std::string script);
+		// DEBUG
+		void	print_client(std::string message = "");
 
 	private:
-		const int				_fd;
-		const std::string		_port;
-		const std::string		_ip;
-		const listen_socket *	_lsocket;
-		struct Request			_request;
+		int				_fd;
+		std::string		_port;
+		std::string		_ip;
+		listen_socket *	_lsocket;
+		struct Request	_request;
 
-		void	_parse_request_line( std::string rline );
+		void	_parse_request_line();
+		void	_parse_request_fields();
 		void	_parse_request_uri( std::string uri );
-		void	_parse_request_headers( std::vector<std::string> list );
-		void	_parse_request_body( size_t pos );
 		void	_parse_port_hostname(std::string host);
 
+		void	_check_request_errors();
 };
 
 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/cgi-bin/php-cgi

@@ -1,9 +1,10 @@
 #! /usr/bin/php
-
 <?php
+	echo   "false: 300\r\n";
+	echo   "server: Webserv/0.2\r\n";
+	echo   "Status: 300\r\n";
+	echo "\r\n";
 	echo "BEGIN PHP-CGI\n-----------\n\n";
-	//phpinfo();
-
 	echo   "AUTH_TYPE:         " . getenv("AUTH_TYPE");
 	echo "\nCONTENT_LENGTH:    " . getenv("CONTENT_LENGTH");
 	echo "\nCONTENT_TYPE:      " . getenv("CONTENT_TYPE");
@@ -26,4 +27,3 @@
 //	echo $_POST['REQUEST_METHOD'];
 	echo "\n\n-----------\nEND PHP-CGI\n\n";
 ?>
-

srcs/colors.h

@@ -0,0 +1,25 @@
+#ifndef COLORS_H
+# define COLORS_H
+
+# define GRAY "\e[0;30m"
+# define RED "\e[0;31m"
+# define GREEN "\e[0;32m"
+# define YELLOW "\e[0;33m"
+# define BLUE "\e[0;34m"
+# define PURPLE "\e[0;35m"
+# define CYAN "\e[0;36m"
+# define WHITE "\e[0;37m"
+
+# define B_GRAY "\e[1;30m"
+# define B_RED "\e[1;31m"
+# define B_GREEN "\e[1;32m"
+# define B_YELLOW "\e[1;33m"
+# define B_BLUE "\e[1;34m"
+# define B_PURPLE "\e[1;35m"
+# define B_CYAN "\e[1;36m"
+# define B_WHITE "\e[1;37m"
+
+# define RESET "\e[0m"
+
+#endif
+

srcs/config/ConfigParser.hpp

@@ -26,8 +26,9 @@
 # include <iostream>		// cout, cin
 # include <fstream>			// ifstream
 //# include <unistd.h>		// access()
-# include <dirent.h>		// opendir()
-
+# include <dirent.h>		// opendir(), doesn't work...
+# include <sys/stat.h>		// stat(), replaces opendir() don't bother with ERRNO ?
+# include <algorithm>		// sort() in Post
 
 class ConfigParser {
 
@@ -51,6 +52,13 @@ public:
 // private member functions from anywhere...
 	void	_print_content() const;
 
+
+// I don't love that this is here but...
+// doesn't work use the operator overload
+//	bool	compareLocationConfigs(const LocationConfig &a, const LocationConfig &b);
+
+
+
 private:
 	std::string		_content;
 
@@ -82,11 +90,20 @@ private:
 
 	void	_post_processing(std::vector<ServerConfig> *servers);
 
+	bool	_find_root_path_location(std::vector<LocationConfig> locations); // const?
 
 
 };
 
 
+// no idea if it should go here...
+//bool	compareLocationConfigs(const LocationConfig &a,
+//			const LocationConfig &b);
+
+
+
+
+
 // def needs work line a better name an do i even need this?
 // should it be in Utils instead?
 class MyException : public std::invalid_argument

srcs/config/ConfigParserPost.cpp

@@ -1,85 +0,0 @@
-
-
-
-#include "ConfigParser.hpp"
-
-
-
-void	ConfigParser::_post_processing(std::vector<ServerConfig> *servers)
-{
-
-	// make certain servers default
-	// fill out empty settings
-	// if special settings are empty throw
-
-	std::vector<ServerConfig>::iterator	it = servers->begin();
-
-	while (it != servers->end())
-	{
-		// host and port should already be set
-		if (it->host == "")
-			throw std::invalid_argument("Config file needs a host and port");
-
-	// is that a good default?
-		if (it->root == "")
-			it->root = "/";
-		if (it->client_body_limit == 0)
-			it->client_body_limit = 5000;	// what is the recomended size?
-
-		// autoindex should already be false by default right?
-
-		// what do we do if Allow methods is left empty?
-			// all ?
-		if (it->allow_methods.empty())
-			throw std::invalid_argument("No methods specified");
-			
-
-		// what to do if index is left empty? index.html?
-		// ok but i still need to check index, no idea how...
-
-		// if error_pages is left empty, we'll use the defaults which
-			// i believe are set elsewhere...
-
-		std::vector<LocationConfig>::iterator	it_l = it->locations.begin();
-		while (it_l != it->locations.end())
-		{
-			// check that path is feasible...
-			// opendir?
-			DIR*	dir = opendir(it_l->path.c_str());
-			if (dir)
-				closedir(dir);
-			else
-				throw std::invalid_argument("location dir could not be opened");
-
-			if (it_l->client_body_limit == 0)
-				it_l->client_body_limit = 5000;	// what is the recomended size?
-			if (it_l->root == "")
-				it_l->root = it->root;
-
-			// fill out allow methods from server?
-			if (it_l->allow_methods.empty())
-				it_l->allow_methods = it->allow_methods;
-
-			// fill out index from Server?
-			// or do a bunch of checks on what is in there...
-
-			// same for redirect status i think
-
-			// maybe do something for Cgi_info?
-
-			++it_l;
-		}
-
-
-		++it;
-	}
-
-	// do the defaults at the end?
-
-
-}
-
-
-
-
-

srcs/config/LocationConfig.hpp

@@ -6,7 +6,7 @@
 /*   By: lperrey <lperrey@student.42.fr>            +#+  +:+       +#+        */
 /*                                                +#+#+#+#+#+   +#+           */
 /*   Created: 2022/07/23 16:08:00 by me                #+#    #+#             */
-/*   Updated: 2022/08/02 14:06:07 by lperrey          ###   ########.fr       */
+/*   Updated: 2022/08/04 19:32:40 by erlazo           ###   ########.fr       */
 /*                                                                            */
 /* ************************************************************************** */
 
@@ -16,7 +16,13 @@
 # include <map>
 # include <vector>
 # include <string>
+# include <iostream>
+# include <sys/stat.h>	// stat()
 
+# include <stdio.h>		// printf(), gotta go
+
+
+# include "utils.hpp"
 
 // again, struct instead?
 class LocationConfig
@@ -24,22 +30,64 @@ class LocationConfig
 public:
 	// canonic stuff?
 
-	std::string		path;
-
-	int				client_body_limit;
-	std::string		root;
+	std::string					path;	// /path and /path/ are fine
+									// i add trailing / if a dir
+	std::string					root;
 	std::vector<std::string>	index;
-	std::vector<http_method>	allow_methods;
-	std::map<std::string, std::string>	cgi_info;
+	unsigned int				allow_methods;
+	std::vector<std::string>	cgi_ext;	// php not .php
+	bool						autoindex;
 
-	// wait if i can call several times, shouldn't it be a map?
-	// wait no there can only be 1 and i think it might have to be in
-	// location only...
-	int			redirect_status;
-	std::string	redirect_uri;
+	std::string					upload_dir;
+
+	int							redirect_status; // only in location
+	std::string					redirect_uri;	// only 1 per location
 	// au pire you do location / { return 301 http://location; }
 	// and that's how you get the redirect from the root.
 
+	void	print_all()
+	{
+		std::cout << "\nPRINTING A LOCATION\n";
+
+		std::cout << "Path: " << path << '\n';
+		std::cout << "root: " << root << '\n';
+		std::cout << "autoindex: " << autoindex << '\n';
+
+		std::cout << "Skipping index...\n";
+
+		std::cout << "Location allow_methods: ";
+		std::cout << ::http_methods_to_str(allow_methods) << "\n";
+
+		std::cout << "Skipping redirect status etc...\n";		
+
+		std::cout << "------\n";
+	}
+
+// works a lot better than  using a compare function...
+	bool operator<(const LocationConfig& rhs) const
+    {
+        int	comp_lhs = 0;
+		int	comp_rhs = 0;
+		size_t	tmp = 0;
+
+		while ((tmp = this->path.find_first_of("/", tmp)) != std::string::npos)
+		{
+			++tmp;
+			++comp_lhs;
+		}
+		if (path[path.find_last_of("/") + 1] != '\0')
+			++comp_lhs; 
+		tmp = 0;
+		while ((tmp = rhs.path.find_first_of("/", tmp)) != std::string::npos)
+		{
+			++tmp;
+			++comp_rhs;
+		}
+		if (rhs.path[rhs.path.find_last_of("/") + 1] != '\0')
+			++comp_rhs; 
+
+		return (comp_lhs < comp_rhs);	// right comparison ? not <= ?
+    };
 
 
 };

srcs/config/ServerConfig.hpp

@@ -14,48 +14,24 @@
 class ServerConfig
 {
 public:
-
 	// do i need some canonic stuff?
 
-
-	// there can be several
 	std::vector<std::string>	server_name;
 	// we could shove default in here if we wanted to...
 
-	// there can only be 1 per server...
 	std::string	host;
 	std::string	port;	// port needs to be something else... not quite an int
-	// should a Server be able to handle several?
 
-	// there can only be one.
-	std::string	root;
+	std::string	root;	// ./www/ or www work www/ and www work
+		// i do remove trailing / tho
 	
 	unsigned int	client_body_limit;	// set to default max if none set
 
-	// might be the only one we let slide if bad input...
-	bool		autoindex;
-
-	// we will check the index in the post processing with access() ?
 	std::vector<std::string>	index;
 	std::map<int, std::string>	error_pages;
 
-	// i'm tempted to do something diff for storing method types...
-	// fuck it, you can only call allow_methods once in Server
-		// once more in each location.
-	std::vector<http_method>		allow_methods;
-
 	std::vector<LocationConfig>		locations;
 
-	// not convinced we need these...
-//	struct timeval	send_timeout;
-//	struct timeval	recv_timeout;
-
-
-// fuck maybe i do need return here...
-	// wait if i can call several times, shouldn't it be a map?
-// i think actually there can only be 1 and it can only be in a location?
-//	int			redirect_status;
-//	std::string	redirect_uri;
 
 	void	print_all()
 	{
@@ -73,15 +49,12 @@ public:
 			std::cout << it->first << "--" << it->second << " ";
 //		for (size_t i = 0; i < error_pages.size(); i++)
 //			std::cout << error_pages->first << "--" << error_pages->second << " ";
-		std::cout << "\nallow_methods: ";
-		for (size_t i = 0; i < allow_methods.size(); i++)
-			std::cout << allow_methods[i] << " ";
-		std::cout << "\nskiping Locations for now...\n";
-		std::cout << "also skiping send_timeout and recv\n";
-		std::cout << "autoindex: " << autoindex << '\n';
+
+//		std::cout << "skiping Locations for now...\n";
+		for (std::vector<LocationConfig>::iterator it = locations.begin(); it < locations.end(); it++)
+			it->print_all();
+
 		std::cout << "client_body_limit: " << client_body_limit << '\n';
-	//	std::cout << "redirect_status: " << redirect_status << '\n';
-	//	std::cout << "redirect_uri: " << redirect_uri << '\n';
 		std::cout << "host: " << host << '\n';
 		std::cout << "port: " << port << '\n';
 

srcs/config/compareLocationConfigs.cpp

@@ -0,0 +1,27 @@
+
+
+// prolly get rid of this file...
+
+
+#include "LocationConfig.hpp"
+
+#include <string>
+#include <algorithm>
+
+// Ok so maybe it can't be a member functions?
+bool	compareLocationConfigs(const LocationConfig &a, const LocationConfig &b)
+{
+	int	len_a;
+	int	len_b;
+	size_t	tmp = 0;
+
+	// consider adding 1 to path that ends in a file not folder.
+
+
+	while ((tmp = a.path.find_first_of("/", tmp)) != std::string::npos)
+		++len_a;
+	tmp = 0;
+	while ((tmp = b.path.find_first_of("/", tmp)) != std::string::npos)
+		++len_b;
+	return (len_a < len_b);	// right comparison ? not <= ?
+}

srcs/config/extraConfig.cpp

@@ -19,8 +19,8 @@ std::string	ConfigParser::_pre_set_val_check(const std::string key, \
 	// there shouldn't be any tabs, right? not between values...
 	if (value.find_first_of("\t") != std::string::npos)
 	{
-		std::cout << value << "\n";
-		throw std::invalid_argument("bad config file arguments 3");
+//		std::cout << value << "\n";
+		throw std::invalid_argument("why would you put tabs between values");
 	}
 
 	size_t	i = value.find_first_of(";");
@@ -33,11 +33,6 @@ std::string	ConfigParser::_pre_set_val_check(const std::string key, \
 		throw std::invalid_argument("bad config file arguments 4");
 
 
-	// we Trim value.
-	// is this valid?
-	// would it be better to shove the result directly in tmp_val?
-		// like call substr in split?
-	//value = value.substr(0, i - 1);
 	return (value.substr(0, i));
 }
 
@@ -68,23 +63,18 @@ std::string	ConfigParser::_get_rest_of_line(size_t *curr)
 	if ((start = _content.find_first_not_of(" \t\n", *curr)) == std::string::npos)
 		throw std::invalid_argument("bad config file arguments");
 
-//	std::cout << "start + 4 = " << _content.substr(start, 4) << "\n";
-//	std::cout << "curr + 4 = " << _content.substr(*curr, 4) << "\n";
-
-
 	if ((*curr = _content.find_first_of("\n", start)) == std::string::npos)
 		throw std::invalid_argument("bad config file arguments");
 
 	std::string	values = _content.substr(start, *curr - start);
-
-//	std::cout << "rest of Line values: " << values << "\n";
-
 	return (values);
 }
 
+
+
+
 void	ConfigParser::_print_content() const
 {
 	std::cout << _content;
 }
 
-// I might need to make my own Exceptions to throw...

srcs/config/parser.cpp

@@ -31,8 +31,6 @@ ConfigParser::ConfigParser(const char*  path)
 	file.open(path);
 	if (file.is_open())
 	{
-	// are there more throws i need to add in case of errors, what would
-		// those errors be?
 		while (!file.eof())
 		{
 			getline(file, buf);
@@ -45,7 +43,7 @@ ConfigParser::ConfigParser(const char*  path)
 			}
 			else if (comment > 0 && (buf.find_first_not_of(" \t")) < comment)
 			{
-				// is there a comment at the end of the line
+				// check for comment at the end of the line
 				std::string	tmp = buf.substr(0, comment - 1);
 				_content.append(tmp + '\n');
 			}
@@ -53,7 +51,7 @@ ConfigParser::ConfigParser(const char*  path)
 		file.close();
 	}
 	else
-		throw std::invalid_argument("open config");
+		throw std::invalid_argument("failed to open config");
 }
 
 ConfigParser::~ConfigParser()
@@ -85,10 +83,11 @@ std::vector<ServerConfig> *	ConfigParser::parse()
 		throw std::invalid_argument("empty config file");
 	while (curr != std::string::npos)
 	{
-// why no checks here
-// if not here do i need them elsewhere?
-		start = _content.find_first_not_of(" \t\n", curr);
-		curr = _content.find_first_of(" \t\n", start);
+		if ((start = _content.find_first_not_of(" \t\n", curr)) == std::string::npos)
+			throw std::invalid_argument("empty config file");
+
+		if ((curr = _content.find_first_of(" \t\n", start)) == std::string::npos)
+			throw std::invalid_argument("empty config file");
 		std::string	key = _content.substr(start, curr - start);
 		if (key != "server")
 			throw std::invalid_argument("bad config file arguments 1");
@@ -104,13 +103,12 @@ ServerConfig	ConfigParser::_parse_server(size_t *start)
 	size_t			curr = _content.find_first_not_of(" \t\n", *start);
 
 	ret.client_body_limit = 0;
-	ret.autoindex = false;
 	if (curr == std::string::npos || _content[curr] != '{')
 		throw std::invalid_argument("bad config file syntax 1");
 
-	curr = _content.find_first_of(" \t\n", curr + 1);
-//	if (curr == std::string::npos)	// are there other things to check for?
-//		throw std::invalid_argument("bad config file syntax");
+	if ((curr = _content.find_first_of(" \t\n", curr + 1)) == std::string::npos)
+		throw std::invalid_argument("bad config file syntax");
+	// are there other things to check for?
 	while (curr != std::string::npos)	// here curr == { + 1
 	{
 	// so this moves curr to past the word...
@@ -142,9 +140,14 @@ LocationConfig	ConfigParser::_parse_location(size_t *start)
 	size_t			curr = *start;
 		// start is after the 1st word aka "location"
 
-	ret.client_body_limit = 0;
+	ret.autoindex = false;
 	ret.redirect_status = 0;
+	ret.allow_methods = 0;
+
 	ret.path = _get_first_word(&curr);
+	if (ret.path[0] != '/')
+		throw std::invalid_argument("Location path require a leading /");
+	//	ret.path.insert(0, "/");
 	// in theory now curr should be right after the "path"
 
 	curr = _content.find_first_not_of(" \t\n", curr);
@@ -152,9 +155,8 @@ LocationConfig	ConfigParser::_parse_location(size_t *start)
 	if (curr == std::string::npos || _content[curr] != '{')
 		throw std::invalid_argument("bad config file syntax 2");
 
-	curr = _content.find_first_of(" \t\n", curr + 1);
-//	if (curr == std::string::npos)	// are there other things to check for?
-//		throw std::invalid_argument("bad config file syntax");
+	if ((curr = _content.find_first_of(" \t\n", curr + 1)) == std::string::npos)
+		throw std::invalid_argument("bad config file syntax");
 	while (curr != std::string::npos)
 	{
 	// so this moves curr to past the word...
@@ -177,11 +179,11 @@ LocationConfig	ConfigParser::_parse_location(size_t *start)
 
 
 
-
-
+// should i be sending pointers or references?
 void	ConfigParser::_set_server_values(ServerConfig *server, \
 			const std::string key, std::string value)
 {
+// should i be sending pointers or references?
 	value = _pre_set_val_check(key, value);
 
 	std::vector<std::string> tmp_val = ::split(value, ' ');
@@ -189,11 +191,12 @@ void	ConfigParser::_set_server_values(ServerConfig *server, \
 
 	if (size < 1)
 		throw std::invalid_argument("missing value");
-	else if (key == "server_name" && size == 1)
+	else if (key == "server_name" && server->server_name.empty())
 	{
-		for (size_t i = 0; i < server->server_name.size(); i++)
+		for (std::vector<std::string>::iterator it = server->server_name.begin(); \
+				it < server->server_name.end(); it++)
 		{
-			if (server->server_name[i].compare(tmp_val[0]) == 0)
+			if (it->compare(tmp_val[0]) == 0)
 				throw std::invalid_argument("server_name already exists");
 		}
 		server->server_name.push_back(tmp_val[0]);
@@ -203,7 +206,6 @@ void	ConfigParser::_set_server_values(ServerConfig *server, \
 	{
 		if (tmp_val[0].find_first_of(":") == std::string::npos)
 		{
-			// should i limit which ports can be used?
 			if (!::isNumeric(tmp_val[0]))
 				throw std::invalid_argument("bad port number");
 			server->host = "0.0.0.0";
@@ -229,19 +231,11 @@ void	ConfigParser::_set_server_values(ServerConfig *server, \
 	}
 	else if (key == "root" && size == 1 && server->root == "")
 	{
-		DIR* dir = opendir(tmp_val[0].c_str());
-		if (dir)
-		    closedir(dir);
-		else
-			throw std::invalid_argument("root dir could not be opened");
+		// remove trailing /
+		if (tmp_val[0][tmp_val[0].size() - 1] == '/')
+			tmp_val[0].erase(tmp_val[0].size() - 1, 1);
 		server->root = tmp_val[0];
 	}
-	else if (key == "autoindex" && size == 1)
-	{
-		// autoindex is a bool, there's no good way for me to see if it has
-		// bet set already
-		server->autoindex = (tmp_val[0] == "on" ? true : false);
-	}
 	else if (key == "client_body_limit" && size == 1 \
 			&& server->client_body_limit == 0)
 	{
@@ -251,69 +245,32 @@ void	ConfigParser::_set_server_values(ServerConfig *server, \
 	}
 	else if (key == "index")
 	{
-		// i think you can call index several times...
-		// should i be doing an access?
-	// since index is at the root, but root might not yet be defined
-	// will check index later in post
 		for (unsigned long i = 0; i != tmp_val.size(); i++)
 			server->index.push_back(tmp_val[i]);
 	}
-	else if (key == "allow_methods" && server->allow_methods.empty())
-	{
-		for (unsigned long i = 0; i != tmp_val.size(); i++)
-		{
-			http_method	m = ::str_to_http_method(tmp_val[i]);
-			if (m == UNKNOWN)
-				throw std::invalid_argument("not a valid method");
-			server->allow_methods.push_back(m);
-		}
-	}
 	else if (key == "error_page")
 	{
-
-	// so it can either be just a /here/is/the/repo
-	// or it can be http://some_domain.com/here
-	// wtf... how should we handle...
-
-
-
-		// you can definitely call Error_pages several times, i think
 		std::string	path = tmp_val[tmp_val.size() - 1];
 		for (unsigned long i = 0; i != tmp_val.size() - 1; i++)
 		{
-			// what are the bounds for Error codes? 
-			if (!(isNumeric_btw(0, 600, tmp_val[i])))
-				throw std::invalid_argument("value not a valid number");
+			if (!(isNumeric_btw(400, 599, tmp_val[i])))
+				throw std::invalid_argument("invalid error code");
 			int	status_code = atoi(tmp_val[i].c_str());
-
-	// yea cuz here we continue.. why suddenly flexible not throw ?
 			if (server->error_pages.find(status_code) != server->error_pages.end())
-				continue ;
+				throw std::invalid_argument("redeclaring error page");
 			server->error_pages[status_code] = path;
 		}
 	}
-/*	else if (key == "recv_timeout" && size == 1 && server->server_name == "")
-	{
-		// what is tv_sec and do i need it?
-// ok so i don't fully understand this part but ok, keep for now...
-		server->recv_timeout.tv_sec = atoi(tmp_val[0].c_str());
-	}
-	else if (key == "send_timeout" && size == 1 && server->server_name == "")
-	{
-		server->send_timeout.tv_sec = atoi(tmp_val[0].c_str());
-	}
-*/	else
-	{
-		// means either you didn't write the right key, or the value is
-		// missing, or the value has already been filled.
+	else
 		throw std::invalid_argument("bad key value pair");
-	}
 }
 
 
+// should i be sending pointers or references?
 void	ConfigParser::_set_location_values(LocationConfig *location, \
 			const std::string key, std::string value)
 {
+// should i be sending pointers or references?
 	value = _pre_set_val_check(key, value);
 
 	std::vector<std::string> tmp_val = ::split(value, ' ');
@@ -323,69 +280,64 @@ void	ConfigParser::_set_location_values(LocationConfig *location, \
 		throw std::invalid_argument("missing value");
 	else if (key == "root" && size == 1 && location->root == "")
 	{
-		DIR* dir = opendir(tmp_val[0].c_str());
-		if (dir)
-		    closedir(dir);
-		else
-			throw std::invalid_argument("root dir could not be opened");
+		// remove trailing /
+		if (tmp_val[0][tmp_val[0].size() - 1] == '/')
+			tmp_val[0].erase(tmp_val[0].size() - 1, 1);
 		location->root = tmp_val[0];
 	}
-	else if (key == "client_body_limit" && size == 1 \
-			&& location->client_body_limit == 0)
-	{
-		if (!::isNumeric(tmp_val[0]))
-			throw std::invalid_argument("client_body_limit not a number");
-		location->client_body_limit = atoi(tmp_val[0].c_str());
-	}
+	else if (key == "autoindex" && size == 1)
+		location->autoindex = (tmp_val[0] == "on" ? true : false);
 	else if (key == "index")
 	{
-		// you can definitely call Index several times, i think
 		for (unsigned long i = 0; i != tmp_val.size(); i++)
 			location->index.push_back(tmp_val[i]);
 	}
-	else if (key == "allow_methods" && location->allow_methods.empty())
+	else if (key == "allow_methods" && location->allow_methods == 0)
 	{
 		for (unsigned long i = 0; i != tmp_val.size(); i++)
 		{
 			http_method	m = ::str_to_http_method(tmp_val[i]);
 			if (m == UNKNOWN)
 				throw std::invalid_argument("not a valid method");
-			location->allow_methods.push_back(m);
+			location->allow_methods |= m;
 		}
 	}
-	else if (key == "cgi_info")
+	else if (key == "cgi_ext")
 	{
-		// you can call cgi_info several times i think.
-// ok wtf is all this even doing, figure that out
-		unsigned long	i = value.find_first_of(" ");
-		if (i == std::string::npos)
-			throw std::invalid_argument("bad config file arguments 8");
-	// ok why an int now, we gotta be more consistent!
-		int	j = value.find_first_not_of(" ", i);
-		location->cgi_info[value.substr(0, i)] = value.substr(j, value.length());
+		for (size_t i = 0; i < tmp_val.size(); i++)
+		{
+			if (tmp_val[i][0] == '.')
+				throw std::invalid_argument("cgi_ext should not have a leading '.'");
+			location->cgi_ext.push_back(tmp_val[i]);
+		}
 	}
-	else if (key == "return" && location->redirect_status == 0 \
+	else if (key == "redirect" && location->redirect_status == 0 \
 			&& location->redirect_uri == "")
 	{
-// actually i think there can only be one per location...
-		// you can definitely call return several times, i think
 		if (tmp_val.size() != 2)
 			throw std::invalid_argument("wrong number of values");
-		// and tmp_val[0] should be a number and tmp_val[1] a string?
-		if (!(::isNumeric(tmp_val[0])))
-			throw std::invalid_argument("value not a number");
+		if (tmp_val[0] != "301" && tmp_val[0] != "302" 
+			&& tmp_val[0] != "303" && tmp_val[0] != "307"
+			&& tmp_val[0] != "308")
+			throw std::invalid_argument("bad redirect status");
+		std::cout << tmp_val[1] << '\n';
+		if (tmp_val[1].compare(0, 7, "http://")
+			&& tmp_val[1].compare(0, 8, "https://"))
+			throw std::invalid_argument("bad redirect uri");
 
-	// somehow check that tmp_val[1] is a string? or valid? how?
-		// something about using access() to see if 
 		location->redirect_status = atoi(tmp_val[0].c_str());
 		location->redirect_uri = tmp_val[1];
 	}
-	else
+	else if (key == "upload_dir" && size == 1 && location->upload_dir == "")
 	{
-		// means either you didn't write the right key, or the value is
-		// missing, or the value has already been filled.
-		throw std::invalid_argument("bad key value pair");
+		// what checks to do?
+		// add trailing /
+		if (tmp_val[0][tmp_val[0].size() - 1] != '/')
+			tmp_val[0].push_back('/');
+		location->upload_dir = tmp_val[0];
 	}
+	else
+		throw std::invalid_argument("bad key value pair");
 }
 
 

srcs/config/postProcessing.cpp

@@ -0,0 +1,88 @@
+
+
+
+#include "ConfigParser.hpp"
+
+void	ConfigParser::_post_processing(std::vector<ServerConfig> *servers)
+{
+	std::vector<ServerConfig>::iterator	it = servers->begin();
+
+	while (it != servers->end())
+	{
+		// host and port are Mandatory
+		if (it->host == "")
+			throw std::invalid_argument("Config file needs a host and port");
+
+		// root is mandatory
+		if (it->root == "")
+			throw std::invalid_argument("Config file needs a root");
+
+		// index is mandatory in Server
+		if (it->index.empty())
+			throw std::invalid_argument("Config file needs an Index");
+
+		if (it->client_body_limit == 0)
+			it->client_body_limit = 5000;	// what is the recomended size?
+		
+
+		// if error_pages is left empty, we'll use the defaults which
+			// i believe are set elsewhere...
+
+
+		if (!_find_root_path_location(it->locations))
+		{
+			LocationConfig	tmp;
+
+			tmp.path = "/";
+			tmp.root = it->root;
+			tmp.index = it->index;
+			tmp.allow_methods = ANY_METHODS;
+			tmp.autoindex = false;
+			tmp.redirect_status = 0;
+			it->locations.push_back(tmp);
+		}
+
+		std::vector<LocationConfig>::iterator	it_l = it->locations.begin();
+
+		while (it_l != it->locations.end())
+		{
+			if (it_l->root == "")
+				it_l->root = it->root;
+
+			if (it_l->allow_methods == UNKNOWN)
+				it_l->allow_methods = ANY_METHODS;
+
+			if (it_l->index.empty())
+				it_l->index = it->index;
+
+	// nothing to be done for cgi_ext, error_pages, redirect
+
+			if (path_is_valid(it_l->root + it_l->path) == 1 \
+				&& it_l->path[it_l->path.size() - 1] != '/')
+				it_l->path.push_back('/');
+
+			++it_l;
+		}
+		std::sort(it->locations.begin(), it->locations.end());
+		std::reverse(it->locations.begin(), it->locations.end());
+
+		++it;
+	}
+}
+
+bool	ConfigParser::_find_root_path_location(std::vector<LocationConfig> locations)
+{
+	std::vector<LocationConfig>::iterator it = locations.begin();
+
+	while (it != locations.end())
+	{
+		if (it->path.compare("/") == 0)
+		{
+	//		std::cout << "in compare: " << it->path << "  --  ";
+			return true;
+		}
+		++it;
+	}
+	return false;
+}
+

srcs/main.cpp

@@ -44,7 +44,7 @@ int	main(int ac, char **av)
 	}
 	catch (std::exception& e)
 	{
-		std::cout << e.what() << '\n';
+		std::cerr << e.what() << '\n';
 	}
 	
 	return (0);

srcs/utils.cpp

@@ -1,6 +1,22 @@
 
 #include "utils.hpp"
 
+void	throw_test()
+{
+	static int i = 0;
+	++i;
+	if (i % 8 == 0)
+		throw std::bad_alloc();
+}
+
+// notice : the use of getline make it such as
+// it doesn't identify multiple delim as one :
+// " something \n else "   -> 1 - something
+//                            2 - else
+// is not the same as :
+// " something \n\n else " -> 1 - something
+//                            2 - 
+//                            3 - else
 std::vector<std::string> split(std::string input, char delimiter)
 {
     std::vector<std::string>	answer;
@@ -13,14 +29,62 @@ std::vector<std::string> split(std::string input, char delimiter)
     return answer;
 }
 
-std::string	trim(std::string str, char c)
+std::vector<std::string>
+	split_trim(std::string input, std::string delim, char ctrim)
 {
-	str = str.substr(str.find_first_not_of(c));
-	str = str.substr(0, str.find_last_not_of(c) + 1);
+	std::vector<std::string>	split_str;
+	std::string					tmp;
+	size_t						start = 0;
+	size_t						end = 0;
+	size_t						len = 0;
+
+	while (end != std::string::npos)
+	{
+		end = input.find(delim, start);
+		len = end - start;
+		if (end == std::string::npos)
+			len = end;
+		tmp = input.substr(start, len);
+		if (ctrim != '\0')
+			tmp = trim(tmp, ctrim);
+		if (tmp.size() != 0)
+			split_str.push_back( tmp );
+		start = end + delim.size();
+	}
+	return split_str;
+}
+
+std::string	trim(std::string str, char del)
+{
+	size_t	pos;
+
+	// delete leadings del
+	pos = str.find_first_not_of(del);
+	if (pos == std::string::npos)
+		pos = str.size();
+	str = str.substr(pos);
+
+	// delete trailing del
+	pos = str.find_last_not_of(del);
+	if (pos != std::string::npos)
+		str = str.substr(0, pos + 1);
 
 	return str;
 }
 
+//// trim a set of char
+//std::string	trim(std::string str, std::string del)
+//{
+//	std::string	new_str;
+//
+//	while (new_str.compare(str) != 0)
+//	{
+//		for (size_t i = 0; i < del.size(); i++)
+//			trim(str, del[i]);
+//	}
+//	return str;
+//}
+
 std::string	itos(int n)
 {
 	std::stringstream	strs;
@@ -85,7 +149,37 @@ std::string	http_methods_to_str(unsigned int methods)
 	return (str);
 }
 
-void	replace_all_substr(std::string &str, const std::string &ori_substr, const std::string &new_substr)
+# include <iostream>
+
+// you could make this &path...
+int	path_is_valid(std::string path)
+{
+	const char *tmp_path = path.c_str();
+	struct stat s;
+
+	if (stat(tmp_path, &s) == 0)
+	{
+		if (S_ISREG(s.st_mode))
+		{
+//			std::cout << "is a file\n";
+			return (2);
+		}
+		else if (S_ISDIR(s.st_mode))
+		{
+//			std::cout << "is a Dir\n";
+			return (1);
+		}
+	}
+//	std::cout << "path is neither dir nor file\n";
+	return (0);
+}
+
+
+void
+	replace_all_substr(
+		std::string &str,
+		const std::string &ori_substr,
+		const std::string &new_substr)
 {
 	if (ori_substr.empty())
 		return;
@@ -100,15 +194,126 @@ void	replace_all_substr(std::string &str, const std::string &ori_substr, const s
 	}
 }
 
-bool	operator==(const listen_socket& lhs, int fd)
-	{ return lhs.fd == fd; }
-
-bool	operator==(int fd, const listen_socket& rhs)
-	{ return fd == rhs.fd; }
-
 std::string	str_tolower(std::string str)
 {
 	std::transform(str.begin(), str.end(), str.begin(), ::tolower);
 	return str;
 }
 
+// identify a line in a string, by delim (ex. '\n')
+// delete this line from the string
+// and return the deleted line
+std::string
+	extract_line(std::string & str, size_t pos, std::string delim)
+{
+	std::string	del_str;
+	size_t		begin;
+	size_t		end;
+	size_t		len;
+
+	begin = str.rfind(delim, pos);
+	if (begin == std::string::npos)
+		begin = 0;
+	else
+		begin += delim.size();
+
+	end = str.find(delim, pos);
+	len = end;
+	if (end != std::string::npos)
+		len = end - begin;
+
+	del_str = str.substr(begin, len);
+	str.erase(begin, len);
+	return del_str;
+}
+
+// get a line in a string, by delim
+// same as extract, except it doesn't delete it
+std::string    get_line(std::string str, size_t pos, std::string delim)
+{
+	std::string	ret;
+
+	ret = ::extract_line(str, pos, delim);
+	return ret;
+}
+
+size_t
+	parse_http_headers (
+		std::string headers,
+		std::map<std::string, std::string> & fields )
+{
+	std::vector<std::string>			list;
+	std::vector<std::string>::iterator	it;
+	std::vector<std::string>::iterator	it_end;
+	size_t								err = 0;
+	size_t								pos;
+	std::string							key;
+	std::string							val;
+
+	list = ::split_trim(headers, CRLF, ' ');
+
+	it_end = list.end();
+	for (it = list.begin(); it != it_end; it++)
+	{
+		pos = (*it).find(':');
+		if (pos == std::string::npos)
+		{
+			err++;
+			continue;
+		}
+		key = (*it).substr(0, pos);
+		if ( key.find(' ') != std::string::npos )
+		{
+			err++;
+			continue;
+		}
+//		bad idea, in cgi we need to have the original value
+//		key = ::str_tolower(key); // to make "key" case_insensitive
+		val = (*it).substr(pos + 1);
+		val = ::trim(val, ' ');
+		fields.insert( std::pair<std::string, std::string>(key, val) );
+	}
+	return err;
+}
+
+void	str_map_key_tolower(std::map<std::string, std::string> & mp)
+{
+	std::map<std::string, std::string>				new_mp;
+	std::map<std::string, std::string>::iterator	it;
+	std::string										key;
+	std::string										value;
+
+	for (it = mp.begin(); it != mp.end(); it++)
+	{
+		key = it->first;
+		value = it->second;
+		key = ::str_tolower(key);
+		new_mp.insert( std::pair<std::string, std::string>(key, value) );
+	}
+	mp.swap(new_mp);
+}
+
+// DEBUG
+void	print_special(std::string str)
+{
+	char	c;
+
+	for (size_t i = 0; i < str.size(); i++)
+	{
+		c = str[i];
+		if (c == '\r')
+			std::cout << YELLOW << "\\r" << RESET;
+		else if (c == '\n')
+			std::cout << YELLOW << "\\n" << RESET << "\n";
+		else
+			std::cout << c;
+		fflush(stdout);
+	}
+}
+
+bool	operator==(const listen_socket& lhs, int fd)
+	{ return lhs.fd == fd; }
+
+bool	operator==(int fd, const listen_socket& rhs)
+	{ return fd == rhs.fd; }
+

srcs/utils.hpp

@@ -3,25 +3,20 @@
 # define UTILS_HPP
 
 # include <vector>
+# include <map>
 # include <string>
 # include <sstream>
 # include <cstdlib>		// atoi
+# include <sys/stat.h>  // stat()
 # include <cctype>		// tolower
 # include <algorithm>	// transform
+# include <cstdio>		// fflush
+# include "colors.h"
 
 # define CR "\r"
 # define LF "\n"
 # define CRLF CR LF
 
-// enum http_method
-// {
-// 	UNKNOWN	= 0b00000000,
-// 	GET		= 0b00000001,
-// 	POST	= 0b00000010,
-// 	DELETE	= 0b00000100,
-//	ANY_METHODS	= 0b11111111,
-// };
-
 enum http_method
 {
 	UNKNOWN = 0b0,
@@ -29,6 +24,8 @@ enum http_method
 	POST = 1 << 1,
 	DELETE = 1 << 2,
 	ANY_METHODS = 0b11111111,
+//	ALL_METHODS = 0b11111111,
+	// i would prefer this...
 };
 
 struct	listen_socket
@@ -41,13 +38,22 @@ bool	operator==(const listen_socket& lhs, int fd);
 bool	operator==(int fd, const listen_socket& rhs);
 
 std::vector<std::string>	split(std::string input, char delimiter);
+std::vector<std::string>	split_trim(std::string input, std::string delim = "\n", char ctrim = '\0');
 bool						isNumeric(std::string str);
 bool						isNumeric_btw(int low, int high, std::string str);
 std::string					itos(int n);
-std::string					trim(std::string str, char c);
+std::string					trim(std::string str, char del);
 http_method					str_to_http_method(std::string &str);
 std::string					http_methods_to_str(unsigned int methods);
+int							path_is_valid(std::string path);
 void						replace_all_substr(std::string &str, const std::string &ori_substr, const std::string &new_substr);
 std::string					str_tolower(std::string str);
+std::string					extract_line(std::string & str, size_t pos = 0, std::string delim = "\n");
+std::string					get_line    (std::string   str, size_t pos = 0, std::string delim = "\n");
+size_t						parse_http_headers (std::string headers, std::map<std::string, std::string> & fields );
+void						str_map_key_tolower(std::map<std::string, std::string> & mp);
+void						throw_test();
+// debug
+void	print_special(std::string str);
 
 #endif

srcs/webserv/Webserv.hpp

@@ -24,20 +24,24 @@
 # include <string>			// string
 # include <cstdio>			// perror, remove
 # include <cstdlib>			// atoi (athough it's already cover by <string>)
+# include <dirent.h>		// opendir()
 
 # include "Client.hpp"
 # include "ServerConfig.hpp"
 # include "utils.hpp"
 # include "http_status.hpp"
+# include "autoindex.hpp"
 
 extern bool	g_run;
 extern int	g_last_signal;
-void	signal_handler(int signum);
+void		signal_handler(int signum);
 
 // these might only be TMP
 # define FAILURE -1
 # define SUCCESS 1
 
+# define MIME_TYPE_DEFAULT "application/octet-stream"
+
 class Webserv
 {
 	public:
@@ -70,37 +74,39 @@ class Webserv
 				_extract_infos(struct sockaddr_in addr);
 	// request.cpp
 		void	_request(Client *client);
-		void	_read_request(Client *client);
+		int		_read_request(Client *client);
 	// response.cpp
-		void	_response(Client *client);
-		void	_send_response(Client *client, ServerConfig &server);
-
-		void	_append_base_headers(Client *client);
-		void	_construct_response(Client *client, ServerConfig &server);
-		void	_process_method(Client *client, ServerConfig &server, LocationConfig &location);
-		void	_insert_status_line(Client *client);
-		void	_error_html_response(Client *client, ServerConfig &server);
-		void	_append_body(Client *client, const char *body, size_t body_size, const std::string &file_extension = "");
-	
-		void	_get(Client *client, ServerConfig &server, LocationConfig &location);
+		void					_response(Client *client);
+		int						_send_response(Client *client);
+		void					_append_base_headers(Client *client);
+		void					_construct_response(Client *client);
+		void					_process_method(Client *client);
+		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
+		void	_get(Client *client);
 		void	_get_file(Client *client, const std::string &path);
-
-		void	_post(Client *client, ServerConfig &server, LocationConfig &location);
+		void	_autoindex(Client *client, std::string &path);
+	// method_post.cpp
+		void	_post(Client *client);
 		void	_post_file(Client *client, const std::string &path);
-
-		void	_delete(Client *client, ServerConfig &server, LocationConfig &location);
+	// method_delete.cpp
+		void	_delete(Client *client);
 		void	_delete_file(Client *client, const std::string &path);
-
-		ServerConfig	&_determine_process_server(Client *client);
-		LocationConfig	&_determine_location(ServerConfig &server, std::string const &path);
-		void	_response_correction(Client *client);
 	// cgi_script.cpp
-		bool	_is_cgi(Client *client);
-		void	_exec_cgi(Client *client);
-		char**	_set_env(Client *client);
-		char*	_dup_env(std::string var, std::string val);
-		char*	_dup_env(std::string var, int i);
-		void	_exec_script(Client *client, char **env);
+		bool		_is_cgi(Client *client);
+		std::string	_exec_cgi(Client *client);
+		char**		_set_env(Client *client);
+		char*		_dup_env(std::string var, std::string val);
+		char*		_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);
 	// 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);
@@ -115,6 +121,8 @@ class Webserv
 		void	_listen(int socket_fd, unsigned int max_connections);
 		void	_init_http_status_map();
 		void	_init_mime_types_map();
+
 };
 
 #endif
+

srcs/webserv/autoindex.hpp

@@ -0,0 +1,32 @@
+
+#ifndef AUTOINDEX_HPP
+# define AUTOINDEX_HPP
+
+// # define HTML_ERROR(STATUS) "\r\n<!DOCTYPE html><html><head><title>"STATUS"</title></head><body><h1 style=\"text-align:center\">"STATUS"</h1><hr><p style=\"text-align:center\">Le Webserv/0.1</p></body></html>"
+
+# define AUTOINDEX_START \
+"<!DOCTYPE html>"\
+"<html>"\
+"<head>"\
+	"<title>Index of "
+
+# define AUTOINDEX_MID1 \
+"</title>"\
+"</head>"\
+"<body>" \
+	"<h1>Index of "
+
+# define AUTOINDEX_MID2 \
+"</h1>"\
+"<hr>"\
+"<pre>"
+
+# define AUTOINDEX_END \
+"</pre>"\
+"<hr>"\
+"</body>"\
+"</html>"
+
+
+
+#endif

srcs/webserv/cgi_script.cpp

@@ -11,16 +11,19 @@ bool	Webserv::_is_cgi(Client *client)
 	return false;
 }
 
-void	Webserv::_exec_cgi(Client *client)
+std::string	Webserv::_exec_cgi(Client *client)
 {
-	char**	env;
+	char**			env;
+	std::string		script_output;
 
 	env = _set_env(client);
-	_exec_script(client, env);
+	script_output = _exec_script(client, env);
 
 	for (int i = 0; env[i]; i++)
 		delete[] env[i];
 	delete[] env;
+
+	return script_output;
 }
 
 char*	Webserv::_dup_env(std::string var, std::string val = "")
@@ -49,7 +52,7 @@ char**	Webserv::_set_env(Client *client)
 	env[0]  = _dup_env("AUTH_TYPE");		// authentification not supported
 	env[1]  = _dup_env("CONTENT_LENGTH"		, client->get_rq_body().size());
 	env[2]  = _dup_env("CONTENT_TYPE"		, client->get_rq_headers("Content-Type"));
-	env[3]  = _dup_env("GATEWAY_INTERFACE"	, "CGI/1.0");
+	env[3]  = _dup_env("GATEWAY_INTERFACE"	, "CGI/1.1"); // https://www.rfc-editor.org/rfc/rfc3875
 	env[4]  = _dup_env("PATH_INFO"			, client->get_rq_script_info());
 	env[5]  = _dup_env("PATH_TRANSLATED");	// not supported
 	env[6]  = _dup_env("QUERY_STRING"		, client->get_rq_query());
@@ -69,7 +72,7 @@ char**	Webserv::_set_env(Client *client)
 	return env;
 }
 
-void	Webserv::_exec_script(Client *client, char **env)
+std::string	Webserv::_exec_script(Client *client, char **env)
 {
 	#define RD 0
 	#define WR 1
@@ -82,10 +85,12 @@ void	Webserv::_exec_script(Client *client, char **env)
 	pid_t			pid;
 	char			buf[CGI_BUF_SIZE]; // WIP define buffer
 	char * const *	nll = NULL;
-	std::string		response;
+	std::string		script_output;
 	std::string		body = client->get_rq_body();
 	int				fd_in[2];
 	int				fd_out[2];
+	int				save_in  = dup(STDIN_FILENO);
+	int				save_out = dup(STDOUT_FILENO);
 
 	pipe(fd_in);
 	pipe(fd_out);
@@ -99,7 +104,11 @@ void	Webserv::_exec_script(Client *client, char **env)
 		close(FD_RD_FR_CHLD);
 		dup2(FD_RD_FR_PRNT, STDIN_FILENO);
 		dup2(FD_WR_TO_PRNT, STDOUT_FILENO);
+		// DEBUG
+		std::cerr << "execve:\n";
 		execve(client->get_rq_script_path().c_str(), nll, env);
+		// for tests execve crash :
+		//execve("wrong", nll, env);
 		std::cerr << "execve crashed.\n";
 	}
 	else
@@ -113,19 +122,76 @@ void	Webserv::_exec_script(Client *client, char **env)
 		memset(buf, '\0', CGI_BUF_SIZE);
 		while (read(FD_RD_FR_CHLD, buf, CGI_BUF_SIZE - 1) > 0)
 		{
-			response += buf;
+			script_output += buf;
 			memset(buf, '\0', CGI_BUF_SIZE);
 		}
+		close(FD_RD_FR_CHLD);
+	}
+	if (script_output.empty())
+		script_output = "Status: 500\r\n\r\n";
+	
+	dup2(save_in, STDIN_FILENO);
+	dup2(save_out, STDOUT_FILENO);
+	return script_output;
+}
+
+void	Webserv::_check_script_output(Client *client, std::string & output)
+{
+	_check_script_status(client, output);
+	_check_script_fields(client, output);
+	// _check_script_empty_lines(client, output);
+	// _check_script_space_colons(client, output);
+	// _check_script_new_lines(client, output);
+}
+
+void	Webserv::_check_script_status(Client *client, std::string & output)
+{
+	size_t	pos;
+	int		status_pos;
+
+	pos = output.find("Status:");
+	if (pos != std::string::npos)
+	{
+		status_pos = pos + std::string("Status:").size();
+		client->status = atoi(output.c_str() + status_pos);
+		::extract_line(output, pos, CRLF);
+	}
+	else
+		client->status = 200;
+}
+
+void	Webserv::_check_script_fields(Client *client, std::string & output)
+{
+	std::map<std::string, std::string>				srv_fld;	// server_field
+	std::map<std::string, std::string>				scr_fld;	// script_field
+	std::map<std::string, std::string>::iterator	it_srv;
+	std::map<std::string, std::string>::iterator	it_scr;
+	std::string										tmp;
+	size_t											pos;
+
+	// put server headers in map
+	tmp = client->response;
+	pos = tmp.find(CRLF CRLF);
+	if (pos != std::string::npos)
+		tmp.erase(pos);
+	::parse_http_headers(tmp, srv_fld);
+	// put script headers in map
+	tmp = output;
+	pos = tmp.find(CRLF CRLF);
+	if (pos != std::string::npos)
+		tmp.erase(pos);
+	::parse_http_headers(tmp, scr_fld);
+	// compare both map to supress duplicates
+	for (it_srv = srv_fld.begin(); it_srv != srv_fld.end(); it_srv++)
+	{
+		for (it_scr = scr_fld.begin(); it_scr != scr_fld.end(); it_scr++)
+		{
+			if (str_tolower(it_srv->first) == str_tolower(it_scr->first))
+			{
+				pos = client->response.find(it_srv->first);
+				::extract_line(client->response, pos, CRLF);
+			}
+		}
 	}
-	if (response.empty())
-		response = "Status: 500\r\n\r\n";
-
-// DEBUG
-std::cout	<< "\n_______response_______\n"
-			<< response
-			<< "\n_____end response_____\n";
-
-	// TODO: see how this must be handled
-	client->response += response;
 }
 

srcs/webserv/close.cpp

@@ -9,10 +9,9 @@ void	Webserv::_close_client(int fd)
 		if (*it == fd)
 		{
 			// _epoll_update(fd, 0, EPOLL_CTL_DEL); // normalement superflu, DEBUG
+			std::cerr << "close fd " << fd << "\n";
 			if (::close(fd) == -1)
 				std::perror("err close()");
-			else
-				std::cerr << "close fd " << fd << "\n";
 			_clients.erase(it);
 			break;
 		}
@@ -25,10 +24,9 @@ void	Webserv::_close_all_clients()
 	while (!_clients.empty())
 	{
 		// _epoll_update(_clients.back().fd, 0, EPOLL_CTL_DEL); // normalement superflu, DEBUG
+		std::cerr << "close fd " << _clients.back().get_cl_fd() << "\n";
 		if (::close(_clients.back().get_cl_fd()) == -1)
 			std::perror("err close()");
-		else
-			std::cerr << "close fd " << _clients.back().get_cl_fd() << "\n";
 		_clients.pop_back();
 	}
 }
@@ -38,10 +36,9 @@ void	Webserv::_close_all_listen_sockets()
 	while (!_listen_sockets.empty())
 	{
 		// _epoll_update(_listen_sockets.back().fd, 0, EPOLL_CTL_DEL); // normalement superflu, DEBUG
+		std::cerr << "close fd " << _listen_sockets.back().fd << "\n";
 		if (::close(_listen_sockets.back().fd) == -1)
 			std::perror("err close()");
-		else
-			std::cerr << "close fd " << _listen_sockets.back().fd << "\n";
 		_listen_sockets.pop_back();
 	}
 }

srcs/webserv/http_status.hpp

@@ -30,6 +30,13 @@
 # define S201 "201 Created"
 # define S204 "204 No Content"
 
+# define S301 "301 Moved Permanently"
+# define S302 "302 Found"
+# define S303 "303 See Other"
+# define S304 "304 Not Modified" // unused
+# define S307 "307 Temporary Redirect"
+# define S308 "308 Permanent Redirect"
+
 # define S400 "400 Bad Request"
 # define S403 "403 Forbidden"
 # define S404 "404 Not Found"
@@ -38,5 +45,6 @@
 
 # define S500 "500 Internal Server Error"
 # define S501 "501 Not Implemented"
+# define S505 "505 HTTP Version Not Supported"
 
 #endif

srcs/webserv/init.cpp

@@ -89,133 +89,152 @@ void	Webserv::_listen(int socket_fd, unsigned int max_connections)
 
 void	Webserv::_init_http_status_map()
 {
-	_http_status[200] = S200;
-	_http_status[201] = S201;
-	_http_status[204] = S204;
+/*	"map.insert()" over "map.operator[]" :
+**	http://www.uml.org.cn/c%2B%2B/pdf/EffectiveSTL.pdf#page=93
+*/
+	typedef std::map<int, std::string>::value_type	status_pair;
 
-	_http_status[400] = S400;
-	_http_status[403] = S403;
-	_http_status[404] = S404;
-	_http_status[405] = S405;
-	_http_status[413] = S413;
+	// _http_status.insert(std::make_pair(200, S200)); // equivalent
+	_http_status.insert(status_pair(200, S200));
+	_http_status.insert(status_pair(201, S201));
+	_http_status.insert(status_pair(204, S204));
 
-	_http_status[500] = S500;
-	_http_status[501] = S501;
+	_http_status.insert(status_pair(301, S301));
+	_http_status.insert(status_pair(302, S302));
+	_http_status.insert(status_pair(303, S303));
+	_http_status.insert(status_pair(304, S304));
+	_http_status.insert(status_pair(307, S307));
+	_http_status.insert(status_pair(308, S308));
+
+	_http_status.insert(status_pair(400, S400));
+	_http_status.insert(status_pair(403, S403));
+	_http_status.insert(status_pair(404, S404));
+	_http_status.insert(status_pair(405, S405));
+	_http_status.insert(status_pair(413, S413));
+
+	_http_status.insert(status_pair(500, S500));
+	_http_status.insert(status_pair(501, S501));
 }
 
 void	Webserv::_init_mime_types_map()
 {
-	_mime_types[""]			= "application/octet-stream";
+/*	From :
+**	http://nginx.org/en/docs/http/ngx_http_core_module.html#types
+*/
+	typedef std::map<std::string, std::string>::value_type	mime_pair;
 
-	_mime_types["html"]		= "text/html";
-	_mime_types["htm"]		= "text/html";
-	_mime_types["shtml"]	= "text/html";
-	_mime_types["css"]		= "text/css";
-	_mime_types["xml"]		= "text/xml";
-	_mime_types["gif"]		= "image/gif";
-	_mime_types["jpeg"]		= "image/jpeg";
-	_mime_types["jpg"]		= "image/jpeg";
-	_mime_types["js"]		= "application/javascript";
-	_mime_types["atom"]		= "application/atom+xml";
-	_mime_types["rss"]		= "application/rss+xml";
+	_mime_types.insert(mime_pair("", MIME_TYPE_DEFAULT));
 
-	_mime_types["mml"]		= "text/mathml";
-	_mime_types["txt"]		= "text/plain";
-	_mime_types["jad"]		= "text/vnd.sun.j2me.app-descriptor";
-	_mime_types["wml"]		= "text/vnd.wap.wml";
-	_mime_types["htc"]		= "text/x-component";
+	_mime_types.insert(mime_pair("html", "text/html"));
+	_mime_types.insert(mime_pair("html", "text/html"));
+	_mime_types.insert(mime_pair("htm", "text/html"));
+	_mime_types.insert(mime_pair("shtml", "text/html"));
+	_mime_types.insert(mime_pair("css", "text/css"));
+	_mime_types.insert(mime_pair("xml", "text/xml"));
+	_mime_types.insert(mime_pair("gif", "image/gif"));
+	_mime_types.insert(mime_pair("jpeg", "image/jpeg"));
+	_mime_types.insert(mime_pair("jpg", "image/jpeg"));
+	_mime_types.insert(mime_pair("js", "application/javascript"));
+	_mime_types.insert(mime_pair("atom", "application/atom+xml"));
+	_mime_types.insert(mime_pair("rss", "application/rss+xml"));
 
-	_mime_types["png"]		= "image/png";
-	_mime_types["tif"]		= "image/tiff";
-	_mime_types["tiff"]		= "image/tiff";
-	_mime_types["wbmp"]		= "image/vnd.wap.wbmp";
-	_mime_types["ico"]		= "image/x-icon";
-	_mime_types["jng"]		= "image/x-jng";
-	_mime_types["bmp"]		= "image/x-ms-bmp";
-	_mime_types["svg"]		= "image/svg+xml";
-	_mime_types["svgz"]		= "image/svg+xml";
-	_mime_types["webp"]		= "image/webp";
+	_mime_types.insert(mime_pair("mml", "text/mathml"));
+	_mime_types.insert(mime_pair("txt", "text/plain"));
+	_mime_types.insert(mime_pair("jad", "text/vnd.sun.j2me.app-descriptor"));
+	_mime_types.insert(mime_pair("wml", "text/vnd.wap.wml"));
+	_mime_types.insert(mime_pair("htc", "text/x-component"));
 
-	_mime_types["woff"]		= "application/font-woff";
-	_mime_types["jar"]		= "application/java-archive";
-	_mime_types["war"]		= "application/java-archive";
-	_mime_types["ear"]		= "application/java-archive";
-	_mime_types["json"]		= "application/json";
-	_mime_types["hqx"]		= "application/mac-binhex40";
-	_mime_types["doc"]		= "application/msword";
-	_mime_types["pdf"]		= "application/pdf";
-	_mime_types["ps"]		= "application/postscript";
-	_mime_types["eps"]		= "application/postscript";
-	_mime_types["ai"]		= "application/postscript";
-	_mime_types["rtf"]		= "application/rtf";
-	_mime_types["m3u8"]		= "application/vnd.apple.mpegurl";
-	_mime_types["xls"]		= "application/vnd.ms-excel";
-	_mime_types["eot"]		= "application/vnd.ms-fontobject";
-	_mime_types["ppt"]		= "application/vnd.ms-powerpoint";
-	_mime_types["wmlc"]		= "application/vnd.wap.wmlc";
-	_mime_types["kml"]		= "application/vnd.google-earth.kml+xml";
-	_mime_types["kmz"]		= "application/vnd.google-earth.kmz";
-	_mime_types["7z"]		= "application/x-7z-compressed";
-	_mime_types["cco"]		= "application/x-cocoa";
-	_mime_types["jardiff"]	= "application/x-java-archive-diff";
-	_mime_types["jnlp"]		= "application/x-java-jnlp-file";
-	_mime_types["run"]		= "application/x-makeself";
-	_mime_types["pl"]		= "application/x-perl";
-	_mime_types["pm"]		= "application/x-perl";
-	_mime_types["prc"]		= "application/x-pilot";
-	_mime_types["pdb"]		= "application/x-pilot";
-	_mime_types["rar"]		= "application/x-rar-compressed";
-	_mime_types["rpm"]		= "application/x-redhat-package-manager";
-	_mime_types["sea"]		= "application/x-sea";
-	_mime_types["swf"]		= "application/x-shockwave-flash";
-	_mime_types["sit"]		= "application/x-stuffit";
-	_mime_types["tcl"]		= "application/x-tcl";
-	_mime_types["tk"]		= "application/x-tcl";
-	_mime_types["der"]		= "application/x-x509-ca-cert";
-	_mime_types["pem"]		= "application/x-x509-ca-cert";
-	_mime_types["crt"]		= "application/x-x509-ca-cert";
-	_mime_types["xpi"]		= "application/x-xpinstall";
-	_mime_types["xhtml"]	= "application/xhtml+xml";
-	_mime_types["xspf"]		= "application/xspf+xml";
-	_mime_types["zip"]		= "application/zip";
+	_mime_types.insert(mime_pair("png", "image/png"));
+	_mime_types.insert(mime_pair("tif", "image/tiff"));
+	_mime_types.insert(mime_pair("tiff", "image/tiff"));
+	_mime_types.insert(mime_pair("wbmp", "image/vnd.wap.wbmp"));
+	_mime_types.insert(mime_pair("ico", "image/x-icon"));
+	_mime_types.insert(mime_pair("jng", "image/x-jng"));
+	_mime_types.insert(mime_pair("bmp", "image/x-ms-bmp"));
+	_mime_types.insert(mime_pair("svg", "image/svg+xml"));
+	_mime_types.insert(mime_pair("svgz", "image/svg+xml"));
+	_mime_types.insert(mime_pair("webp", "image/webp"));
 
-	_mime_types["bin"]		= "application/octet-stream";
-	_mime_types["exe"]		= "application/octet-stream";
-	_mime_types["dll"]		= "application/octet-stream";
-	_mime_types["deb"]		= "application/octet-stream";
-	_mime_types["dmg"]		= "application/octet-stream";
-	_mime_types["iso"]		= "application/octet-stream";
-	_mime_types["img"]		= "application/octet-stream";
-	_mime_types["msi"]		= "application/octet-stream";
-	_mime_types["msp"]		= "application/octet-stream";
-	_mime_types["msm"]		= "application/octet-stream";
+	_mime_types.insert(mime_pair("woff", "application/font-woff"));
+	_mime_types.insert(mime_pair("jar", "application/java-archive"));
+	_mime_types.insert(mime_pair("war", "application/java-archive"));
+	_mime_types.insert(mime_pair("ear", "application/java-archive"));
+	_mime_types.insert(mime_pair("json", "application/json"));
+	_mime_types.insert(mime_pair("hqx", "application/mac-binhex40"));
+	_mime_types.insert(mime_pair("doc", "application/msword"));
+	_mime_types.insert(mime_pair("pdf", "application/pdf"));
+	_mime_types.insert(mime_pair("ps", "application/postscript"));
+	_mime_types.insert(mime_pair("eps", "application/postscript"));
+	_mime_types.insert(mime_pair("ai", "application/postscript"));
+	_mime_types.insert(mime_pair("rtf", "application/rtf"));
+	_mime_types.insert(mime_pair("m3u8", "application/vnd.apple.mpegurl"));
+	_mime_types.insert(mime_pair("xls", "application/vnd.ms-excel"));
+	_mime_types.insert(mime_pair("eot", "application/vnd.ms-fontobject"));
+	_mime_types.insert(mime_pair("ppt", "application/vnd.ms-powerpoint"));
+	_mime_types.insert(mime_pair("wmlc", "application/vnd.wap.wmlc"));
+	_mime_types.insert(mime_pair("kml", "application/vnd.google-earth.kml+xml"));
+	_mime_types.insert(mime_pair("kmz", "application/vnd.google-earth.kmz"));
+	_mime_types.insert(mime_pair("7z", "application/x-7z-compressed"));
+	_mime_types.insert(mime_pair("cco", "application/x-cocoa"));
+	_mime_types.insert(mime_pair("jardiff", "application/x-java-archive-diff"));
+	_mime_types.insert(mime_pair("jnlp", "application/x-java-jnlp-file"));
+	_mime_types.insert(mime_pair("run", "application/x-makeself"));
+	_mime_types.insert(mime_pair("pl", "application/x-perl"));
+	_mime_types.insert(mime_pair("pm", "application/x-perl"));
+	_mime_types.insert(mime_pair("prc", "application/x-pilot"));
+	_mime_types.insert(mime_pair("pdb", "application/x-pilot"));
+	_mime_types.insert(mime_pair("rar", "application/x-rar-compressed"));
+	_mime_types.insert(mime_pair("rpm", "application/x-redhat-package-manager"));
+	_mime_types.insert(mime_pair("sea", "application/x-sea"));
+	_mime_types.insert(mime_pair("swf", "application/x-shockwave-flash"));
+	_mime_types.insert(mime_pair("sit", "application/x-stuffit"));
+	_mime_types.insert(mime_pair("tcl", "application/x-tcl"));
+	_mime_types.insert(mime_pair("tk", "application/x-tcl"));
+	_mime_types.insert(mime_pair("der", "application/x-x509-ca-cert"));
+	_mime_types.insert(mime_pair("pem", "application/x-x509-ca-cert"));
+	_mime_types.insert(mime_pair("crt", "application/x-x509-ca-cert"));
+	_mime_types.insert(mime_pair("xpi", "application/x-xpinstall"));
+	_mime_types.insert(mime_pair("xhtml", "application/xhtml+xml"));
+	_mime_types.insert(mime_pair("xspf", "application/xspf+xml"));
+	_mime_types.insert(mime_pair("zip", "application/zip"));
 
-	_mime_types["docx"]		= "application/vnd.openxmlformats-officedocument.wordprocessingml.document";
-	_mime_types["xlsx"]		= "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet";
-	_mime_types["pptx"]		= "application/vnd.openxmlformats-officedocument.presentationml.presentation";
+	_mime_types.insert(mime_pair("bin", "application/octet-stream"));
+	_mime_types.insert(mime_pair("exe", "application/octet-stream"));
+	_mime_types.insert(mime_pair("dll", "application/octet-stream"));
+	_mime_types.insert(mime_pair("deb", "application/octet-stream"));
+	_mime_types.insert(mime_pair("dmg", "application/octet-stream"));
+	_mime_types.insert(mime_pair("iso", "application/octet-stream"));
+	_mime_types.insert(mime_pair("img", "application/octet-stream"));
+	_mime_types.insert(mime_pair("msi", "application/octet-stream"));
+	_mime_types.insert(mime_pair("msp", "application/octet-stream"));
+	_mime_types.insert(mime_pair("msm", "application/octet-stream"));
 
-	_mime_types["mid"]		= "audio/midi";
-	_mime_types["midi"]		= "audio/midi";
-	_mime_types["kar"]		= "audio/midi";
-	_mime_types["mp3"]		= "audio/mpeg";
-	_mime_types["ogg"]		= "audio/ogg";
-	_mime_types["m4a"]		= "audio/x-m4a";
-	_mime_types["ra"]		= "audio/x-realaudio";
+	_mime_types.insert(mime_pair("docx", "application/vnd.openxmlformats-officedocument.wordprocessingml.document"));
+	_mime_types.insert(mime_pair("xlsx", "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet"));
+	_mime_types.insert(mime_pair("pptx", "application/vnd.openxmlformats-officedocument.presentationml.presentation"));
 
-	_mime_types["3gpp"]		= "video/3gpp";
-	_mime_types["3gp"]		= "video/3gpp";
-	_mime_types["ts"]		= "video/mp2t";
-	_mime_types["mp4"]		= "video/mp4";
-	_mime_types["mpeg"]		= "video/mpeg";
-	_mime_types["mpg"]		= "video/mpeg";
-	_mime_types["mov"]		= "video/quicktime";
-	_mime_types["webm"]		= "video/webm";
-	_mime_types["flv"]		= "video/x-flv";
-	_mime_types["m4v"]		= "video/x-m4v";
-	_mime_types["mng"]		= "video/x-mng";
-	_mime_types["asx"]		= "video/x-ms-asf";
-	_mime_types["asf"]		= "video/x-ms-asf";
-	_mime_types["wmv"]		= "video/x-ms-wmv";
-	_mime_types["avi"]		= "video/x-msvideo";
+	_mime_types.insert(mime_pair("mid", "audio/midi"));
+	_mime_types.insert(mime_pair("midi", "audio/midi"));
+	_mime_types.insert(mime_pair("kar", "audio/midi"));
+	_mime_types.insert(mime_pair("mp3", "audio/mpeg"));
+	_mime_types.insert(mime_pair("ogg", "audio/ogg"));
+	_mime_types.insert(mime_pair("m4a", "audio/x-m4a"));
+	_mime_types.insert(mime_pair("ra", "audio/x-realaudio"));
+
+	_mime_types.insert(mime_pair("3gpp", "video/3gpp"));
+	_mime_types.insert(mime_pair("3gp", "video/3gpp"));
+	_mime_types.insert(mime_pair("ts", "video/mp2t"));
+	_mime_types.insert(mime_pair("mp4", "video/mp4"));
+	_mime_types.insert(mime_pair("mpeg", "video/mpeg"));
+	_mime_types.insert(mime_pair("mpg", "video/mpeg"));
+	_mime_types.insert(mime_pair("mov", "video/quicktime"));
+	_mime_types.insert(mime_pair("webm", "video/webm"));
+	_mime_types.insert(mime_pair("flv", "video/x-flv"));
+	_mime_types.insert(mime_pair("m4v", "video/x-m4v"));
+	_mime_types.insert(mime_pair("mng", "video/x-mng"));
+	_mime_types.insert(mime_pair("asx", "video/x-ms-asf"));
+	_mime_types.insert(mime_pair("asf", "video/x-ms-asf"));
+	_mime_types.insert(mime_pair("wmv", "video/x-ms-wmv"));
+	_mime_types.insert(mime_pair("avi", "video/x-msvideo"));
 }
 

srcs/webserv/method_delete.cpp

@@ -0,0 +1,40 @@
+
+#include "Webserv.hpp"
+
+void	Webserv::_delete(Client *client)
+{
+/*
+	WIP
+	https://www.rfc-editor.org/rfc/rfc9110.html#name-delete
+*/
+	std::string		path = client->get_rq_abs_path();
+	path.insert(0, client->assigned_location->root);
+
+	/* CGI Here ? */
+
+	_delete_file(client, path);
+}
+
+void	Webserv::_delete_file(Client *client, const std::string &path)
+{
+	if (access(path.c_str(), F_OK) == -1)
+	{
+		std::perror("err access()");
+		client->status = 404;
+		return ;
+	}
+
+	if (access(path.c_str(), W_OK) == -1)
+	{
+		std::perror("err access()");
+		client->status = 403;
+		return ;
+	}
+
+	if (remove(path.c_str()) == -1)
+	{
+		std::perror("err remove()");
+		client->status = 500;
+		return ;
+	}
+}

srcs/webserv/method_get.cpp

@@ -0,0 +1,191 @@
+
+#include "Webserv.hpp"
+
+void	Webserv::_get(Client *client)
+{
+/* RULES **
+
+if path is a valid dir check if index is specified and serve that
+if no index and autoindex, server that
+if file, server that!
+
+Where does cgi fit in in all this ???
+
+
+*/
+	std::string		path = client->get_rq_abs_path();
+
+	// this might not be the best thing, a voir
+	path.insert(0, client->assigned_location->root);
+
+	std::cerr << "path = " << path << "\n";
+
+	// path = root + location.path
+	// we will tack on an index if there is a valid one
+	// or autoindex if allowed
+	// or let _get_file sort out the error otherwise.
+
+	if (path_is_valid(path) == 1)
+	{
+	//	std::cout << "path is valid\n";
+		if (path[path.size() - 1] != '/')
+			path.push_back('/');
+		for (size_t i = 0; i < client->assigned_location->index.size(); i++)
+		{
+//			std::cout << "location path: " << client->assigned_location->path << '\n';
+//			std::cout << "location index: " << client->assigned_location->index[i] << '\n';
+//			std::cout << "path with index: " << path + assigned_location->index[i] << '\n';
+			if (path_is_valid(path + client->assigned_location->index[i]) == 2)
+			{
+	//			std::cout << "found a valid index\n";
+				path.append(client->assigned_location->index[i]);
+				_get_file(client, path);
+				return ;
+			}
+		}
+		if (client->assigned_location->autoindex == true)
+		{
+			_autoindex(client, path);
+			return ;
+		}
+	}
+//	else
+//		_get_file(client, path);
+	// what about cgi ???
+
+
+
+	// TMP HUGO
+	//
+	std::string	script_output;
+	if (_is_cgi(client))
+	{
+		script_output = _exec_cgi(client);
+		_check_script_output(client, script_output);
+		client->response += script_output;
+		return;
+	}
+	//
+	// END TMP HUGO
+
+	_get_file(client, path);
+}
+
+# define MAX_FILESIZE 1000000 // (1Mo)
+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.
+*/
+	std::ifstream		ifd; // For chunk, ifstream directly in struct CLient for multiples read without close() ?
+	std::stringstream	buf;
+
+	std::cout << "made it to get_file\n";
+
+	if (access(path.c_str(), F_OK) == -1)
+	{
+		std::perror("err access()");
+		client->status = 404;
+		return ;
+	}
+
+	if (access(path.c_str(), R_OK) == -1)
+	{
+		std::perror("err access()");
+		client->status = 403;
+		return ;
+	}
+
+	ifd.open(path.c_str(), std::ios::ate);
+	if (!ifd)
+	{
+		std::cerr << path << ": ifd.open fail" << '\n';
+		client->status = 500;
+	}
+	else
+	{
+		/* std::streampos	size = ifd.tellg();
+		// WIP Low priority : Chunk or not chunk (if filesize too big)
+		if (size > MAX_FILESIZE)
+		{
+			// Then chunk
+			client->status = 500; // WIP temp
+			std::cerr << "File too large for non chunk body\n";
+			return ;
+		} */
+
+		ifd.seekg(0, std::ios::beg);
+		buf << ifd.rdbuf();
+		if (!ifd || !buf)
+		{
+			std::cerr << path << ": ifd.read fail" << '\n';
+			client->status = 500;
+		}
+		else
+		{
+			client->status = 200;
+			std::string	file_ext = _determine_file_extension(path);
+			_append_body(client, buf.str(), file_ext);
+		}
+	}
+}
+
+// i only sort of need &path...
+// def can improve but works for now...
+//void	Webserv::_autoindex(Client *client, LocationConfig &location, std::string &path)
+void	Webserv::_autoindex(Client *client, std::string &path)
+{
+//	std::cout << "made it to _autoindex\n";
+
+	(void)path;
+	std::string		dir_list;
+	DIR				*dir;
+	struct dirent	*ent;
+
+//	std::cout << "location root: " << client->assigned_location->root << " location path: "
+//		<< client->assigned_location->path << '\n';
+
+//	if ((dir = opendir (path.c_str())) != NULL)
+	if ((dir = opendir ((client->assigned_location->root + client->assigned_location->path).c_str())) != NULL)
+	{
+	/* print all the files and directories within directory */
+		dir_list.append(AUTOINDEX_START);
+		dir_list.append(client->assigned_location->path);
+		dir_list.append(AUTOINDEX_MID1);
+		dir_list.append(client->assigned_location->path);
+		dir_list.append(AUTOINDEX_MID2);
+		while ((ent = readdir (dir)) != NULL)
+		{
+			if (strcmp(".", ent->d_name) == 0)
+				continue ;
+			dir_list.append("<a href=\"");
+			dir_list.append(client->assigned_location->path.c_str());
+			dir_list.append(ent->d_name);
+			dir_list.append("\">");
+			dir_list.append(ent->d_name);
+			dir_list.append("</a>");
+			dir_list.append("\r\n");	// is this right?
+		}
+
+// <a href="http://nginx.org/">nginx.org</a>.<br/>
+// <a href="/test/test_deeper/index1.html">index1.html</a>
+
+// apparently this is more than good enough!
+// <a href="/test/test_deeper/..">..</a>
+
+		dir_list.append(AUTOINDEX_END);
+//		std::cout << "\n\n" << dir_list << '\n';
+		closedir (dir);
+		_append_body(client, dir_list, "html");
+	}
+	else
+	{
+		// in theory not possible cuz we already checked...
+	/* could not open directory */
+//	perror ("");
+		std::cout << "could not open dir\n";
+		return ;
+	}
+}

srcs/webserv/method_post.cpp

@@ -0,0 +1,63 @@
+
+#include "Webserv.hpp"
+
+
+void	Webserv::_post(Client *client)
+{
+/*
+	WIP
+	https://www.rfc-editor.org/rfc/rfc9110.html#name-post
+*/
+	std::string		path = client->get_rq_abs_path();
+	path.insert(0, client->assigned_location->root);
+
+	/* CGI Here ? */
+
+	_post_file(client, path);
+}
+
+void	Webserv::_post_file(Client *client, const std::string &path)
+{
+	std::ofstream	ofd;
+
+	bool	file_existed;
+	if (access(path.c_str(), F_OK) == -1)
+		file_existed = false;
+	else
+		file_existed = true;
+
+	// How to determine status 403 for file that dont already exist ?
+	if (file_existed && access(path.c_str(), W_OK) == -1)
+	{
+		std::perror("err access()");
+		client->status = 403;
+		return ;
+	}
+
+	ofd.open(path.c_str(), std::ios::trunc);
+	if (!ofd)
+	{
+		std::cerr << path << ": ofd.open fail" << '\n';
+		client->status = 500;
+	}
+	else
+	{
+		// Content-Length useless at this point ?
+		ofd << client->get_rq_body();
+		if (!ofd)
+		{
+			std::cerr << path << ": ofd.write fail" << '\n';
+			client->status = 500;
+		}
+		else if (file_existed)
+		{
+			client->status = 200;
+			// WIP https://www.rfc-editor.org/rfc/rfc9110.html#name-200-ok
+		}
+		else
+		{
+			client->status = 201;
+			// WIP https://www.rfc-editor.org/rfc/rfc9110.html#section-9.3.3-4
+		}
+	}
+}

srcs/webserv/parsing_message_http.cpp

@@ -0,0 +1,17 @@
+
+#include "parsing_message_http.hpp"
+
+std::string
+	parse_http_body(std::string message)
+{
+	std::string							body;
+	size_t								pos;
+
+	pos = message.find(CRLF CRLF);
+	pos += std::string(CRLF CRLF).size();
+	// TODO: copying just like that might fail in case of binary or images
+	body = message.substr(pos);
+
+	return body;
+}
+

srcs/webserv/parsing_message_http.hpp

@@ -0,0 +1,32 @@
+
+#ifndef PARSING_MESSAGE_HTTP_HPP
+# define PARSING_MESSAGE_HTTP_HPP
+
+# include <iostream>
+# include <string>
+# include <vector>
+# include <map>
+# include "utils.hpp"
+
+std::map<std::string, std::string>
+	parse_http_headers (
+		std::string headers,
+		std::map<std::string, std::string> fields )
+
+std::string
+	parse_http_body(std::string message);
+
+//	http message structure :
+//
+//	start-line
+//		request-line
+//			method SP target SP version
+//		response-line
+//			version SP status SP reason
+//	header-fields
+//		name ":" SP value
+//	CRLF
+//	body
+
+#endif
+

srcs/webserv/parsing_request.cpp

@@ -0,0 +1,2 @@
+#include "parsing_request.hpp"
+

srcs/webserv/request.cpp

@@ -1,44 +1,90 @@
 
 #include "Webserv.hpp"
 
-#define BUFSIZE 8192
+// Arbitrary values
+#define BUFSIZE 8192 // (8Ko)
+#define MAX_HEADER_SIZE 16384 // (16Ko)
+
+enum read_return
+{
+	READ_IN_PROGRESS,
+	READ_COMPLETE,
+	READ_CLOSE,
+};
 
 void	Webserv::_request(Client *client)
 {
-	_read_request(client);
+	int ret = _read_request(client);
+
 	if (g_last_signal)
-		_handle_last_signal();	
+		_handle_last_signal();
+
+	if (ret == READ_CLOSE)
+	{
+		_close_client(client->get_cl_fd());
+	}
+	else if (ret == READ_COMPLETE)
+	{
+		_epoll_update(client->get_cl_fd(), EPOLLOUT, EPOLL_CTL_MOD);
+	}
 }
 
-void	Webserv::_read_request(Client *client)
+int	Webserv::_read_request(Client *client) // Messy, Need refactoring
 {
-	char	buf[BUFSIZE+1];
+	char	buf[BUFSIZE];
 	ssize_t	ret;
 
+	std::cerr << "call recv()" << "\n" ;
 	ret = ::recv(client->get_cl_fd(), buf, BUFSIZE, 0);
 	std::cerr << "recv() on fd(" << client->get_cl_fd() << ") returned = " << ret << "\n" ;
 	if (ret == -1)
 	{
 		std::perror("err recv()");
 		std::cerr << "client ptr =" << client << "\n"; // DEBUG
-		std::cerr << "client.get_cl_fd() =" << client->get_cl_fd() << "\n"; // DEBUG
-		_close_client(client->get_cl_fd());
-		return ;
+		std::cerr << "client.fd =" << client->get_cl_fd() << "\n"; // DEBUG
+		return READ_CLOSE;
 	}
-	if (ret == 0) // Not sure what to do in case of 0. Just close ?
+	if (ret == 0)
 	{
-		_close_client(client->get_cl_fd());
-		return ;
+		std::cerr << "recv() read 0, then close client" << "\n"; // DEBUG
+		return READ_CLOSE;
 	}
-	/*
-	if (ret == BUFSIZE)
-		// send error like "request too long" to client
-	*/
 
-	buf[ret] = '\0';
-	client->raw_request.append(buf);
-	client->parse_request();
+	client->raw_request.append(buf, ret);
+	if (!client->header_complete)
+	{
+		if (client->raw_request.find(CRLF CRLF) != std::string::npos)
+		{
+			client->header_complete = true;
+			client->parse_request_headers(_servers);
+			std::cerr << client->get_rq_method_str() << " " << client->get_rq_uri() << " " << client->get_rq_version() << "\n"; // DEBUG
+			if (client->status)
+				return READ_COMPLETE;
 
-	_epoll_update(client->get_cl_fd(), EPOLLOUT, EPOLL_CTL_MOD);
+			if (client->get_rq_headers("Content-Type").empty() && client->get_rq_headers("Content-Length").empty()) // No body case
+				return READ_COMPLETE;
+		}
+		else if (client->raw_request.size() > MAX_HEADER_SIZE)
+		{
+			// 413 or 400 ? 413 seems common among http server, but don't fit perfectly.
+			client->status = 413;
+			return READ_COMPLETE;
+		}
+	}
+	else if (client->header_complete)
+	{
+		client->read_body_size += ret;
+		if (client->read_body_size > client->assigned_server->client_body_limit)
+		{
+			client->status = 413;
+			return READ_COMPLETE;
+		}
+		if ((int)client->read_body_size >= ::atoi(client->get_rq_headers("Content-Length").c_str()))
+		{
+			client->parse_request_body();
+			return READ_COMPLETE;
+		}
+	}
+
+	return READ_IN_PROGRESS;
 }
-

srcs/webserv/response.cpp

@@ -1,57 +1,60 @@
 
 #include "Webserv.hpp"
 
+enum send_return
+{
+	SEND_IN_PROGRESS, // unused
+	SEND_COMPLETE,
+	SEND_CLOSE,
+};
+
 void	Webserv::_response(Client *client)
 {
-	client->status = 200; // default value
+	int ret = _send_response(client);
 
-	ServerConfig	&server = _determine_process_server(client);
-	_send_response(client, server);
 	if (g_last_signal)
 		_handle_last_signal();
+
+	if (ret == SEND_CLOSE)
+	{
+		_close_client(client->get_cl_fd());
+	}
+	else if (ret == SEND_COMPLETE)
+	{
+		if (client->get_rq_headers("Connection") == "close")
+			_close_client(client->get_cl_fd());
+		else
+		{
+			_epoll_update(client->get_cl_fd(), EPOLLIN, EPOLL_CTL_MOD);
+			client->clear();
+		}
+	}
 }
 
-void	Webserv::_send_response(Client *client, ServerConfig &server)
+int	Webserv::_send_response(Client *client)
 {
 	ssize_t	ret;
 
 	std::cerr << "send()\n";
 
 	_append_base_headers(client);
-	_construct_response(client, server);
+	if (!client->status)
+		_construct_response(client);
 	_insert_status_line(client);
 	if (client->status >= 400)
-		_error_html_response(client, server);
+		_error_html_response(client);
 
+	std::cerr << "client->response.size() = " << client->response.size() << "\n"; // DEBUG
 	ret = ::send(client->get_cl_fd(), client->response.c_str(), client->response.size(), 0);
 	if (ret == -1)
 	{
 		std::perror("err send()");
-		std::cerr << "client.get_cl_fd() =" << client->get_cl_fd() << "\n"; // DEBUG
-		_close_client(client->get_cl_fd());
-		return ;
+		std::cerr << "client.fd =" << client->get_cl_fd() << "\n"; // DEBUG
+		return SEND_CLOSE;
 	}
+	std::cerr << "ret send() = " << ret << "\n"; // DEBUG
 
-	// Body send (WIP for working binary files)
-	if (client->body_size)
-	{
-		ret = ::send(client->get_cl_fd(), client->buf, client->body_size, 0);
-		if (ret == -1)
-		{
-			std::perror("err send()");
-			std::cerr << "client.get_cl_fd() =" << client->get_cl_fd() << "\n"; // DEBUG
-			_close_client(client->get_cl_fd());
-			return ;
-		}
-	}
-
-	if (client->get_rq_headers("Connection") == "close")
-		_close_client(client->get_cl_fd());
-	else
-	{
-		_epoll_update(client->get_cl_fd(), EPOLLIN, EPOLL_CTL_MOD);
-		client->clear();
-	}
+	return SEND_COMPLETE;
 }
 
 void	Webserv::_append_base_headers(Client *client)
@@ -64,48 +67,28 @@ void	Webserv::_append_base_headers(Client *client)
 		client->response.append("Connection: keep-alive" CRLF);
 }
 
-void	Webserv::_construct_response(Client *client, ServerConfig &server)
+void	Webserv::_construct_response(Client *client)
 {
-	// TODO : Move this in read(), stop read if content too large
-	if (client->get_rq_body().size() > server.client_body_limit)
-	{
-		client->status = 413;
-		return;
-	}
-	LocationConfig	&location = _determine_location(server, client->get_rq_uri());
-	_process_method(client, server, location);
+	/* Switch between normal behavior or CGI here ?
+	maybe better than in _get(), _post(), ...*/
+	_process_method(client);
 }
 
-void	Webserv::_process_method(Client *client, ServerConfig &server, LocationConfig &location)
+void	Webserv::_process_method(Client *client)
 {
-	unsigned int	allow_methods = ANY_METHODS; // TEMP VARIABLE
-	// after update in ConfigParser, use the "allow_methods" of location.
-	// TODO in ConfigParser : by default if no field in config file, "allow_methods" must be set to ANY_METHODS
+	std::cerr << "assigned_location->path = " << client->assigned_location->path << "\n"; // debug
+	std::cerr << "allow_methods = " << client->assigned_location->allow_methods << "\n"; // debug
 
-	if (client->get_rq_method() == UNKNOWN)
+	switch (client->get_rq_method())
 	{
-		client->status = 501;
-	}
-	else if (allow_methods & client->get_rq_method())
-	{
-		switch (client->get_rq_method())
-		{
-			case (GET):
-				_get(client, server, location); break;
-			case (POST):
-				_post(client, server, location); break;
-			case (DELETE):
-				_delete(client, server, location); break;
-			default:
-				break;
-		}
-	}
-	else
-	{
-		client->status = 405;
-		client->response.append("Allow: ");
-		client->response.append(::http_methods_to_str(allow_methods));
-		client->response.append(CRLF);
+		case (GET):
+			_get(client); break;
+		case (POST):
+			_post(client); break;
+		case (DELETE):
+			_delete(client); break;
+		default:
+			break;
 	}
 }
 
@@ -119,248 +102,44 @@ void	Webserv::_insert_status_line(Client *client)
 	client->response.insert(0, status_line);
 }
 
-void	Webserv::_error_html_response(Client *client, ServerConfig &server)
+void	Webserv::_error_html_response(Client *client)
 {
-	if (server.error_pages[client->status].empty())
+	if (!client->assigned_server || client->assigned_server->error_pages[client->status].empty())
 	{
 		std::string	html_page = HTML_ERROR;
 		::replace_all_substr(html_page, STATUS_PLACEHOLDER, _http_status[client->status]);
-		_append_body(client, html_page.c_str(), html_page.size(), "html");
+		_append_body(client, html_page, "html");
 	}
 	else
-		_get_file(client, server.error_pages[client->status]);
+		_get_file(client, client->assigned_server->error_pages[client->status]);
 }
 
-#define INDEX "index.html" // temp wip
-void	Webserv::_get(Client *client, ServerConfig &server, LocationConfig &location)
+void	Webserv::_append_body(Client *client, const std::string &body, const std::string &file_extension)
 {
-	(void)server; // To remove from arg if we determine its useless	
-	std::string		path = client->get_rq_uri();
+	const std::string	&mime_type = _mime_types[file_extension];
 
-	if (path == "/") // TODO : index and autoindex
-		path.append(INDEX);
-	path.insert(0, location.root);
-
-	std::cerr << "path = " << path << "\n";
-
-	// TMP HUGO
-	//
-	if (_is_cgi(client))
-	{
-		_exec_cgi(client);
-		_response_correction(client);
-		return;
-	}
-	//
-	// END TMP HUGO
-
-	_get_file(client, path);
-}
-
-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() ?
-	// char			buf[MAX_FILESIZE+1];
-
-	if (access(path.c_str(), F_OK) == -1)
-	{
-		std::perror("err access()");
-		client->status = 404;
-		return ;
-	}
-
-	if (access(path.c_str(), R_OK) == -1)
-	{
-		std::perror("err access()");
-		client->status = 403;
-		return ;
-	}
-
-	ifd.open(path.c_str(), std::ios::binary | std::ios::ate); // std::ios::binary (binary for files like images ?)
-	if (!ifd)
-	{
-		std::cerr << path << ": ifd.open fail" << '\n';
-		client->status = 500;
-	}
+	client->response.append("Content-Type: ");
+	if (mime_type.empty())
+		client->response.append(MIME_TYPE_DEFAULT);
 	else
 	{
-		std::streampos	size = ifd.tellg();
-
-		// WIP : Chunk or not chunk (if filesize too big)
-		if (size > MAX_FILESIZE)
-		{
-			// Then chunk
-			client->status = 500; // WIP temp
-			std::cerr << "File too large for non chunk body\n";
-			ifd.close();
-			return ;
-		}
-
-		ifd.seekg(0, std::ios::beg);
-		ifd.read(client->buf, size);
-		if (!ifd)
-		{
-			std::cerr << path << ": ifd.read fail" << '\n';
-			client->status = 500;
-		}
-		else
-		{
-			client->status = 200;
-			client->buf[ifd.gcount()] = '\0';
-
-			std::string	file_ext = "";
-			size_t	dot_pos = path.rfind(".");
-			std::cerr << "dot_pos = " << dot_pos << "\n";
-			if (dot_pos != std::string::npos && dot_pos + 1 < path.size())
-				file_ext = path.substr(dot_pos + 1);
-			std::cerr << "file_ext = " << file_ext << "\n";
-
-			client->body_size = ifd.gcount();
-			// WIP, pass empty body argument because append to string mess up binary file
-			_append_body(client, "", client->body_size, file_ext);
-		}
-
-		ifd.close();
-	}
-}
-
-// WIP HUGO
-void	Webserv::_response_correction(Client *client)
-{
-	(void)client;
-}
-
-void	Webserv::_append_body(Client *client, const char *body, size_t body_size, const std::string &file_extension)
-{
-/*
-	TODO : determine Content-Type
-	how ? read the body ?
-	or before in other way (like based and file extension) and pass here as argument ?
-	http://nginx.org/en/docs/http/ngx_http_core_module.html#types
-	Need to look "conf/mime.types" of nginx. Maybe make a map<> based on that.
-*/
-		const std::string	&mime_type = _mime_types[file_extension];
-		client->response.append("Content-Type: ");
 		client->response.append(mime_type);
 		if (mime_type.find("text/") != std::string::npos)
 			client->response.append("; charset=UTF-8");
-		client->response.append(CRLF);
+	}
+	client->response.append(CRLF);
 
-		client->response.append("Content-Length: ");
-		std::string	tmp = ::itos(body_size);
-		client->response.append(tmp);
-		client->response.append(CRLF);
+	client->response.append("Content-Length: ");
+	std::string	tmp = ::itos(body.size());
+	client->response.append(tmp);
+	client->response.append(CRLF);
 
-		client->response.append(CRLF);
-		client->response.append(body);
+	client->response.append(CRLF);
+	client->response.append(body);
 }
 
-void	Webserv::_post(Client *client, ServerConfig &server, LocationConfig &location)
-{
-	(void)server; // To remove from arg if we determine its useless	
-/*
-	WIP
-	https://www.rfc-editor.org/rfc/rfc9110.html#name-post
-*/
-	std::string		path = client->get_rq_uri();
-	path.insert(0, location.root);
-
-	/* CGI Here ? */
-
-	_post_file(client, path);
-}
-
-void	Webserv::_post_file(Client *client, const std::string &path)
-{
-	std::ofstream	ofd;
-
-	bool	file_existed;
-	if (access(path.c_str(), F_OK) == -1)
-		file_existed = false;
-	else
-		file_existed = true;
-
-	// How to determine status 403 for file that dont already exist ?
-	if (file_existed && access(path.c_str(), W_OK) == -1)
-	{
-		std::perror("err access()");
-		client->status = 403;
-		return ;
-	}
-
-	ofd.open(path.c_str(), std::ios::binary | std::ios::trunc);
-	if (!ofd)
-	{
-		std::cerr << path << ": ofd.open fail" << '\n';
-		client->status = 500;
-	}
-	else
-	{
-		// Used body.size() so Content-Length useless at this point ?
-		// Maybe usefull in _read_request() for rejecting too big content.
-		// Need to _determine_process_server() as soon as possible,
-		// like in _read_request() for stopping read if body is too big ?
-		ofd.write(client->get_rq_body().c_str(), client->get_rq_body().size());
-		if (!ofd)
-		{
-			std::cerr << path << ": ofd.write fail" << '\n';
-			client->status = 500;
-		}
-		else if (file_existed)
-		{
-			client->status = 200;
-			// WIP https://www.rfc-editor.org/rfc/rfc9110.html#name-200-ok
-		}
-		else
-		{
-			client->status = 201;
-			// WIP https://www.rfc-editor.org/rfc/rfc9110.html#section-9.3.3-4
-		}
-
-		ofd.close();
-	}
-}
-
-void	Webserv::_delete(Client *client, ServerConfig &server, LocationConfig &location)
-{
-	(void)server; // To remove from arg if we determine its useless
-/*
-	WIP
-	https://www.rfc-editor.org/rfc/rfc9110.html#name-delete
-*/
-	std::string		path = client->get_rq_uri();
-	path.insert(0, location.root);
-
-	/* CGI Here ? */
-
-	_delete_file(client, path);
-}
-
-void	Webserv::_delete_file(Client *client, const std::string &path)
-{
-	if (access(path.c_str(), F_OK) == -1)
-	{
-		std::perror("err access()");
-		client->status = 404;
-		return ;
-	}
-
-	if (access(path.c_str(), W_OK) == -1)
-	{
-		std::perror("err access()");
-		client->status = 403;
-		return ;
-	}
-
-	if (remove(path.c_str()) == -1)
-	{
-		std::perror("err remove()");
-		client->status = 500;
-		return ;
-	}
-}
-
-ServerConfig	&Webserv::_determine_process_server(Client *client)
+// Temporary Global Scope. Probably move to Client in the future.
+ServerConfig	*_determine_process_server(Client *client, std::vector<ServerConfig> &servers)
 {
 /*
 	http://nginx.org/en/docs/http/request_processing.html
@@ -369,45 +148,246 @@ ServerConfig	&Webserv::_determine_process_server(Client *client)
 */
 
 	std::string	const &server_name = client->get_rq_headers("Host");
-	std::vector<ServerConfig>::iterator	it = _servers.begin();
-	std::vector<ServerConfig>::iterator	default_server = _servers.end();
+	std::vector<ServerConfig>::iterator	it = servers.begin();
+	std::vector<ServerConfig>::iterator	default_server = servers.end();
 
-	while (it != _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())
+			else if (default_server == servers.end())
 				default_server = it;
 		}
 		++it;
 	}
-	if (it != _servers.end())
-		return (*it);
+	if (it != servers.end())
+		return (&(*it));
 	else
-		return (*default_server);
+		return (&(*default_server));
 }
 
-LocationConfig	&Webserv::_determine_location(ServerConfig &server, std::string const &path)
-{
-/*
-	Assume there is at least one location in vector<LocationConfig> for path "/"
-	TODO in ConfigParser :
-	If no location block in config file, one need to be generated
-	for path "/", and filled with fields "root" and "index" based on parent server block
-*/
 
-	std::vector<LocationConfig>::iterator	it = server.locations.begin();
+const LocationConfig	*_determine_location_COOP_FIX(const ServerConfig &server, const std::string &path)
+{
+/* Pseudo-code :
+- comparer les size(), si location.path > client.path, stop comparaison.
+- client.path.compare(0, location.path.size(), location.path)
+if ( == 0)
+{
+	if (location.path.size() == client.path.size())
+	{
+		FOUND;
+	}
+	else if (client.path[location.path.size()] == '/')
+	{
+		FOUND;
+	}
+}
+else
+{
+	NOT FOUND;
+	++next;
+}
+*/
+	std::vector<LocationConfig>::const_iterator	it = server.locations.begin();
 	while (it != server.locations.end())
 	{
-		if (it->path.compare(0, path.size(), path))
-			break;
+		if (it->path.size() > path.size())
+		{
+			// prendre en compte l'éventuel "/" final si location est un dossier
+			if (it->path.size()-1 > path.size() || it->path[it->path.size()-1] != '/')
+			{
+				++it;
+				continue;
+			}
+		}
+		if (path.compare(0, it->path.size(), it->path) == 0)
+		{
+			if (it->path.size() == path.size())
+				break;
+			else if (path[it->path.size()-1] == '/')
+				break;
+		}
 		++it;
 	}
 	if (it != server.locations.end())
-		return (*it);
+		return (&(*it));
 	else
-		return (server.locations.front());
+		return (&(server.locations.back()));
+}
+
+// Temporary Global Scope. Probably move to Client in the future.
+const LocationConfig	*_determine_location(const ServerConfig &server, const std::string &path)
+{
+	std::cout << "determin location path sent: " << path << '\n';
+
+
+/// NO FUCKING IDEA WHY BUT...
+// basically if 2 strings are identical to a point, compare from 
+// longer one or it'll freak out cuz of \0 or something idk
+//// Basically: str.compare() from the larger string...
+
+/* 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
+
+*/
+	for (std::vector<LocationConfig>::const_iterator it = server.locations.begin(); it != server.locations.end(); it++)
+	{
+		std::cout << it->path << "  --  ";
+//		std::cout << it->path[it->path.size() - 1] << " ";
+// it->path.size() -1 only when path ends in / because
+// if path doesn't end in / then we are looking for a file
+// meaning all it->paths that end in / are wrong if they >=
+	//	if (it->path[it->path.size() - 1] == '/' ? it->path.size() - 1 > path.size() : it->path.size() > path.size())
+		if (path[path.size() - 1] == '/' ? it->path.size() > path.size() : it->path.size() - 1 > path.size())
+		{
+			std::cout << "skipping this one\n";
+			continue ;
+		}
+
+//		if (it->path.size() > path.size()) // Warning : il faut aussi prendre en compte l'éventuel "/" final
+//			continue;
+
+
+// IS THERE A WAY TO SIMPLIFY THIS LOGIC ???
+
+
+//		if (it->path[it->path.size() - 1] == '/')
+		if (path[path.size() - 1] == '/')
+		{
+			if (path.compare(0, it->path.size(), it->path) == 0)
+			{
+				std::cout << "checking with last /\n";
+				if (it->path.size() == path.size())
+				{
+					std::cout << "path sizes are equal \n";
+					return (&(*it));
+				}
+				else if (path[it->path.size() - 1] == '/')
+				{
+					std::cout << "ends in /\n";
+					return (&(*it));
+				}
+			}
+		}
+		else
+		{
+			if (path.size() <= it->path.size())
+			{
+				std::cout << "path is missing a /\n";
+				if (it->path.compare(0, path.size(), path) == 0)
+					return (&(*it));
+				// means we are looking for /test/test_deeper/
+				// with /test/test_deeper
+			}
+			else
+			{
+//				if (it->path.compare(0, it->path.size() - 1, path) == 0)
+				if (path.compare(0, it->path.size(), it->path) == 0)
+				{
+					std::cout << "checking without last /\n";
+					if (it->path.size() - 1 == path.size())
+						return (&(*it));
+					else if (path[it->path.size() - 1] == '/')
+						return (&(*it));
+				}
+			}
+		}
+//	/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
+
+
+*/
+
+
+
+
+	}
+//	if (it != server.locations.end())
+//		return (&(*it));
+//	else
+		return (&(server.locations.back()));
+
+
+/*
+
+	std::vector<LocationConfig>::const_iterator best;
+	std::cout << "\nMade it to weird location picker case.\n";
+
+	for (std::vector<LocationConfig>::const_iterator it = server.locations.begin(); it != server.locations.end(); it++)
+	{
+*/
+		//	if (rit->path.size() > path.size())
+/*			if ((rit->path[rit->path.size() - 1] == '/' ? rit->path.size() : rit->path.size() - 1) > path.size())
+			{
+				std::cout << "skipping this one\n";
+				continue ;
+			}
+*/
+// OK I REALLY DON"T LOVE THIS PART, BUT IT DOES WORK FOR NOW...
+	//	if (it->path[it->path.size() - 1] == '/'
+	//		&& it->path.compare(0, it->path.size(), path + "/") == 0)
+// HOLD ON THIS MIGHT BE GOOD, BUT I COULD USE SOME HELP...
+
+//test /test/ 
+
+//test/redirect/index /test/redirect
+//test/redirec_something
+
+// thing is reverse sorted
+// if location path is longer than path sent, don't look at it
+// otherwise compare if (path.compare(0, it->path.size(), it->path) == 0)
+// do another size compare and look for / if location smaller than client otherwise /redirect_something /redirect
+
+// compare everything 
+// 
+
+/*		if (it->path[it->path.size() - 1] == '/'
+			&& it->path.compare(0, it->path.size() - 1, path) == 0)
+		{
+			best = it;
+			std::cout << "Picked a best! 1\n";
+		}
+			// int	comp = path.compare(0, rit->path.size(), rit->path);
+			//int	comp = rit->path.compare(0, rit->path.size() - 1, path);
+			// std::cout << "rit path size: " << rit->path.size() << " comp: " << comp << '\n';
+	//		if (rit->path.compare(0, rit->path.size(), path) == 0)
+			// if (comp == 0)
+
+		if (path.compare(0, it->path.size(), it->path) == 0)
+		{
+			best = it;
+			std::cout << "Picked a best! 2\n";
+		}
+	}
+//	//test.comare(0, 5, /test/something)
+	// /test /test/something 
+	return (&(*best));
+*/
+}
+
+std::string	Webserv::_determine_file_extension(const std::string &path) const
+{
+	size_t	dot_pos = path.rfind(".");
+	if (dot_pos != std::string::npos && dot_pos + 1 < path.size())
+		return ( path.substr(dot_pos + 1) );
+	return (std::string(""));
 }

srcs/webserv/run_loop.cpp

@@ -34,17 +34,34 @@ void	Webserv::run()
 		i = 0;
 		while (i < nfds)
 		{
-			// TODO : handle EPOLLERR and EPOLLHUP
-			it_socket = std::find(_listen_sockets.begin(), _listen_sockets.end(), events[i].data.fd);
-			if (it_socket != _listen_sockets.end() && events[i].events & EPOLLIN)
-				_accept_connection(*it_socket);
-			else if (events[i].events & EPOLLIN)
-				_request( &(*std::find(_clients.begin(), _clients.end(), events[i].data.fd)) );
-			else if (events[i].events & EPOLLOUT)
-				_response( &(*std::find(_clients.begin(), _clients.end(), events[i].data.fd)) );
-			++i;
-			if (!g_run)
+			try
+			{
+				// TODO : handle EPOLLERR and EPOLLHUP
+				it_socket = std::find(_listen_sockets.begin(), _listen_sockets.end(), events[i].data.fd);
+				if (it_socket != _listen_sockets.end() && events[i].events & EPOLLIN)
+					_accept_connection(*it_socket);
+				else if (events[i].events & EPOLLIN)
+					_request( &(*std::find(_clients.begin(), _clients.end(), events[i].data.fd)) );
+				else if (events[i].events & EPOLLOUT)
+					_response( &(*std::find(_clients.begin(), _clients.end(), events[i].data.fd)) );
+				++i;
+				if (!g_run)
+					break;
+			}
+			catch (const std::bad_alloc& e)
+			{
+				std::cerr << e.what() << '\n';
+				_close_all_clients();
+				/* Swap to free the memory
+				From : http://www.uml.org.cn/c%2B%2B/pdf/EffectiveSTL.pdf#page=66 */
+				std::vector<Client>().swap(_clients);
 				break;
+			}
+			catch (const std::exception& e)
+			{
+				std::cerr << e.what() << '\n';
+				++i;
+			}
 		}
 	}
 }

stylesheet/style.css

@@ -0,0 +1,4 @@
+h1 {
+  color: red;
+  text-align: center;
+}

www/error_pages/error_404.html

@@ -0,0 +1,11 @@
+<!DOCTYPE html>
+<html>
+<head>
+	<title>404 Not Found</title>
+</head>
+<body>
+	<h1 style="text-align:center">Check it UP 404 Not Found</h1>
+	<hr>
+	<p style=\"text-align:center\">Le Webserv/0.1</p>
+</body>
+</html>

www/form_get.html

@@ -0,0 +1,13 @@
+<!DOCTYPE html>
+<html>
+	<head>
+		<meta charset="UTF-8">
+		<title></title>
+	</head>
+	<body>
+		<form method="get">
+			<input type="submit" value="submit">
+		</form>
+	</body>
+</html>
+

www/form_post.html

@@ -0,0 +1,13 @@
+<!DOCTYPE html>
+<html>
+	<head>
+		<meta charset="UTF-8">
+		<title></title>
+	</head>
+	<body>
+		<form method="post">
+			<input type="submit" value="submit">
+		</form>
+	</body>
+</html>
+

www/index.html

@@ -8,4 +8,4 @@
 	<hr>
 	<p style="text-align:center">(˚3˚)</p>
 </body>
-</html>
+</html>

www/test/index1.html

@@ -0,0 +1,14 @@
+<!DOCTYPE html>
+<html>
+<head>
+	<title>Webserv test index</title>
+<!--	<link rel="stylesheet" href="stylesheet/style.css"> -->
+	<link rel="stylesheet" href="/stylesheet/style.css">
+	
+</head>
+<body>
+	<h1>Webserv Test Index</h1>
+	<hr>
+	<p style="text-align:center">(˚3˚)</p>
+</body>
+</html>

www/test/something.html

@@ -0,0 +1,11 @@
+<!DOCTYPE html>
+<html>
+<head>
+	<title>Webserv test Something</title>
+</head>
+<body>
+	<h1 style="text-align:center">Webserv Test Something</h1>
+	<hr>
+	<p style="text-align:center">(˚3˚)</p>
+</body>
+</html>

www/test/test_deeper/index1.html

@@ -0,0 +1,11 @@
+<!DOCTYPE html>
+<html>
+<head>
+	<title>Webserv Test Deeper Index</title>
+</head>
+<body>
+	<h1 style="text-align:center">Webserv Test Deeper Index</h1>
+	<hr>
+	<p style="text-align:center">(˚3˚)</p>
+</body>
+</html>

www/test/test_deeper/something.html

@@ -0,0 +1,11 @@
+<!DOCTYPE html>
+<html>
+<head>
+	<title>Webserv test deeper Something</title>
+</head>
+<body>
+	<h1 style="text-align:center">Webserv Test Deeper Something</h1>
+	<hr>
+	<p style="text-align:center">(˚3˚)</p>
+</body>
+</html>

www/test/test_deeper/super_deep/something.html

@@ -0,0 +1,11 @@
+<!DOCTYPE html>
+<html>
+<head>
+	<title>Webserv test deeper Something</title>
+</head>
+<body>
+	<h1 style="text-align:center">Webserv Test Super Deep Something</h1>
+	<hr>
+	<p style="text-align:center">(˚3˚)</p>
+</body>
+</html>