3 #define PROJECT_H_VERSION "$Id: project.h,v 1.21 2001/07/13 14:03:19 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 1.21 2001/07/13 14:03:19 oes
40 * - Reorganized regex header inclusion and #defines to
41 * comply to the scheme in configure.in
42 * - Added csp->content_type and its CT_* keys
43 * - Added ACTION_DEANIMATE
44 * - Removed all #ifdef PCRS
46 * Revision 1.20 2001/06/29 21:45:41 oes
47 * Indentation, CRLF->LF, Tab-> Space
49 * Revision 1.19 2001/06/29 13:33:36 oes
51 * - Introduced http_request.host_ip_addr_str
52 * - Introduced http_response.head_length
53 * - Introduced config.my_ip_addr_str, config.my_hostname,
54 * config.admin_address and config.proxy_info_url
55 * - Removed config.proxy_args_header and config.proxy_args_trailer,
56 * renamed config.proxy_args_invocation to config.proxy_args
57 * - Removed HTML snipplets and GIFs
58 * - Removed logentry from cancelled commit
60 * Revision 1.18 2001/06/09 10:57:39 jongfoster
61 * Adding definition of BUFFER_SIZE.
62 * Changing struct cgi_dispatcher to use "const" strings.
64 * Revision 1.17 2001/06/07 23:15:09 jongfoster
65 * Merging ACL and forward files into config file.
66 * Moving struct gateway members into struct forward_spec
67 * Removing config->proxy_args_gateways
68 * Cosmetic: Adding a few comments
70 * Revision 1.16 2001/06/04 18:31:58 swa
71 * files are now prefixed with either `confdir' or `logdir'.
72 * `make redhat-dist' replaces both entries confdir and logdir
75 * Revision 1.15 2001/06/04 11:28:53 swa
76 * redirect did not work due to missing /
78 * Revision 1.14 2001/06/03 11:03:48 oes
80 * added struct http_response,
81 * changed struct interceptors to struct cgi_dispatcher,
82 * moved HTML stuff to cgi.h
84 * Revision 1.13 2001/06/01 20:05:36 jongfoster
85 * Support for +image-blocker{}: added ACTION_IMAGE_BLOCKER
86 * constant, and removed csp->tinygif.
88 * Revision 1.12 2001/06/01 18:49:17 jongfoster
89 * Replaced "list_share" with "list" - the tiny memory gain was not
90 * worth the extra complexity.
92 * Revision 1.11 2001/06/01 10:32:47 oes
93 * Added constants for anchoring selection bitmap
95 * Revision 1.10 2001/05/31 21:33:53 jongfoster
96 * Changes for new actions file, replacing permissionsfile
97 * and parts of the config file. Also added support for
100 * Revision 1.9 2001/05/31 17:32:31 oes
102 * - Enhanced domain part globbing with infix and prefix asterisk
103 * matching and optional unanchored operation
105 * Revision 1.8 2001/05/29 20:09:15 joergs
106 * HTTP_REDIRECT_TEMPLATE fixed.
108 * Revision 1.7 2001/05/29 09:50:24 jongfoster
109 * Unified blocklist/imagelist/actionslist.
110 * File format is still under discussion, but the internal changes
113 * Also modified interceptor behaviour:
114 * - We now intercept all URLs beginning with one of the following
115 * prefixes (and *only* these prefixes):
117 * * http://ijbswa.sf.net/config/
118 * * http://ijbswa.sourceforge.net/config/
119 * - New interceptors "home page" - go to http://i.j.b/ to see it.
120 * - Internal changes so that intercepted and fast redirect pages
121 * are not replaced with an image.
122 * - Interceptors now have the option to send a binary page direct
123 * to the client. (i.e. ijb-send-banner uses this)
124 * - Implemented show-url-info interceptor. (Which is why I needed
125 * the above interceptors changes - a typical URL is
126 * "http://i.j.b/show-url-info?url=www.somesite.com/banner.gif".
127 * The previous mechanism would not have intercepted that, and
128 * if it had been intercepted then it then it would have replaced
131 * Revision 1.6 2001/05/27 22:17:04 oes
133 * - re_process_buffer no longer writes the modified buffer
134 * to the client, which was very ugly. It now returns the
135 * buffer, which it is then written by chat.
137 * - content_length now adjusts the Content-Length: header
138 * for modified documents rather than crunch()ing it.
139 * (Length info in csp->content_length, which is 0 for
140 * unmodified documents)
142 * - For this to work, sed() is called twice when filtering.
144 * Revision 1.5 2001/05/26 00:28:36 jongfoster
145 * Automatic reloading of config file.
146 * Removed obsolete SIGHUP support (Unix) and Reload menu option (Win32).
147 * Most of the global variables have been moved to a new
148 * struct configuration_spec, accessed through csp->config->globalname
149 * Most of the globals remaining are used by the Win32 GUI.
151 * Revision 1.4 2001/05/22 18:46:04 oes
153 * - Enabled filtering banners by size rather than URL
154 * by adding patterns that replace all standard banner
155 * sizes with the "Junkbuster" gif to the re_filterfile
157 * - Enabled filtering WebBugs by providing a pattern
158 * which kills all 1x1 images
160 * - Added support for PCRE_UNGREEDY behaviour to pcrs,
161 * which is selected by the (nonstandard and therefore
162 * capital) letter 'U' in the option string.
163 * It causes the quantifiers to be ungreedy by default.
164 * Appending a ? turns back to greedy (!).
166 * - Added a new interceptor ijb-send-banner, which
167 * sends back the "Junkbuster" gif. Without imagelist or
168 * MSIE detection support, or if tinygif = 1, or the
169 * URL isn't recognized as an imageurl, a lame HTML
170 * explanation is sent instead.
172 * - Added new feature, which permits blocking remote
173 * script redirects and firing back a local redirect
175 * The feature is conditionally compiled, i.e. it
176 * can be disabled with --disable-fast-redirects,
177 * plus it must be activated by a "fast-redirects"
178 * line in the config file, has its own log level
179 * and of course wants to be displayed by show-proxy-args
180 * Note: Boy, all the #ifdefs in 1001 locations and
181 * all the fumbling with configure.in and acconfig.h
182 * were *way* more work than the feature itself :-(
184 * - Because a generic redirect template was needed for
185 * this, tinygif = 3 now uses the same.
187 * - Moved GIFs, and other static HTTP response templates
192 * - Removed some >400 CRs again (Jon, you really worked
195 * Revision 1.3 2001/05/20 01:21:20 jongfoster
196 * Version 2.9.4 checkin.
197 * - Merged popupfile and cookiefile, and added control over PCRS
198 * filtering, in new "actionsfile".
199 * - Implemented LOG_LEVEL_FATAL, so that if there is a configuration
200 * file error you now get a message box (in the Win32 GUI) rather
201 * than the program exiting with no explanation.
202 * - Made killpopup use the PCRS MIME-type checking and HTTP-header
204 * - Removed tabs from "config"
205 * - Moved duplicated url parsing code in "loaders.c" to a new funcition.
206 * - Bumped up version number.
208 * Revision 1.2 2001/05/17 23:01:01 oes
209 * - Cleaned CRLF's from the sources and related files
211 * Revision 1.1.1.1 2001/05/15 13:59:03 oes
212 * Initial import of version 2.9.3 source tree
215 *********************************************************************/
218 /* Declare struct FILE for vars and funcs. */
221 /* Need time_t for file_list */
225 * Include appropriate regular expression libraries.
226 * Note that pcrs and pcre (native) are needed for cgi
227 * and are included anyway.
230 #if defined(REGEX_PCRE) || defined (REGEX_GNU)
232 #endif /* defined(REGEX_PCRE) || defined (REGEX_GNU) */
246 #if defined(REGEX_PCRE)
248 # include "pcreposix.h"
250 # include <pcreposix.h>
252 #endif /* defined(REGEX_PCRE) */
254 #if defined(REGEX_GNU)
255 # include "gnu_regex.h"
260 #endif /* def AMIGA */
266 #define BUFFER_SIZE 5000
270 /* Default IP and port to listen on */
271 #define HADDR_DEFAULT "127.0.0.1"
272 #define HADDR_PORT 8000
275 /* Need this for struct client_state */
276 struct configuration_spec;
278 /* Generic linked list of strings */
279 struct list /* FIXME: Why not separate entries and header? */
281 char * str; /* valid in an entry */
282 struct list *last; /* valid in header */
298 char *host_ip_addr_str; /* NULL before connect_to() */
302 char *hostport; /* "host[:port]" */
306 /* Response generated by CGI, blocker, or error handler */
309 char *status; /* HTTP status (string)*/
310 struct list headers[1]; /* List of header lines */
311 char *head; /* Formatted http response head */
312 int head_length; /* Length of http response head */
313 char *body; /* HTTP document body */
314 int content_length; /* Length of body, REQUIRED if binary body*/
320 char *spec; /* The string which was parsed to produce this */
321 /* url_spec. Used for debugging or display only. */
323 /* Hostname matching: */
324 char *domain; /* Fully qalified domain name (FQDN) pattern. */
325 /* May contain "*". */
326 char *dbuf; /* Buffer with '\0'-delimited fqdn */
327 char **dvec; /* Domain ptr vector into dbuf */
328 int dcnt; /* How many domains in fqdn? */
329 int unanchored; /* Bitmap - flags are ANCHOR_LEFT and ANCHOR_RIGHT */
332 int port; /* The port number, or 0 to match all ports. */
335 char *path; /* The path prefix (if not using regex), or source */
337 int pathlen; /* ==strlen(path). Needed for prefix matching. */
339 regex_t *preg; /* Regex for matching path part */
343 #define ANCHOR_LEFT 1
344 #define ANCHOR_RIGHT 2
357 #define IOB_PEEK(CSP) ((CSP->iob->cur > CSP->iob->eod) ? (CSP->iob->eod - CSP->iob->cur) : 0)
358 #define IOB_RESET(CSP) if(CSP->iob->buf) free(CSP->iob->buf); memset(CSP->iob, '\0', sizeof(CSP->iob));
360 /* Keys for csp->content_type */
361 #define CT_TEXT 0x01U
364 #define ACTION_MASK_ALL (~0U)
366 #define ACTION_MOST_COMPATIBLE 0x0000U
368 #define ACTION_BLOCK 0x0001U
369 #define ACTION_DEANIMATE 0x2000U
370 #define ACTION_FAST_REDIRECTS 0x0002U
371 #define ACTION_FILTER 0x0004U
372 #define ACTION_HIDE_FORWARDED 0x0008U
373 #define ACTION_HIDE_FROM 0x0010U
374 #define ACTION_HIDE_REFERER 0x0020U /* sic - follow HTTP, not English */
375 #define ACTION_HIDE_USER_AGENT 0x0040U
376 #define ACTION_IMAGE 0x0080U
377 #define ACTION_IMAGE_BLOCKER 0x0100U
378 #define ACTION_NO_COOKIE_READ 0x0200U
379 #define ACTION_NO_COOKIE_SET 0x0400U
380 #define ACTION_NO_POPUPS 0x0800U
381 #define ACTION_VANILLA_WAFER 0x1000U
383 #define ACTION_STRING_FROM 0
384 #define ACTION_STRING_IMAGE_BLOCKER 1
385 #define ACTION_STRING_REFERER 2
386 #define ACTION_STRING_USER_AGENT 3
387 #define ACTION_STRING_COUNT 4
389 #define ACTION_MULTI_ADD_HEADER 0
390 #define ACTION_MULTI_WAFER 1
391 #define ACTION_MULTI_COUNT 2
394 * This structure contains a list of actions to apply to a URL.
395 * It only contains positive instructions - no "-" options.
396 * It is not used to store the actions list itself, only for
397 * url_actions() to return the current values.
399 struct current_action_spec
401 unsigned flags; /* a bit set to "1" = add action */
403 /* For those actions that require parameters: */
405 /* each entry is valid if & only if corresponding entry in "add" set. */
406 char * string[ACTION_STRING_COUNT];
409 struct list multi[ACTION_MULTI_COUNT][1];
414 * This structure contains a set of changes to actions.
415 * It can contain both positive and negative instructions.
416 * It is used to store an entry in the actions list.
420 unsigned mask; /* a bit set to "0" = remove action */
421 unsigned add; /* a bit set to "1" = add action */
423 /* For those actions that require parameters: */
425 /* each entry is valid if & only if corresponding entry in "add" set. */
426 char * string[ACTION_STRING_COUNT];
428 /* Strings to remove. */
429 struct list multi_remove[ACTION_MULTI_COUNT][1];
431 /* If nonzero, remove *all* strings. */
432 int multi_remove_all[ACTION_MULTI_COUNT];
435 struct list multi_add[ACTION_MULTI_COUNT][1];
439 * This structure is used to store the actions list.
441 * It contains a URL pattern, and the chages to the actions.
442 * It is a linked list.
446 struct url_spec url[1];
448 struct action_spec action[1];
450 struct url_actions * next;
454 /* Constants defining bitmask for csp->accept_types */
456 #ifdef DETECT_MSIE_IMAGES
458 /* MSIE detected by user-agent string */
459 #define ACCEPT_TYPE_IS_MSIE 0x0001
462 * *If* this is MSIE, it wants an image. (Or this is a shift-reload, or
463 * it's got an image from this URL before... yuck!)
464 * Only meaningful if ACCEPT_TYPE_IS_MSIE set
466 #define ACCEPT_TYPE_MSIE_IMAGE 0x0002
469 * *If* this is MSIE, it wants a HTML document.
470 * Only meaningful if ACCEPT_TYPE_IS_MSIE set
472 #define ACCEPT_TYPE_MSIE_HTML 0x0004
474 #endif /* def DETECT_MSIE_IMAGES */
479 /* The proxy's configuration */
480 struct configuration_spec * config;
482 /* The actions to perform on the current request */
483 struct current_action_spec action[1];
485 /* socket to talk to client (web browser) */
488 /* socket to talk to server (web server or proxy) */
493 /* 1 if this URL was rejected, 0 otherwise. Allows actual stats inc to
494 * occur in main thread only for thread-safety.
497 #endif /* def STATISTICS */
501 #endif /* def FORCE_LOAD */
505 #endif /* def TOGGLE */
508 * Client PC's IP address, as reported by the accept()_ function.
509 * Both as string and number
515 /* Our IP address and hostname, i.e. the IP address that
516 the client used to reach us, and the associated hostname,
519 char *my_ip_addr_str;
523 /* The referer in this request, if one was specified. */
525 #endif /* def TRUST_FILES */
527 #if defined(DETECT_MSIE_IMAGES)
528 /* Types the client will accept.
529 * Bitmask - see ACCEPT_TYPE_XXX constants.
532 #endif /* defined(DETECT_MSIE_IMAGES) */
534 /* The URL that was requested */
535 struct http_request http[1];
537 /* An I/O buffer used for buffering data read from the client */
540 /* List of all headers for this request */
541 struct list headers[1];
543 /* List of all cookies for this request */
544 struct list cookie_list[1];
546 /* MIME-Type bitmap, see CT_* above */
547 unsigned char content_type;
549 /* The "X-Forwarded-For:" header sent by the client */
553 * Nonzero if this client is processing data.
554 * Set to zero when the thread associated with this structure dies.
558 /* files associated with this client */
559 struct file_list *actions_list;
561 struct file_list *rlist; /* pcrs job file */
562 size_t content_length; /* Length after content modification */
565 struct file_list *tlist; /* trustfile */
566 #endif /* def TRUST_FILES */
568 struct client_state *next;
576 char *(*parser)(const struct parsers *, char *, struct client_state *);
579 struct cgi_dispatcher
583 int (*handler)(struct client_state *csp, struct http_response *rsp, struct map *parameters);
584 const char *description;
590 * this is a pointer to the data structures associated with the file.
591 * Read-only once the structure has been created.
595 /* Normally NULL. When we are finished with file (i.e. when we have
596 * loaded a new one), set to a pointer to an unloader function.
597 * Unloader will be called by sweep() (called from main loop) when
598 * all clients using this file are done. This prevents threading
601 void (*unloader)(void *);
603 /* Used internally by sweep(). Do not access from elsewhere. */
606 #ifndef SPLIT_PROXY_ARGS
607 /* String to be displayed as part of show-proxy-args display.
608 * Read-only once the structure has been created.
611 #endif /* ndef SPLIT_PROXY_ARGS */
613 /* Following variables allow us to check if file has been changed.
614 * Read-only once the structure has been created.
619 /* Pointer to next entry in the linked list of all "file_list"s.
620 * This linked list is so that sweep() can navigate it.
621 * Since sweep() can remove items from the list, we must be careful
622 * to only access this value from main thread (when we know sweep
625 struct file_list *next;
632 struct url_spec url[1];
634 struct block_spec *next;
636 #endif /* def TRUST_FILES */
639 #define SOCKS_NONE 0 /* Don't use a SOCKS server */
640 #define SOCKS_4 40 /* original SOCKS 4 protocol */
641 #define SOCKS_4A 41 /* as modified for hosts w/o external DNS */
645 struct url_spec url[1];
647 /* Connection type - must be a SOCKS_xxx constant */
654 /* Parent HTTP proxy */
658 /* For the linked list */
659 struct forward_spec *next;
662 struct re_filterfile_spec
664 struct list patterns[1];
669 #define ACL_PERMIT 1 /* accept connection request */
670 #define ACL_DENY 2 /* reject connection request */
672 struct access_control_addr
679 struct access_control_list
681 struct access_control_addr src[1];
682 struct access_control_addr dst[1];
685 struct access_control_list *next;
687 #endif /* def ACL_FILES */
690 /* Maximum number of loaders (actions, re_filter, ...) */
694 * Data loaded from the configuration file.
696 * (Anomaly: toggle is still handled through a global, not this structure)
698 struct configuration_spec
707 const char *actions_file;
709 /* The administrator's email address */
712 /* A URL with info on this proxy */
713 char *proxy_info_url;
715 const char *re_filterfile;
718 const char * jarfile;
720 #endif /* def JAR_FILES */
723 * Port and IP to bind to.
724 * Defaults to HADDR_DEFAULT:HADDR_PORT == 127.0.0.1:8000
729 #ifndef SPLIT_PROXY_ARGS
730 const char *suppress_message;
731 #endif /* ndef SPLIT_PROXY_ARGS */
733 #ifndef SPLIT_PROXY_ARGS
734 /* suppress listing config files */
735 int suppress_blocklists;
736 #endif /* ndef SPLIT_PROXY_ARGS */
739 const char * trustfile;
741 struct list trust_info[1];
742 struct url_spec *trust_list[64];
743 #endif /* def TRUST_FILES */
746 struct access_control_list *acl;
747 #endif /* def ACL_FILES */
749 struct forward_spec *forward;
751 /* All options from the config file, HTML-formatted */
754 /* the configuration file object. */
755 struct file_list *config_file_list;
757 /* List of loaders */
758 int (*loaders[NLOADERS])(struct client_state *);
760 /* bool, nonzero if we need to bind() to the new port */
765 #define SZ(X) (sizeof(X) / sizeof(*X))
768 #define FORCE_PREFIX "/IJB-FORCE-LOAD"
769 #endif /* def FORCE_LOAD */
772 #define HOME_PAGE_URL "http://ijbswa.sourceforge.net"
773 #define REDIRECT_URL HOME_PAGE_URL "/redirect.php?v=" VERSION "&to="
774 #define CGI_PREFIX_HOST "i.j.b"
777 static const char CSUCCEED[] =
778 "HTTP/1.0 200 Connection established\n"
779 "Proxy-Agent: IJ/" VERSION "\n\n";
781 static const char CHEADER[] =
782 "HTTP/1.0 400 Invalid header received from browser\n\n";
788 #endif /* ndef _PROJECT_H */