1 #ifndef PROJECT_H_INCLUDED
2 #define PROJECT_H_INCLUDED
3 #define PROJECT_H_VERSION "$Id: project.h,v 1.30 2001/09/13 23:52:00 jongfoster Exp $"
4 /*********************************************************************
6 * File : $Source: /cvsroot/ijbswa/current/project.h,v $
8 * Purpose : Defines data structures which are widely used in the
9 * project. Does not define any variables or functions
10 * (though it does declare some macros).
12 * Copyright : Written by and Copyright (C) 2001 the SourceForge
13 * IJBSWA team. http://ijbswa.sourceforge.net
15 * Based on the Internet Junkbuster originally written
16 * by and Copyright (C) 1997 Anonymous Coders and
17 * Junkbusters Corporation. http://www.junkbusters.com
19 * This program is free software; you can redistribute it
20 * and/or modify it under the terms of the GNU General
21 * Public License as published by the Free Software
22 * Foundation; either version 2 of the License, or (at
23 * your option) any later version.
25 * This program is distributed in the hope that it will
26 * be useful, but WITHOUT ANY WARRANTY; without even the
27 * implied warranty of MERCHANTABILITY or FITNESS FOR A
28 * PARTICULAR PURPOSE. See the GNU General Public
29 * License for more details.
31 * The GNU General Public License should be included with
32 * this file. If not, you can view it at
33 * http://www.gnu.org/copyleft/gpl.html
34 * or write to the Free Software Foundation, Inc., 59
35 * Temple Place - Suite 330, Boston, MA 02111-1307, USA.
39 * Revision 1.30 2001/09/13 23:52:00 jongfoster
40 * Support for both static and dynamically generated CGI pages
42 * Revision 1.29 2001/09/13 23:29:43 jongfoster
43 * Defining FORWARD_SPEC_INITIALIZER
45 * Revision 1.28 2001/09/13 23:05:50 jongfoster
46 * Changing the string paramater to the header parsers a "const".
48 * Revision 1.27 2001/08/05 16:06:20 jongfoster
49 * Modifiying "struct map" so that there are now separate header and
50 * "map_entry" structures. This means that functions which modify a
51 * map no longer need to return a pointer to the modified map.
52 * Also, it no longer reverses the order of the entries (which may be
53 * important with some advanced template substitutions).
55 * Revision 1.26 2001/07/30 22:08:36 jongfoster
56 * Tidying up #defines:
57 * - All feature #defines are now of the form FEATURE_xxx
58 * - Permanently turned off WIN_GUI_EDIT
59 * - Permanently turned on WEBDAV and SPLIT_PROXY_ARGS
61 * Revision 1.25 2001/07/29 18:43:08 jongfoster
62 * Changing #ifdef _FILENAME_H to FILENAME_H_INCLUDED, to conform to
65 * Revision 1.24 2001/07/25 17:20:27 oes
66 * Introduced http->user_agent
68 * Revision 1.23 2001/07/18 12:32:23 oes
69 * - Added ACTION_STRING_DEANIMATE
70 * - moved #define freez from jcc.h to project.h
72 * Revision 1.22 2001/07/15 17:51:41 jongfoster
73 * Renaming #define STATIC to STATIC_PCRE
75 * Revision 1.21 2001/07/13 14:03:19 oes
76 * - Reorganized regex header inclusion and #defines to
77 * comply to the scheme in configure.in
78 * - Added csp->content_type and its CT_* keys
79 * - Added ACTION_DEANIMATE
80 * - Removed all #ifdef PCRS
82 * Revision 1.20 2001/06/29 21:45:41 oes
83 * Indentation, CRLF->LF, Tab-> Space
85 * Revision 1.19 2001/06/29 13:33:36 oes
87 * - Introduced http_request.host_ip_addr_str
88 * - Introduced http_response.head_length
89 * - Introduced config.my_ip_addr_str, config.my_hostname,
90 * config.admin_address and config.proxy_info_url
91 * - Removed config.proxy_args_header and config.proxy_args_trailer,
92 * renamed config.proxy_args_invocation to config.proxy_args
93 * - Removed HTML snipplets and GIFs
94 * - Removed logentry from cancelled commit
96 * Revision 1.18 2001/06/09 10:57:39 jongfoster
97 * Adding definition of BUFFER_SIZE.
98 * Changing struct cgi_dispatcher to use "const" strings.
100 * Revision 1.17 2001/06/07 23:15:09 jongfoster
101 * Merging ACL and forward files into config file.
102 * Moving struct gateway members into struct forward_spec
103 * Removing config->proxy_args_gateways
104 * Cosmetic: Adding a few comments
106 * Revision 1.16 2001/06/04 18:31:58 swa
107 * files are now prefixed with either `confdir' or `logdir'.
108 * `make redhat-dist' replaces both entries confdir and logdir
111 * Revision 1.15 2001/06/04 11:28:53 swa
112 * redirect did not work due to missing /
114 * Revision 1.14 2001/06/03 11:03:48 oes
116 * added struct http_response,
117 * changed struct interceptors to struct cgi_dispatcher,
118 * moved HTML stuff to cgi.h
120 * Revision 1.13 2001/06/01 20:05:36 jongfoster
121 * Support for +image-blocker{}: added ACTION_IMAGE_BLOCKER
122 * constant, and removed csp->tinygif.
124 * Revision 1.12 2001/06/01 18:49:17 jongfoster
125 * Replaced "list_share" with "list" - the tiny memory gain was not
126 * worth the extra complexity.
128 * Revision 1.11 2001/06/01 10:32:47 oes
129 * Added constants for anchoring selection bitmap
131 * Revision 1.10 2001/05/31 21:33:53 jongfoster
132 * Changes for new actions file, replacing permissionsfile
133 * and parts of the config file. Also added support for
136 * Revision 1.9 2001/05/31 17:32:31 oes
138 * - Enhanced domain part globbing with infix and prefix asterisk
139 * matching and optional unanchored operation
141 * Revision 1.8 2001/05/29 20:09:15 joergs
142 * HTTP_REDIRECT_TEMPLATE fixed.
144 * Revision 1.7 2001/05/29 09:50:24 jongfoster
145 * Unified blocklist/imagelist/actionslist.
146 * File format is still under discussion, but the internal changes
149 * Also modified interceptor behaviour:
150 * - We now intercept all URLs beginning with one of the following
151 * prefixes (and *only* these prefixes):
153 * * http://ijbswa.sf.net/config/
154 * * http://ijbswa.sourceforge.net/config/
155 * - New interceptors "home page" - go to http://i.j.b/ to see it.
156 * - Internal changes so that intercepted and fast redirect pages
157 * are not replaced with an image.
158 * - Interceptors now have the option to send a binary page direct
159 * to the client. (i.e. ijb-send-banner uses this)
160 * - Implemented show-url-info interceptor. (Which is why I needed
161 * the above interceptors changes - a typical URL is
162 * "http://i.j.b/show-url-info?url=www.somesite.com/banner.gif".
163 * The previous mechanism would not have intercepted that, and
164 * if it had been intercepted then it then it would have replaced
167 * Revision 1.6 2001/05/27 22:17:04 oes
169 * - re_process_buffer no longer writes the modified buffer
170 * to the client, which was very ugly. It now returns the
171 * buffer, which it is then written by chat.
173 * - content_length now adjusts the Content-Length: header
174 * for modified documents rather than crunch()ing it.
175 * (Length info in csp->content_length, which is 0 for
176 * unmodified documents)
178 * - For this to work, sed() is called twice when filtering.
180 * Revision 1.5 2001/05/26 00:28:36 jongfoster
181 * Automatic reloading of config file.
182 * Removed obsolete SIGHUP support (Unix) and Reload menu option (Win32).
183 * Most of the global variables have been moved to a new
184 * struct configuration_spec, accessed through csp->config->globalname
185 * Most of the globals remaining are used by the Win32 GUI.
187 * Revision 1.4 2001/05/22 18:46:04 oes
189 * - Enabled filtering banners by size rather than URL
190 * by adding patterns that replace all standard banner
191 * sizes with the "Junkbuster" gif to the re_filterfile
193 * - Enabled filtering WebBugs by providing a pattern
194 * which kills all 1x1 images
196 * - Added support for PCRE_UNGREEDY behaviour to pcrs,
197 * which is selected by the (nonstandard and therefore
198 * capital) letter 'U' in the option string.
199 * It causes the quantifiers to be ungreedy by default.
200 * Appending a ? turns back to greedy (!).
202 * - Added a new interceptor ijb-send-banner, which
203 * sends back the "Junkbuster" gif. Without imagelist or
204 * MSIE detection support, or if tinygif = 1, or the
205 * URL isn't recognized as an imageurl, a lame HTML
206 * explanation is sent instead.
208 * - Added new feature, which permits blocking remote
209 * script redirects and firing back a local redirect
211 * The feature is conditionally compiled, i.e. it
212 * can be disabled with --disable-fast-redirects,
213 * plus it must be activated by a "fast-redirects"
214 * line in the config file, has its own log level
215 * and of course wants to be displayed by show-proxy-args
216 * Note: Boy, all the #ifdefs in 1001 locations and
217 * all the fumbling with configure.in and acconfig.h
218 * were *way* more work than the feature itself :-(
220 * - Because a generic redirect template was needed for
221 * this, tinygif = 3 now uses the same.
223 * - Moved GIFs, and other static HTTP response templates
228 * - Removed some >400 CRs again (Jon, you really worked
231 * Revision 1.3 2001/05/20 01:21:20 jongfoster
232 * Version 2.9.4 checkin.
233 * - Merged popupfile and cookiefile, and added control over PCRS
234 * filtering, in new "actionsfile".
235 * - Implemented LOG_LEVEL_FATAL, so that if there is a configuration
236 * file error you now get a message box (in the Win32 GUI) rather
237 * than the program exiting with no explanation.
238 * - Made killpopup use the PCRS MIME-type checking and HTTP-header
240 * - Removed tabs from "config"
241 * - Moved duplicated url parsing code in "loaders.c" to a new funcition.
242 * - Bumped up version number.
244 * Revision 1.2 2001/05/17 23:01:01 oes
245 * - Cleaned CRLF's from the sources and related files
247 * Revision 1.1.1.1 2001/05/15 13:59:03 oes
248 * Initial import of version 2.9.3 source tree
251 *********************************************************************/
254 /* Declare struct FILE for vars and funcs. */
257 /* Need time_t for file_list */
261 * Include appropriate regular expression libraries.
262 * Note that pcrs and pcre (native) are needed for cgi
263 * and are included anyway.
266 #if defined(REGEX_PCRE) || defined (REGEX_GNU)
268 #endif /* defined(REGEX_PCRE) || defined (REGEX_GNU) */
282 #if defined(REGEX_PCRE)
284 # include "pcreposix.h"
286 # include <pcreposix.h>
288 #endif /* defined(REGEX_PCRE) */
290 #if defined(REGEX_GNU)
291 # include "gnu_regex.h"
296 #endif /* def AMIGA */
302 #define freez(X) if(X) free(X); X = NULL
304 #define BUFFER_SIZE 5000
308 /* Default IP and port to listen on */
309 #define HADDR_DEFAULT "127.0.0.1"
310 #define HADDR_PORT 8000
313 /* Forward defs for various structures */
315 /* Need this for struct client_state */
316 struct configuration_spec;
319 /* Generic linked list of strings */
324 struct list_entry *next;
329 struct list_entry *first;
330 struct list_entry *last;
334 /* A map from a string to another string */
340 struct map_entry *next;
345 struct map_entry *first;
346 struct map_entry *last;
355 char *host_ip_addr_str; /* NULL before connect_to() */
359 char *hostport; /* "host[:port]" */
361 char *user_agent; /* Client's User-Agent: header value */
364 /* Response generated by CGI, blocker, or error handler */
367 char *status; /* HTTP status (string)*/
368 struct list headers[1]; /* List of header lines */
369 char *head; /* Formatted http response head */
370 int head_length; /* Length of http response head */
371 char *body; /* HTTP document body */
372 int content_length; /* Length of body, REQUIRED if binary body */
373 int is_static; /* Nonzero if the content will never change and
374 * should be cached by the broser (e.g. images) */
380 char *spec; /* The string which was parsed to produce this */
381 /* url_spec. Used for debugging or display only. */
383 /* Hostname matching: */
384 char *domain; /* Fully qalified domain name (FQDN) pattern. */
385 /* May contain "*". */
386 char *dbuf; /* Buffer with '\0'-delimited fqdn */
387 char **dvec; /* Domain ptr vector into dbuf */
388 int dcnt; /* How many domains in fqdn? */
389 int unanchored; /* Bitmap - flags are ANCHOR_LEFT and ANCHOR_RIGHT */
392 int port; /* The port number, or 0 to match all ports. */
395 char *path; /* The path prefix (if not using regex), or source */
397 int pathlen; /* ==strlen(path). Needed for prefix matching. */
399 regex_t *preg; /* Regex for matching path part */
403 #define URL_SPEC_INITIALIZER { NULL, NULL, NULL, NULL, 0, 0, 0, NULL, 0, NULL }
404 #else /* ifndef REGEX */
405 #define URL_SPEC_INITIALIZER { NULL, NULL, NULL, NULL, 0, 0, 0, NULL, 0 }
406 #endif /* ndef REGEX */
408 /* Constants for host part matching in URLs */
409 #define ANCHOR_LEFT 1
410 #define ANCHOR_RIGHT 2
422 #define IOB_PEEK(CSP) ((CSP->iob->cur > CSP->iob->eod) ? (CSP->iob->eod - CSP->iob->cur) : 0)
423 #define IOB_RESET(CSP) if(CSP->iob->buf) free(CSP->iob->buf); memset(CSP->iob, '\0', sizeof(CSP->iob));
425 /* Keys for csp->content_type */
426 #define CT_TEXT 0x01U
429 #define ACTION_MASK_ALL (~0U)
431 #define ACTION_MOST_COMPATIBLE 0x0000U
433 #define ACTION_BLOCK 0x0001U
434 #define ACTION_DEANIMATE 0x2000U
435 #define ACTION_FAST_REDIRECTS 0x0002U
436 #define ACTION_FILTER 0x0004U
437 #define ACTION_HIDE_FORWARDED 0x0008U
438 #define ACTION_HIDE_FROM 0x0010U
439 #define ACTION_HIDE_REFERER 0x0020U /* sic - follow HTTP, not English */
440 #define ACTION_HIDE_USER_AGENT 0x0040U
441 #define ACTION_IMAGE 0x0080U
442 #define ACTION_IMAGE_BLOCKER 0x0100U
443 #define ACTION_NO_COOKIE_READ 0x0200U
444 #define ACTION_NO_COOKIE_SET 0x0400U
445 #define ACTION_NO_POPUPS 0x0800U
446 #define ACTION_VANILLA_WAFER 0x1000U
448 #define ACTION_STRING_DEANIMATE 0
449 #define ACTION_STRING_FROM 1
450 #define ACTION_STRING_IMAGE_BLOCKER 2
451 #define ACTION_STRING_REFERER 3
452 #define ACTION_STRING_USER_AGENT 4
453 #define ACTION_STRING_COUNT 5
456 #define ACTION_MULTI_ADD_HEADER 0
457 #define ACTION_MULTI_WAFER 1
458 #define ACTION_MULTI_COUNT 2
461 * This structure contains a list of actions to apply to a URL.
462 * It only contains positive instructions - no "-" options.
463 * It is not used to store the actions list itself, only for
464 * url_actions() to return the current values.
466 struct current_action_spec
468 unsigned flags; /* a bit set to "1" = add action */
470 /* For those actions that require parameters: */
472 /* each entry is valid if & only if corresponding entry in "add" set. */
473 char * string[ACTION_STRING_COUNT];
476 struct list multi[ACTION_MULTI_COUNT][1];
481 * This structure contains a set of changes to actions.
482 * It can contain both positive and negative instructions.
483 * It is used to store an entry in the actions list.
487 unsigned mask; /* a bit set to "0" = remove action */
488 unsigned add; /* a bit set to "1" = add action */
490 /* For those actions that require parameters: */
492 /* each entry is valid if & only if corresponding entry in "add" set. */
493 char * string[ACTION_STRING_COUNT];
495 /* Strings to remove. */
496 struct list multi_remove[ACTION_MULTI_COUNT][1];
498 /* If nonzero, remove *all* strings. */
499 int multi_remove_all[ACTION_MULTI_COUNT];
502 struct list multi_add[ACTION_MULTI_COUNT][1];
506 * This structure is used to store the actions list.
508 * It contains a URL pattern, and the chages to the actions.
509 * It is a linked list.
513 struct url_spec url[1];
515 struct action_spec action[1];
517 struct url_actions * next;
521 /* Constants defining bitmask for csp->accept_types */
523 #ifdef FEATURE_IMAGE_DETECT_MSIE
525 /* MSIE detected by user-agent string */
526 #define ACCEPT_TYPE_IS_MSIE 0x0001
529 * *If* this is MSIE, it wants an image. (Or this is a shift-reload, or
530 * it's got an image from this URL before... yuck!)
531 * Only meaningful if ACCEPT_TYPE_IS_MSIE set
533 #define ACCEPT_TYPE_MSIE_IMAGE 0x0002
536 * *If* this is MSIE, it wants a HTML document.
537 * Only meaningful if ACCEPT_TYPE_IS_MSIE set
539 #define ACCEPT_TYPE_MSIE_HTML 0x0004
541 #endif /* def FEATURE_IMAGE_DETECT_MSIE */
546 /* The proxy's configuration */
547 struct configuration_spec * config;
549 /* The actions to perform on the current request */
550 struct current_action_spec action[1];
552 /* socket to talk to client (web browser) */
555 /* socket to talk to server (web server or proxy) */
559 #ifdef FEATURE_STATISTICS
560 /* 1 if this URL was rejected, 0 otherwise. Allows actual stats inc to
561 * occur in main thread only for thread-safety.
564 #endif /* def FEATURE_STATISTICS */
566 #ifdef FEATURE_FORCE_LOAD
568 #endif /* def FEATURE_FORCE_LOAD */
570 #ifdef FEATURE_TOGGLE
572 #endif /* def FEATURE_TOGGLE */
575 * Client PC's IP address, as reported by the accept()_ function.
576 * Both as string and number
582 /* Our IP address and hostname, i.e. the IP address that
583 the client used to reach us, and the associated hostname,
586 char *my_ip_addr_str;
590 /* The referer in this request, if one was specified. */
592 #endif /* def FEATURE_TRUST */
594 #if defined(FEATURE_IMAGE_DETECT_MSIE)
595 /* Types the client will accept.
596 * Bitmask - see ACCEPT_TYPE_XXX constants.
599 #endif /* defined(FEATURE_IMAGE_DETECT_MSIE) */
601 /* The URL that was requested */
602 struct http_request http[1];
604 /* An I/O buffer used for buffering data read from the client */
607 /* List of all headers for this request */
608 struct list headers[1];
610 /* List of all cookies for this request */
611 struct list cookie_list[1];
613 /* MIME-Type bitmap, see CT_* above */
614 unsigned char content_type;
616 /* The "X-Forwarded-For:" header sent by the client */
620 * Nonzero if this client is processing data.
621 * Set to zero when the thread associated with this structure dies.
625 /* files associated with this client */
626 struct file_list *actions_list;
628 struct file_list *rlist; /* pcrs job file */
629 size_t content_length; /* Length after content modification */
632 struct file_list *tlist; /* trustfile */
633 #endif /* def FEATURE_TRUST */
635 struct client_state *next;
643 char *(*parser)(const struct parsers *, const char *, struct client_state *);
646 struct cgi_dispatcher
650 int (*handler)(struct client_state *csp, struct http_response *rsp, struct map *parameters);
651 const char *description;
657 * this is a pointer to the data structures associated with the file.
658 * Read-only once the structure has been created.
662 /* Normally NULL. When we are finished with file (i.e. when we have
663 * loaded a new one), set to a pointer to an unloader function.
664 * Unloader will be called by sweep() (called from main loop) when
665 * all clients using this file are done. This prevents threading
668 void (*unloader)(void *);
670 /* Used internally by sweep(). Do not access from elsewhere. */
673 /* Following variables allow us to check if file has been changed.
674 * Read-only once the structure has been created.
679 /* Pointer to next entry in the linked list of all "file_list"s.
680 * This linked list is so that sweep() can navigate it.
681 * Since sweep() can remove items from the list, we must be careful
682 * to only access this value from main thread (when we know sweep
685 struct file_list *next;
692 struct url_spec url[1];
694 struct block_spec *next;
696 #endif /* def FEATURE_TRUST */
699 #define SOCKS_NONE 0 /* Don't use a SOCKS server */
700 #define SOCKS_4 40 /* original SOCKS 4 protocol */
701 #define SOCKS_4A 41 /* as modified for hosts w/o external DNS */
705 struct url_spec url[1];
707 /* Connection type - must be a SOCKS_xxx constant */
714 /* Parent HTTP proxy */
718 /* For the linked list */
719 struct forward_spec *next;
721 #define FORWARD_SPEC_INITIALIZER { URL_SPEC_INITIALIZER, 0, NULL, 0, NULL, 0, NULL }
724 struct re_filterfile_spec
728 struct list patterns[1];
733 #define ACL_PERMIT 1 /* accept connection request */
734 #define ACL_DENY 2 /* reject connection request */
736 struct access_control_addr
743 struct access_control_list
745 struct access_control_addr src[1];
746 struct access_control_addr dst[1];
749 struct access_control_list *next;
751 #endif /* def FEATURE_ACL */
754 /* Maximum number of loaders (actions, re_filter, ...) */
758 * Data loaded from the configuration file.
760 * (Anomaly: toggle is still handled through a global, not this structure)
762 struct configuration_spec
771 const char *actions_file;
773 /* The administrator's email address */
776 /* A URL with info on this proxy */
777 char *proxy_info_url;
779 const char *re_filterfile;
781 #ifdef FEATURE_COOKIE_JAR
782 const char * jarfile;
784 #endif /* def FEATURE_COOKIE_JAR */
787 * Port and IP to bind to.
788 * Defaults to HADDR_DEFAULT:HADDR_PORT == 127.0.0.1:8000
794 const char * trustfile;
796 struct list trust_info[1];
797 struct url_spec *trust_list[64];
798 #endif /* def FEATURE_TRUST */
801 struct access_control_list *acl;
802 #endif /* def FEATURE_ACL */
804 struct forward_spec *forward;
806 /* All options from the config file, HTML-formatted */
809 /* the configuration file object. */
810 struct file_list *config_file_list;
812 /* List of loaders */
813 int (*loaders[NLOADERS])(struct client_state *);
815 /* bool, nonzero if we need to bind() to the new port */
820 #define SZ(X) (sizeof(X) / sizeof(*X))
822 #ifdef FEATURE_FORCE_LOAD
823 #define FORCE_PREFIX "/IJB-FORCE-LOAD"
824 #endif /* def FEATURE_FORCE_LOAD */
827 #define HOME_PAGE_URL "http://ijbswa.sourceforge.net"
828 #define REDIRECT_URL HOME_PAGE_URL "/redirect.php?v=" VERSION "&to="
829 #define CGI_PREFIX_HOST "i.j.b"
832 static const char CSUCCEED[] =
833 "HTTP/1.0 200 Connection established\n"
834 "Proxy-Agent: IJ/" VERSION "\n\n";
836 static const char CHEADER[] =
837 "HTTP/1.0 400 Invalid header received from browser\n\n";
843 #endif /* ndef PROJECT_H_INCLUDED */