op public repos

Blame

Date:: Fri Dec 25 12:13:12 2020 UTC
Message:: implement a valid RFC3986 (URI) parser Up until now I used a "poor man" approach: the uri parser is barely a parser, it tries to extract the path from the request, with some minor checking, and that's all. This obviously is not RFC3986-compliant. The new RFC3986 (URI) parser should be fully compliant. It may accept some invalid URI, but shouldn't reject or mis-parse valid URI. (in particular, the rule for the path is way more relaxed in this parser than it is in the RFC text). A difference with RFC3986 is that we don't even try to parse the (optional) userinfo part of a URI: following the Gemini spec we treat it as an error. A further caveats is that %2F in the path part of the URI is indistinguishable from a literal '/': this is NOT conforming, but due to the scope and use of gmid, I don't see how treat a %2F sequence in the path (reject the URI?).
Actions:: History | Blob | Raw File

  1 d5aba4c7 2020-12-24 op /*
  2 d5aba4c7 2020-12-24 op  * Copyright (c) 2020 Omar Polo <op@omarpolo.com>
  3 d5aba4c7 2020-12-24 op  *
  4 d5aba4c7 2020-12-24 op  * Permission to use, copy, modify, and distribute this software for any
  5 d5aba4c7 2020-12-24 op  * purpose with or without fee is hereby granted, provided that the above
  6 d5aba4c7 2020-12-24 op  * copyright notice and this permission notice appear in all copies.
  7 d5aba4c7 2020-12-24 op  *
  8 d5aba4c7 2020-12-24 op  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
  9 d5aba4c7 2020-12-24 op  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
 10 d5aba4c7 2020-12-24 op  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
 11 d5aba4c7 2020-12-24 op  * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
 12 d5aba4c7 2020-12-24 op  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
 13 d5aba4c7 2020-12-24 op  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
 14 d5aba4c7 2020-12-24 op  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
 15 d5aba4c7 2020-12-24 op  */
 16 d5aba4c7 2020-12-24 op 
 17 488f059a 2020-12-24 op #ifndef GMID_H
 18 488f059a 2020-12-24 op #define GMID_H
 19 488f059a 2020-12-24 op 
 20 488f059a 2020-12-24 op #include <arpa/inet.h>
 21 488f059a 2020-12-24 op #include <netinet/in.h>
 22 488f059a 2020-12-24 op 
 23 488f059a 2020-12-24 op #include <poll.h>
 24 488f059a 2020-12-24 op #include <stdio.h>
 25 488f059a 2020-12-24 op #include <stdlib.h>
 26 488f059a 2020-12-24 op #include <syslog.h>
 27 488f059a 2020-12-24 op #include <tls.h>
 28 488f059a 2020-12-24 op #include <unistd.h>
 29 488f059a 2020-12-24 op 
 30 488f059a 2020-12-24 op #ifndef __OpenBSD__
 31 488f059a 2020-12-24 op # define pledge(a, b) 0
 32 488f059a 2020-12-24 op # define unveil(a, b) 0
 33 488f059a 2020-12-24 op #endif
 34 488f059a 2020-12-24 op 
 35 488f059a 2020-12-24 op #ifndef INFTIM
 36 488f059a 2020-12-24 op # define INFTIM -1
 37 488f059a 2020-12-24 op #endif
 38 488f059a 2020-12-24 op 
 39 488f059a 2020-12-24 op #define GEMINI_URL_LEN (1024+3)	/* URL max len + \r\n + \0 */
 40 488f059a 2020-12-24 op 
 41 488f059a 2020-12-24 op /* large enough to hold a copy of a gemini URL and still have extra room */
 42 488f059a 2020-12-24 op #define PATHBUF		2048
 43 488f059a 2020-12-24 op 
 44 488f059a 2020-12-24 op #define SUCCESS		20
 45 488f059a 2020-12-24 op #define TEMP_FAILURE	40
 46 488f059a 2020-12-24 op #define NOT_FOUND	51
 47 488f059a 2020-12-24 op #define BAD_REQUEST	59
 48 488f059a 2020-12-24 op 
 49 488f059a 2020-12-24 op #ifndef MAX_USERS
 50 488f059a 2020-12-24 op #define MAX_USERS	64
 51 488f059a 2020-12-24 op #endif
 52 488f059a 2020-12-24 op 
 53 488f059a 2020-12-24 op #define SAFE_SETENV(var, val) do {		\
 54 488f059a 2020-12-24 op 		const char *_tmp = (val);	\
 55 488f059a 2020-12-24 op 		if (_tmp == NULL)		\
 56 488f059a 2020-12-24 op 			_tmp = "";		\
 57 488f059a 2020-12-24 op 		setenv((var), _tmp, 1);		\
 58 488f059a 2020-12-24 op 	} while(0)
 59 488f059a 2020-12-24 op 
 60 488f059a 2020-12-24 op #define LOG(priority, c, fmt, ...)					\
 61 488f059a 2020-12-24 op 	do {								\
 62 488f059a 2020-12-24 op 		char buf[INET_ADDRSTRLEN];				\
 63 488f059a 2020-12-24 op 		if (inet_ntop((c)->af, &(c)->addr,			\
 64 488f059a 2020-12-24 op 		    buf, sizeof(buf)) == NULL)				\
 65 488f059a 2020-12-24 op 			FATAL("inet_ntop: %s", strerror(errno));	\
 66 488f059a 2020-12-24 op 		if (foreground)						\
 67 488f059a 2020-12-24 op 			fprintf(stderr,					\
 68 488f059a 2020-12-24 op 			    "%s " fmt "\n", buf, __VA_ARGS__);		\
 69 488f059a 2020-12-24 op 		else							\
 70 488f059a 2020-12-24 op 			syslog((priority) | LOG_DAEMON,			\
 71 488f059a 2020-12-24 op 			    "%s " fmt, buf, __VA_ARGS__);		\
 72 488f059a 2020-12-24 op 	} while (0)
 73 488f059a 2020-12-24 op 
 74 488f059a 2020-12-24 op #define LOGE(c, fmt, ...) LOG(LOG_ERR,    c, fmt, __VA_ARGS__)
 75 488f059a 2020-12-24 op #define LOGN(c, fmt, ...) LOG(LOG_NOTICE, c, fmt, __VA_ARGS__)
 76 488f059a 2020-12-24 op #define LOGI(c, fmt, ...) LOG(LOG_INFO,   c, fmt, __VA_ARGS__)
 77 488f059a 2020-12-24 op #define LOGD(c, fmt, ...) LOG(LOG_DEBUG,  c, fmt, __VA_ARGS__)
 78 488f059a 2020-12-24 op 
 79 488f059a 2020-12-24 op #define FATAL(fmt, ...)							\
 80 488f059a 2020-12-24 op 	do {								\
 81 488f059a 2020-12-24 op 		if (foreground)						\
 82 488f059a 2020-12-24 op 			fprintf(stderr, fmt "\n", __VA_ARGS__);		\
 83 488f059a 2020-12-24 op 		else							\
 84 488f059a 2020-12-24 op 			syslog(LOG_DAEMON | LOG_CRIT,			\
 85 488f059a 2020-12-24 op 			    fmt, __VA_ARGS__);				\
 86 488f059a 2020-12-24 op 		exit(1);						\
 87 488f059a 2020-12-24 op 	} while (0)
 88 488f059a 2020-12-24 op 
 89 488f059a 2020-12-24 op enum {
 90 488f059a 2020-12-24 op 	S_OPEN,
 91 488f059a 2020-12-24 op 	S_INITIALIZING,
 92 488f059a 2020-12-24 op 	S_SENDING,
 93 488f059a 2020-12-24 op 	S_CLOSING,
 94 488f059a 2020-12-24 op };
 95 488f059a 2020-12-24 op 
 96 488f059a 2020-12-24 op struct client {
 97 488f059a 2020-12-24 op 	struct tls	*ctx;
 98 488f059a 2020-12-24 op 	int		 state;
 99 488f059a 2020-12-24 op 	int		 code;
100 488f059a 2020-12-24 op 	const char	*meta;
101 488f059a 2020-12-24 op 	int		 fd, waiting_on_child;
102 488f059a 2020-12-24 op 	pid_t		 child;
103 488f059a 2020-12-24 op 	char		 sbuf[1024];	  /* static buffer */
104 488f059a 2020-12-24 op 	void		*buf, *i;	  /* mmap buffer */
105 488f059a 2020-12-24 op 	ssize_t		 len, off;	  /* mmap/static buffer  */
106 488f059a 2020-12-24 op 	int		 af;
107 488f059a 2020-12-24 op 	struct in_addr	 addr;
108 488f059a 2020-12-24 op };
109 488f059a 2020-12-24 op 
110 33d32d1f 2020-12-25 op 
111 33d32d1f 2020-12-25 op struct uri {
112 33d32d1f 2020-12-25 op 	char		*schema;
113 33d32d1f 2020-12-25 op 	char		*host;
114 33d32d1f 2020-12-25 op 	char		*port;
115 33d32d1f 2020-12-25 op 	uint16_t	 port_no;
116 33d32d1f 2020-12-25 op 	char		*path;
117 33d32d1f 2020-12-25 op 	char		*query;
118 33d32d1f 2020-12-25 op 	char		*fragment;
119 33d32d1f 2020-12-25 op };
120 33d32d1f 2020-12-25 op 
121 488f059a 2020-12-24 op enum {
122 488f059a 2020-12-24 op 	FILE_EXISTS,
123 488f059a 2020-12-24 op 	FILE_EXECUTABLE,
124 488f059a 2020-12-24 op 	FILE_DIRECTORY,
125 488f059a 2020-12-24 op 	FILE_MISSING,
126 488f059a 2020-12-24 op };
127 488f059a 2020-12-24 op 
128 33d32d1f 2020-12-25 op /* gmid.c */
129 488f059a 2020-12-24 op void		 siginfo_handler(int);
130 488f059a 2020-12-24 op int		 starts_with(const char*, const char*);
131 488f059a 2020-12-24 op 
132 488f059a 2020-12-24 op ssize_t		 filesize(int);
133 488f059a 2020-12-24 op 
134 488f059a 2020-12-24 op int		 start_reply(struct pollfd*, struct client*, int, const char*);
135 488f059a 2020-12-24 op const char	*path_ext(const char*);
136 488f059a 2020-12-24 op const char	*mime(const char*);
137 488f059a 2020-12-24 op int		 check_path(struct client*, const char*, int*);
138 488f059a 2020-12-24 op int		 check_for_cgi(char *, char*, struct pollfd*, struct client*);
139 488f059a 2020-12-24 op int		 open_file(char*, char*, struct pollfd*, struct client*);
140 488f059a 2020-12-24 op int		 start_cgi(const char*, const char*, const char*, struct pollfd*, struct client*);
141 488f059a 2020-12-24 op void		 cgi_setpoll_on_child(struct pollfd*, struct client*);
142 488f059a 2020-12-24 op void		 cgi_setpoll_on_client(struct pollfd*, struct client*);
143 488f059a 2020-12-24 op void		 handle_cgi(struct pollfd*, struct client*);
144 488f059a 2020-12-24 op void		 send_file(char*, char*, struct pollfd*, struct client*);
145 488f059a 2020-12-24 op void		 send_dir(char*, struct pollfd*, struct client*);
146 488f059a 2020-12-24 op void		 handle(struct pollfd*, struct client*);
147 488f059a 2020-12-24 op 
148 488f059a 2020-12-24 op void		 mark_nonblock(int);
149 488f059a 2020-12-24 op int		 make_soket(int);
150 488f059a 2020-12-24 op void		 do_accept(int, struct tls*, struct pollfd*, struct client*);
151 488f059a 2020-12-24 op void		 goodbye(struct pollfd*, struct client*);
152 488f059a 2020-12-24 op void		 loop(struct tls*, int);
153 488f059a 2020-12-24 op 
154 488f059a 2020-12-24 op void		 usage(const char*);
155 488f059a 2020-12-24 op 
156 33d32d1f 2020-12-25 op /* uri.c */
157 33d32d1f 2020-12-25 op int		 parse_uri(char*, struct uri*, const char**);
158 33d32d1f 2020-12-25 op int		 trim_req_uri(char*);
159 33d32d1f 2020-12-25 op 
160 488f059a 2020-12-24 op #endif