3 #define PROJECT_H_VERSION "$Id: project.h,v 1.20 2001/06/29 21:45:41 oes 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 1.20 2001/06/29 21:45:41 oes
40 * Indentation, CRLF->LF, Tab-> Space
42 * Revision 1.19 2001/06/29 13:33:36 oes
44 * - Introduced http_request.host_ip_addr_str
45 * - Introduced http_response.head_length
46 * - Introduced config.my_ip_addr_str, config.my_hostname,
47 * config.admin_address and config.proxy_info_url
48 * - Removed config.proxy_args_header and config.proxy_args_trailer,
49 * renamed config.proxy_args_invocation to config.proxy_args
50 * - Removed HTML snipplets and GIFs
51 * - Removed logentry from cancelled commit
53 * Revision 1.18 2001/06/09 10:57:39 jongfoster
54 * Adding definition of BUFFER_SIZE.
55 * Changing struct cgi_dispatcher to use "const" strings.
57 * Revision 1.17 2001/06/07 23:15:09 jongfoster
58 * Merging ACL and forward files into config file.
59 * Moving struct gateway members into struct forward_spec
60 * Removing config->proxy_args_gateways
61 * Cosmetic: Adding a few comments
63 * Revision 1.16 2001/06/04 18:31:58 swa
64 * files are now prefixed with either `confdir' or `logdir'.
65 * `make redhat-dist' replaces both entries confdir and logdir
68 * Revision 1.15 2001/06/04 11:28:53 swa
69 * redirect did not work due to missing /
71 * Revision 1.14 2001/06/03 11:03:48 oes
73 * added struct http_response,
74 * changed struct interceptors to struct cgi_dispatcher,
75 * moved HTML stuff to cgi.h
77 * Revision 1.13 2001/06/01 20:05:36 jongfoster
78 * Support for +image-blocker{}: added ACTION_IMAGE_BLOCKER
79 * constant, and removed csp->tinygif.
81 * Revision 1.12 2001/06/01 18:49:17 jongfoster
82 * Replaced "list_share" with "list" - the tiny memory gain was not
83 * worth the extra complexity.
85 * Revision 1.11 2001/06/01 10:32:47 oes
86 * Added constants for anchoring selection bitmap
88 * Revision 1.10 2001/05/31 21:33:53 jongfoster
89 * Changes for new actions file, replacing permissionsfile
90 * and parts of the config file. Also added support for
93 * Revision 1.9 2001/05/31 17:32:31 oes
95 * - Enhanced domain part globbing with infix and prefix asterisk
96 * matching and optional unanchored operation
98 * Revision 1.8 2001/05/29 20:09:15 joergs
99 * HTTP_REDIRECT_TEMPLATE fixed.
101 * Revision 1.7 2001/05/29 09:50:24 jongfoster
102 * Unified blocklist/imagelist/actionslist.
103 * File format is still under discussion, but the internal changes
106 * Also modified interceptor behaviour:
107 * - We now intercept all URLs beginning with one of the following
108 * prefixes (and *only* these prefixes):
110 * * http://ijbswa.sf.net/config/
111 * * http://ijbswa.sourceforge.net/config/
112 * - New interceptors "home page" - go to http://i.j.b/ to see it.
113 * - Internal changes so that intercepted and fast redirect pages
114 * are not replaced with an image.
115 * - Interceptors now have the option to send a binary page direct
116 * to the client. (i.e. ijb-send-banner uses this)
117 * - Implemented show-url-info interceptor. (Which is why I needed
118 * the above interceptors changes - a typical URL is
119 * "http://i.j.b/show-url-info?url=www.somesite.com/banner.gif".
120 * The previous mechanism would not have intercepted that, and
121 * if it had been intercepted then it then it would have replaced
124 * Revision 1.6 2001/05/27 22:17:04 oes
126 * - re_process_buffer no longer writes the modified buffer
127 * to the client, which was very ugly. It now returns the
128 * buffer, which it is then written by chat.
130 * - content_length now adjusts the Content-Length: header
131 * for modified documents rather than crunch()ing it.
132 * (Length info in csp->content_length, which is 0 for
133 * unmodified documents)
135 * - For this to work, sed() is called twice when filtering.
137 * Revision 1.5 2001/05/26 00:28:36 jongfoster
138 * Automatic reloading of config file.
139 * Removed obsolete SIGHUP support (Unix) and Reload menu option (Win32).
140 * Most of the global variables have been moved to a new
141 * struct configuration_spec, accessed through csp->config->globalname
142 * Most of the globals remaining are used by the Win32 GUI.
144 * Revision 1.4 2001/05/22 18:46:04 oes
146 * - Enabled filtering banners by size rather than URL
147 * by adding patterns that replace all standard banner
148 * sizes with the "Junkbuster" gif to the re_filterfile
150 * - Enabled filtering WebBugs by providing a pattern
151 * which kills all 1x1 images
153 * - Added support for PCRE_UNGREEDY behaviour to pcrs,
154 * which is selected by the (nonstandard and therefore
155 * capital) letter 'U' in the option string.
156 * It causes the quantifiers to be ungreedy by default.
157 * Appending a ? turns back to greedy (!).
159 * - Added a new interceptor ijb-send-banner, which
160 * sends back the "Junkbuster" gif. Without imagelist or
161 * MSIE detection support, or if tinygif = 1, or the
162 * URL isn't recognized as an imageurl, a lame HTML
163 * explanation is sent instead.
165 * - Added new feature, which permits blocking remote
166 * script redirects and firing back a local redirect
168 * The feature is conditionally compiled, i.e. it
169 * can be disabled with --disable-fast-redirects,
170 * plus it must be activated by a "fast-redirects"
171 * line in the config file, has its own log level
172 * and of course wants to be displayed by show-proxy-args
173 * Note: Boy, all the #ifdefs in 1001 locations and
174 * all the fumbling with configure.in and acconfig.h
175 * were *way* more work than the feature itself :-(
177 * - Because a generic redirect template was needed for
178 * this, tinygif = 3 now uses the same.
180 * - Moved GIFs, and other static HTTP response templates
185 * - Removed some >400 CRs again (Jon, you really worked
188 * Revision 1.3 2001/05/20 01:21:20 jongfoster
189 * Version 2.9.4 checkin.
190 * - Merged popupfile and cookiefile, and added control over PCRS
191 * filtering, in new "actionsfile".
192 * - Implemented LOG_LEVEL_FATAL, so that if there is a configuration
193 * file error you now get a message box (in the Win32 GUI) rather
194 * than the program exiting with no explanation.
195 * - Made killpopup use the PCRS MIME-type checking and HTTP-header
197 * - Removed tabs from "config"
198 * - Moved duplicated url parsing code in "loaders.c" to a new funcition.
199 * - Bumped up version number.
201 * Revision 1.2 2001/05/17 23:01:01 oes
202 * - Cleaned CRLF's from the sources and related files
204 * Revision 1.1.1.1 2001/05/15 13:59:03 oes
205 * Initial import of version 2.9.3 source tree
208 *********************************************************************/
211 /* Declare struct FILE for vars and funcs. */
214 /* Need time_t for file_list */
218 * Include appropriate regular expression libraries.
219 * Note that pcrs and pcre (native) are needed for cgi
220 * and are included anyway.
223 #if defined(REGEX_PCRE) || defined (REGEX_GNU)
225 #endif /* defined(REGEX_PCRE) || defined (REGEX_GNU) */
239 #if defined(REGEX_PCRE)
241 # include "pcreposix.h"
243 # include <pcreposix.h>
245 #endif /* defined(REGEX_PCRE) */
247 #if defined(REGEX_GNU)
248 # include "gnu_regex.h"
253 #endif /* def AMIGA */
259 #define BUFFER_SIZE 5000
263 /* Default IP and port to listen on */
264 #define HADDR_DEFAULT "127.0.0.1"
265 #define HADDR_PORT 8000
268 /* Need this for struct client_state */
269 struct configuration_spec;
271 /* Generic linked list of strings */
272 struct list /* FIXME: Why not separate entries and header? */
274 char * str; /* valid in an entry */
275 struct list *last; /* valid in header */
291 char *host_ip_addr_str; /* NULL before connect_to() */
295 char *hostport; /* "host[:port]" */
299 /* Response generated by CGI, blocker, or error handler */
302 char *status; /* HTTP status (string)*/
303 struct list headers[1]; /* List of header lines */
304 char *head; /* Formatted http response head */
305 int head_length; /* Length of http response head */
306 char *body; /* HTTP document body */
307 int content_length; /* Length of body, REQUIRED if binary body*/
313 char *spec; /* The string which was parsed to produce this */
314 /* url_spec. Used for debugging or display only. */
316 /* Hostname matching: */
317 char *domain; /* Fully qalified domain name (FQDN) pattern. */
318 /* May contain "*". */
319 char *dbuf; /* Buffer with '\0'-delimited fqdn */
320 char **dvec; /* Domain ptr vector into dbuf */
321 int dcnt; /* How many domains in fqdn? */
322 int unanchored; /* Bitmap - flags are ANCHOR_LEFT and ANCHOR_RIGHT */
325 int port; /* The port number, or 0 to match all ports. */
328 char *path; /* The path prefix (if not using regex), or source */
330 int pathlen; /* ==strlen(path). Needed for prefix matching. */
332 regex_t *preg; /* Regex for matching path part */
336 #define ANCHOR_LEFT 1
337 #define ANCHOR_RIGHT 2
350 #define IOB_PEEK(CSP) ((CSP->iob->cur > CSP->iob->eod) ? (CSP->iob->eod - CSP->iob->cur) : 0)
351 #define IOB_RESET(CSP) if(CSP->iob->buf) free(CSP->iob->buf); memset(CSP->iob, '\0', sizeof(CSP->iob));
353 /* Keys for csp->content_type */
354 #define CT_TEXT 0x01U
357 #define ACTION_MASK_ALL (~0U)
359 #define ACTION_MOST_COMPATIBLE 0x0000U
361 #define ACTION_BLOCK 0x0001U
362 #define ACTION_DEANIMATE 0x2000U
363 #define ACTION_FAST_REDIRECTS 0x0002U
364 #define ACTION_FILTER 0x0004U
365 #define ACTION_HIDE_FORWARDED 0x0008U
366 #define ACTION_HIDE_FROM 0x0010U
367 #define ACTION_HIDE_REFERER 0x0020U /* sic - follow HTTP, not English */
368 #define ACTION_HIDE_USER_AGENT 0x0040U
369 #define ACTION_IMAGE 0x0080U
370 #define ACTION_IMAGE_BLOCKER 0x0100U
371 #define ACTION_NO_COOKIE_READ 0x0200U
372 #define ACTION_NO_COOKIE_SET 0x0400U
373 #define ACTION_NO_POPUPS 0x0800U
374 #define ACTION_VANILLA_WAFER 0x1000U
376 #define ACTION_STRING_FROM 0
377 #define ACTION_STRING_IMAGE_BLOCKER 1
378 #define ACTION_STRING_REFERER 2
379 #define ACTION_STRING_USER_AGENT 3
380 #define ACTION_STRING_COUNT 4
382 #define ACTION_MULTI_ADD_HEADER 0
383 #define ACTION_MULTI_WAFER 1
384 #define ACTION_MULTI_COUNT 2
387 * This structure contains a list of actions to apply to a URL.
388 * It only contains positive instructions - no "-" options.
389 * It is not used to store the actions list itself, only for
390 * url_actions() to return the current values.
392 struct current_action_spec
394 unsigned flags; /* a bit set to "1" = add action */
396 /* For those actions that require parameters: */
398 /* each entry is valid if & only if corresponding entry in "add" set. */
399 char * string[ACTION_STRING_COUNT];
402 struct list multi[ACTION_MULTI_COUNT][1];
407 * This structure contains a set of changes to actions.
408 * It can contain both positive and negative instructions.
409 * It is used to store an entry in the actions list.
413 unsigned mask; /* a bit set to "0" = remove action */
414 unsigned add; /* a bit set to "1" = add action */
416 /* For those actions that require parameters: */
418 /* each entry is valid if & only if corresponding entry in "add" set. */
419 char * string[ACTION_STRING_COUNT];
421 /* Strings to remove. */
422 struct list multi_remove[ACTION_MULTI_COUNT][1];
424 /* If nonzero, remove *all* strings. */
425 int multi_remove_all[ACTION_MULTI_COUNT];
428 struct list multi_add[ACTION_MULTI_COUNT][1];
432 * This structure is used to store the actions list.
434 * It contains a URL pattern, and the chages to the actions.
435 * It is a linked list.
439 struct url_spec url[1];
441 struct action_spec action[1];
443 struct url_actions * next;
447 /* Constants defining bitmask for csp->accept_types */
449 #ifdef DETECT_MSIE_IMAGES
451 /* MSIE detected by user-agent string */
452 #define ACCEPT_TYPE_IS_MSIE 0x0001
455 * *If* this is MSIE, it wants an image. (Or this is a shift-reload, or
456 * it's got an image from this URL before... yuck!)
457 * Only meaningful if ACCEPT_TYPE_IS_MSIE set
459 #define ACCEPT_TYPE_MSIE_IMAGE 0x0002
462 * *If* this is MSIE, it wants a HTML document.
463 * Only meaningful if ACCEPT_TYPE_IS_MSIE set
465 #define ACCEPT_TYPE_MSIE_HTML 0x0004
467 #endif /* def DETECT_MSIE_IMAGES */
472 /* The proxy's configuration */
473 struct configuration_spec * config;
475 /* The actions to perform on the current request */
476 struct current_action_spec action[1];
478 /* socket to talk to client (web browser) */
481 /* socket to talk to server (web server or proxy) */
486 /* 1 if this URL was rejected, 0 otherwise. Allows actual stats inc to
487 * occur in main thread only for thread-safety.
490 #endif /* def STATISTICS */
494 #endif /* def FORCE_LOAD */
498 #endif /* def TOGGLE */
501 * Client PC's IP address, as reported by the accept()_ function.
502 * Both as string and number
508 /* Our IP address and hostname, i.e. the IP address that
509 the client used to reach us, and the associated hostname,
512 char *my_ip_addr_str;
516 /* The referer in this request, if one was specified. */
518 #endif /* def TRUST_FILES */
520 #if defined(DETECT_MSIE_IMAGES)
521 /* Types the client will accept.
522 * Bitmask - see ACCEPT_TYPE_XXX constants.
525 #endif /* defined(DETECT_MSIE_IMAGES) */
527 /* The URL that was requested */
528 struct http_request http[1];
530 /* An I/O buffer used for buffering data read from the client */
533 /* List of all headers for this request */
534 struct list headers[1];
536 /* List of all cookies for this request */
537 struct list cookie_list[1];
539 /* MIME-Type bitmap, see CT_* above */
540 unsigned char content_type;
542 /* The "X-Forwarded-For:" header sent by the client */
546 * Nonzero if this client is processing data.
547 * Set to zero when the thread associated with this structure dies.
551 /* files associated with this client */
552 struct file_list *actions_list;
554 struct file_list *rlist; /* pcrs job file */
555 size_t content_length; /* Length after content modification */
558 struct file_list *tlist; /* trustfile */
559 #endif /* def TRUST_FILES */
561 struct client_state *next;
569 char *(*parser)(const struct parsers *, char *, struct client_state *);
572 struct cgi_dispatcher
576 int (*handler)(struct client_state *csp, struct http_response *rsp, struct map *parameters);
577 const char *description;
583 * this is a pointer to the data structures associated with the file.
584 * Read-only once the structure has been created.
588 /* Normally NULL. When we are finished with file (i.e. when we have
589 * loaded a new one), set to a pointer to an unloader function.
590 * Unloader will be called by sweep() (called from main loop) when
591 * all clients using this file are done. This prevents threading
594 void (*unloader)(void *);
596 /* Used internally by sweep(). Do not access from elsewhere. */
599 #ifndef SPLIT_PROXY_ARGS
600 /* String to be displayed as part of show-proxy-args display.
601 * Read-only once the structure has been created.
604 #endif /* ndef SPLIT_PROXY_ARGS */
606 /* Following variables allow us to check if file has been changed.
607 * Read-only once the structure has been created.
612 /* Pointer to next entry in the linked list of all "file_list"s.
613 * This linked list is so that sweep() can navigate it.
614 * Since sweep() can remove items from the list, we must be careful
615 * to only access this value from main thread (when we know sweep
618 struct file_list *next;
625 struct url_spec url[1];
627 struct block_spec *next;
629 #endif /* def TRUST_FILES */
632 #define SOCKS_NONE 0 /* Don't use a SOCKS server */
633 #define SOCKS_4 40 /* original SOCKS 4 protocol */
634 #define SOCKS_4A 41 /* as modified for hosts w/o external DNS */
638 struct url_spec url[1];
640 /* Connection type - must be a SOCKS_xxx constant */
647 /* Parent HTTP proxy */
651 /* For the linked list */
652 struct forward_spec *next;
655 struct re_filterfile_spec
657 struct list patterns[1];
662 #define ACL_PERMIT 1 /* accept connection request */
663 #define ACL_DENY 2 /* reject connection request */
665 struct access_control_addr
672 struct access_control_list
674 struct access_control_addr src[1];
675 struct access_control_addr dst[1];
678 struct access_control_list *next;
680 #endif /* def ACL_FILES */
683 /* Maximum number of loaders (actions, re_filter, ...) */
687 * Data loaded from the configuration file.
689 * (Anomaly: toggle is still handled through a global, not this structure)
691 struct configuration_spec
700 const char *actions_file;
702 /* The administrator's email address */
705 /* A URL with info on this proxy */
706 char *proxy_info_url;
708 const char *re_filterfile;
711 const char * jarfile;
713 #endif /* def JAR_FILES */
716 * Port and IP to bind to.
717 * Defaults to HADDR_DEFAULT:HADDR_PORT == 127.0.0.1:8000
722 #ifndef SPLIT_PROXY_ARGS
723 const char *suppress_message;
724 #endif /* ndef SPLIT_PROXY_ARGS */
726 #ifndef SPLIT_PROXY_ARGS
727 /* suppress listing config files */
728 int suppress_blocklists;
729 #endif /* ndef SPLIT_PROXY_ARGS */
732 const char * trustfile;
734 struct list trust_info[1];
735 struct url_spec *trust_list[64];
736 #endif /* def TRUST_FILES */
739 struct access_control_list *acl;
740 #endif /* def ACL_FILES */
742 struct forward_spec *forward;
744 /* All options from the config file, HTML-formatted */
747 /* the configuration file object. */
748 struct file_list *config_file_list;
750 /* List of loaders */
751 int (*loaders[NLOADERS])(struct client_state *);
753 /* bool, nonzero if we need to bind() to the new port */
758 #define SZ(X) (sizeof(X) / sizeof(*X))
761 #define FORCE_PREFIX "/IJB-FORCE-LOAD"
762 #endif /* def FORCE_LOAD */
765 #define HOME_PAGE_URL "http://ijbswa.sourceforge.net"
766 #define REDIRECT_URL HOME_PAGE_URL "/redirect.php?v=" VERSION "&to="
767 #define CGI_PREFIX_HOST "i.j.b"
770 static const char CSUCCEED[] =
771 "HTTP/1.0 200 Connection established\n"
772 "Proxy-Agent: IJ/" VERSION "\n\n";
774 static const char CHEADER[] =
775 "HTTP/1.0 400 Invalid header received from browser\n\n";
781 #endif /* ndef _PROJECT_H */