3 #define PROJECT_H_VERSION "$Id: project.h,v 1.19 2001/06/29 13:33:36 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 1.19 2001/06/29 13:33:36 oes
41 * - Introduced http_request.host_ip_addr_str
42 * - Introduced http_response.head_length
43 * - Introduced config.my_ip_addr_str, config.my_hostname,
44 * config.admin_address and config.proxy_info_url
45 * - Removed config.proxy_args_header and config.proxy_args_trailer,
46 * renamed config.proxy_args_invocation to config.proxy_args
47 * - Removed HTML snipplets and GIFs
48 * - Removed logentry from cancelled commit
50 * Revision 1.18 2001/06/09 10:57:39 jongfoster
51 * Adding definition of BUFFER_SIZE.
52 * Changing struct cgi_dispatcher to use "const" strings.
54 * Revision 1.17 2001/06/07 23:15:09 jongfoster
55 * Merging ACL and forward files into config file.
56 * Moving struct gateway members into struct forward_spec
57 * Removing config->proxy_args_gateways
58 * Cosmetic: Adding a few comments
60 * Revision 1.16 2001/06/04 18:31:58 swa
61 * files are now prefixed with either `confdir' or `logdir'.
62 * `make redhat-dist' replaces both entries confdir and logdir
65 * Revision 1.15 2001/06/04 11:28:53 swa
66 * redirect did not work due to missing /
68 * Revision 1.14 2001/06/03 11:03:48 oes
70 * added struct http_response,
71 * changed struct interceptors to struct cgi_dispatcher,
72 * moved HTML stuff to cgi.h
74 * Revision 1.13 2001/06/01 20:05:36 jongfoster
75 * Support for +image-blocker{}: added ACTION_IMAGE_BLOCKER
76 * constant, and removed csp->tinygif.
78 * Revision 1.12 2001/06/01 18:49:17 jongfoster
79 * Replaced "list_share" with "list" - the tiny memory gain was not
80 * worth the extra complexity.
82 * Revision 1.11 2001/06/01 10:32:47 oes
83 * Added constants for anchoring selection bitmap
85 * Revision 1.10 2001/05/31 21:33:53 jongfoster
86 * Changes for new actions file, replacing permissionsfile
87 * and parts of the config file. Also added support for
90 * Revision 1.9 2001/05/31 17:32:31 oes
92 * - Enhanced domain part globbing with infix and prefix asterisk
93 * matching and optional unanchored operation
95 * Revision 1.8 2001/05/29 20:09:15 joergs
96 * HTTP_REDIRECT_TEMPLATE fixed.
98 * Revision 1.7 2001/05/29 09:50:24 jongfoster
99 * Unified blocklist/imagelist/actionslist.
100 * File format is still under discussion, but the internal changes
103 * Also modified interceptor behaviour:
104 * - We now intercept all URLs beginning with one of the following
105 * prefixes (and *only* these prefixes):
107 * * http://ijbswa.sf.net/config/
108 * * http://ijbswa.sourceforge.net/config/
109 * - New interceptors "home page" - go to http://i.j.b/ to see it.
110 * - Internal changes so that intercepted and fast redirect pages
111 * are not replaced with an image.
112 * - Interceptors now have the option to send a binary page direct
113 * to the client. (i.e. ijb-send-banner uses this)
114 * - Implemented show-url-info interceptor. (Which is why I needed
115 * the above interceptors changes - a typical URL is
116 * "http://i.j.b/show-url-info?url=www.somesite.com/banner.gif".
117 * The previous mechanism would not have intercepted that, and
118 * if it had been intercepted then it then it would have replaced
121 * Revision 1.6 2001/05/27 22:17:04 oes
123 * - re_process_buffer no longer writes the modified buffer
124 * to the client, which was very ugly. It now returns the
125 * buffer, which it is then written by chat.
127 * - content_length now adjusts the Content-Length: header
128 * for modified documents rather than crunch()ing it.
129 * (Length info in csp->content_length, which is 0 for
130 * unmodified documents)
132 * - For this to work, sed() is called twice when filtering.
134 * Revision 1.5 2001/05/26 00:28:36 jongfoster
135 * Automatic reloading of config file.
136 * Removed obsolete SIGHUP support (Unix) and Reload menu option (Win32).
137 * Most of the global variables have been moved to a new
138 * struct configuration_spec, accessed through csp->config->globalname
139 * Most of the globals remaining are used by the Win32 GUI.
141 * Revision 1.4 2001/05/22 18:46:04 oes
143 * - Enabled filtering banners by size rather than URL
144 * by adding patterns that replace all standard banner
145 * sizes with the "Junkbuster" gif to the re_filterfile
147 * - Enabled filtering WebBugs by providing a pattern
148 * which kills all 1x1 images
150 * - Added support for PCRE_UNGREEDY behaviour to pcrs,
151 * which is selected by the (nonstandard and therefore
152 * capital) letter 'U' in the option string.
153 * It causes the quantifiers to be ungreedy by default.
154 * Appending a ? turns back to greedy (!).
156 * - Added a new interceptor ijb-send-banner, which
157 * sends back the "Junkbuster" gif. Without imagelist or
158 * MSIE detection support, or if tinygif = 1, or the
159 * URL isn't recognized as an imageurl, a lame HTML
160 * explanation is sent instead.
162 * - Added new feature, which permits blocking remote
163 * script redirects and firing back a local redirect
165 * The feature is conditionally compiled, i.e. it
166 * can be disabled with --disable-fast-redirects,
167 * plus it must be activated by a "fast-redirects"
168 * line in the config file, has its own log level
169 * and of course wants to be displayed by show-proxy-args
170 * Note: Boy, all the #ifdefs in 1001 locations and
171 * all the fumbling with configure.in and acconfig.h
172 * were *way* more work than the feature itself :-(
174 * - Because a generic redirect template was needed for
175 * this, tinygif = 3 now uses the same.
177 * - Moved GIFs, and other static HTTP response templates
182 * - Removed some >400 CRs again (Jon, you really worked
185 * Revision 1.3 2001/05/20 01:21:20 jongfoster
186 * Version 2.9.4 checkin.
187 * - Merged popupfile and cookiefile, and added control over PCRS
188 * filtering, in new "actionsfile".
189 * - Implemented LOG_LEVEL_FATAL, so that if there is a configuration
190 * file error you now get a message box (in the Win32 GUI) rather
191 * than the program exiting with no explanation.
192 * - Made killpopup use the PCRS MIME-type checking and HTTP-header
194 * - Removed tabs from "config"
195 * - Moved duplicated url parsing code in "loaders.c" to a new funcition.
196 * - Bumped up version number.
198 * Revision 1.2 2001/05/17 23:01:01 oes
199 * - Cleaned CRLF's from the sources and related files
201 * Revision 1.1.1.1 2001/05/15 13:59:03 oes
202 * Initial import of version 2.9.3 source tree
205 *********************************************************************/
208 /* Declare struct FILE for vars and funcs. */
211 /* Need time_t for file_list */
215 * Include appropriate regular expression libraries.
217 * PCRS ==> Include pcre
218 * REGEX && PCRE ==> Include pcre and pcreposix
219 * REGEX && !PCRE ==> Include gnu_regex
221 * STATIC ==> Use #include "pcre.h" (compiling at same time)
222 * !STATIC ==> Use #include <pcre.h> (System library)
225 #if (defined(REGEX) && defined(PCRE)) || defined(PCRS)
231 #endif /* (defined(REGEX) && defined(PCRE)) || defined(PCRS) */
233 #if defined(REGEX) && defined(PCRE)
235 # include "pcreposix.h"
237 # include <pcreposix.h>
239 #endif /* defined(REGEX) && defined(PCRE) */
241 #if defined(REGEX) && !defined(PCRE)
242 # include "gnu_regex.h"
247 #endif /* def PCRS */
251 #endif /* def AMIGA */
257 #define BUFFER_SIZE 5000
261 /* Default IP and port to listen on */
262 #define HADDR_DEFAULT "127.0.0.1"
263 #define HADDR_PORT 8000
266 /* Need this for struct client_state */
267 struct configuration_spec;
269 /* Generic linked list of strings */
270 struct list /* FIXME: Why not separate entries and header? */
272 char * str; /* valid in an entry */
273 struct list *last; /* valid in header */
289 char *host_ip_addr_str; /* NULL before connect_to() */
293 char *hostport; /* "host[:port]" */
297 /* Response generated by CGI, blocker, or error handler */
300 char *status; /* HTTP status (string)*/
301 struct list headers[1]; /* List of header lines */
302 char *head; /* Formatted http response head */
303 int head_length; /* Length of http response head */
304 char *body; /* HTTP document body */
305 int content_length; /* Length of body, REQUIRED if binary body*/
311 char *spec; /* The string which was parsed to produce this */
312 /* url_spec. Used for debugging or display only. */
314 /* Hostname matching: */
315 char *domain; /* Fully qalified domain name (FQDN) pattern. */
316 /* May contain "*". */
317 char *dbuf; /* Buffer with '\0'-delimited fqdn */
318 char **dvec; /* Domain ptr vector into dbuf */
319 int dcnt; /* How many domains in fqdn? */
320 int unanchored; /* Bitmap - flags are ANCHOR_LEFT and ANCHOR_RIGHT */
323 int port; /* The port number, or 0 to match all ports. */
326 char *path; /* The path prefix (if not using regex), or source */
328 int pathlen; /* ==strlen(path). Needed for prefix matching. */
330 regex_t *preg; /* Regex for matching path part */
334 #define ANCHOR_LEFT 1
335 #define ANCHOR_RIGHT 2
348 #define IOB_PEEK(CSP) ((CSP->iob->cur > CSP->iob->eod) ? (CSP->iob->eod - CSP->iob->cur) : 0)
349 #define IOB_RESET(CSP) if(CSP->iob->buf) free(CSP->iob->buf); memset(CSP->iob, '\0', sizeof(CSP->iob));
353 #define ACTION_MASK_ALL (~0U)
355 #define ACTION_MOST_COMPATIBLE 0x0000U
357 #define ACTION_BLOCK 0x0001U
358 #define ACTION_FAST_REDIRECTS 0x0002U
359 #define ACTION_FILTER 0x0004U
360 #define ACTION_HIDE_FORWARDED 0x0008U
361 #define ACTION_HIDE_FROM 0x0010U
362 #define ACTION_HIDE_REFERER 0x0020U /* sic - follow HTTP, not English */
363 #define ACTION_HIDE_USER_AGENT 0x0040U
364 #define ACTION_IMAGE 0x0080U
365 #define ACTION_IMAGE_BLOCKER 0x0100U
366 #define ACTION_NO_COOKIE_READ 0x0200U
367 #define ACTION_NO_COOKIE_SET 0x0400U
368 #define ACTION_NO_POPUPS 0x0800U
369 #define ACTION_VANILLA_WAFER 0x1000U
371 #define ACTION_STRING_FROM 0
372 #define ACTION_STRING_IMAGE_BLOCKER 1
373 #define ACTION_STRING_REFERER 2
374 #define ACTION_STRING_USER_AGENT 3
375 #define ACTION_STRING_COUNT 4
377 #define ACTION_MULTI_ADD_HEADER 0
378 #define ACTION_MULTI_WAFER 1
379 #define ACTION_MULTI_COUNT 2
382 * This structure contains a list of actions to apply to a URL.
383 * It only contains positive instructions - no "-" options.
384 * It is not used to store the actions list itself, only for
385 * url_actions() to return the current values.
387 struct current_action_spec
389 unsigned flags; /* a bit set to "1" = add action */
391 /* For those actions that require parameters: */
393 /* each entry is valid if & only if corresponding entry in "add" set. */
394 char * string[ACTION_STRING_COUNT];
397 struct list multi[ACTION_MULTI_COUNT][1];
402 * This structure contains a set of changes to actions.
403 * It can contain both positive and negative instructions.
404 * It is used to store an entry in the actions list.
408 unsigned mask; /* a bit set to "0" = remove action */
409 unsigned add; /* a bit set to "1" = add action */
411 /* For those actions that require parameters: */
413 /* each entry is valid if & only if corresponding entry in "add" set. */
414 char * string[ACTION_STRING_COUNT];
416 /* Strings to remove. */
417 struct list multi_remove[ACTION_MULTI_COUNT][1];
419 /* If nonzero, remove *all* strings. */
420 int multi_remove_all[ACTION_MULTI_COUNT];
423 struct list multi_add[ACTION_MULTI_COUNT][1];
427 * This structure is used to store the actions list.
429 * It contains a URL pattern, and the chages to the actions.
430 * It is a linked list.
434 struct url_spec url[1];
436 struct action_spec action[1];
438 struct url_actions * next;
442 /* Constants defining bitmask for csp->accept_types */
444 #ifdef DETECT_MSIE_IMAGES
446 /* MSIE detected by user-agent string */
447 #define ACCEPT_TYPE_IS_MSIE 0x0001
450 * *If* this is MSIE, it wants an image. (Or this is a shift-reload, or
451 * it's got an image from this URL before... yuck!)
452 * Only meaningful if ACCEPT_TYPE_IS_MSIE set
454 #define ACCEPT_TYPE_MSIE_IMAGE 0x0002
457 * *If* this is MSIE, it wants a HTML document.
458 * Only meaningful if ACCEPT_TYPE_IS_MSIE set
460 #define ACCEPT_TYPE_MSIE_HTML 0x0004
462 #endif /* def DETECT_MSIE_IMAGES */
467 /* The proxy's configuration */
468 struct configuration_spec * config;
470 /* The actions to perform on the current request */
471 struct current_action_spec action[1];
473 /* socket to talk to client (web browser) */
476 /* socket to talk to server (web server or proxy) */
481 /* 1 if this URL was rejected, 0 otherwise. Allows actual stats inc to
482 * occur in main thread only for thread-safety.
485 #endif /* def STATISTICS */
489 #endif /* def FORCE_LOAD */
493 #endif /* def TOGGLE */
496 * Client PC's IP address, as reported by the accept()_ function.
497 * Both as string and number
503 /* Our IP address and hostname, i.e. the IP address that
504 the client used to reach us, and the associated hostname,
507 char *my_ip_addr_str;
511 /* The referer in this request, if one was specified. */
513 #endif /* def TRUST_FILES */
515 #if defined(DETECT_MSIE_IMAGES)
516 /* Types the client will accept.
517 * Bitmask - see ACCEPT_TYPE_XXX constants.
520 #endif /* defined(DETECT_MSIE_IMAGES) */
522 /* The URL that was requested */
523 struct http_request http[1];
525 /* An I/O buffer used for buffering data read from the client */
528 /* List of all headers for this request */
529 struct list headers[1];
531 /* List of all cookies for this request */
532 struct list cookie_list[1];
534 #if defined(PCRS) || defined(KILLPOPUPS)
535 /* Nonzero if this has a text MIME type */
537 #endif /* defined(PCRS) || defined(KILLPOPUPS) */
539 /* The "X-Forwarded-For:" header sent by the client */
543 * Nonzero if this client is processing data.
544 * Set to zero when the thread associated with this structure dies.
548 /* files associated with this client */
549 struct file_list *actions_list;
552 struct file_list *rlist; /* Perl re_filterfile */
553 size_t content_length; /* Length after processing */
554 #endif /* def PCRS */
557 struct file_list *tlist; /* trustfile */
558 #endif /* def TRUST_FILES */
560 struct client_state *next;
568 char *(*parser)(const struct parsers *, char *, struct client_state *);
571 struct cgi_dispatcher
575 int (*handler)(struct client_state *csp, struct http_response *rsp, struct map *parameters);
576 const char *description;
582 * this is a pointer to the data structures associated with the file.
583 * Read-only once the structure has been created.
587 /* Normally NULL. When we are finished with file (i.e. when we have
588 * loaded a new one), set to a pointer to an unloader function.
589 * Unloader will be called by sweep() (called from main loop) when
590 * all clients using this file are done. This prevents threading
593 void (*unloader)(void *);
595 /* Used internally by sweep(). Do not access from elsewhere. */
598 #ifndef SPLIT_PROXY_ARGS
599 /* String to be displayed as part of show-proxy-args display.
600 * Read-only once the structure has been created.
603 #endif /* ndef SPLIT_PROXY_ARGS */
605 /* Following variables allow us to check if file has been changed.
606 * Read-only once the structure has been created.
611 /* Pointer to next entry in the linked list of all "file_list"s.
612 * This linked list is so that sweep() can navigate it.
613 * Since sweep() can remove items from the list, we must be careful
614 * to only access this value from main thread (when we know sweep
617 struct file_list *next;
624 struct url_spec url[1];
626 struct block_spec *next;
628 #endif /* def TRUST_FILES */
631 #define SOCKS_NONE 0 /* Don't use a SOCKS server */
632 #define SOCKS_4 40 /* original SOCKS 4 protocol */
633 #define SOCKS_4A 41 /* as modified for hosts w/o external DNS */
637 struct url_spec url[1];
639 /* Connection type - must be a SOCKS_xxx constant */
646 /* Parent HTTP proxy */
650 /* For the linked list */
651 struct forward_spec *next;
656 struct re_filterfile_spec
658 struct list patterns[1];
661 #endif /* def PCRS */
665 #define ACL_PERMIT 1 /* accept connection request */
666 #define ACL_DENY 2 /* reject connection request */
668 struct access_control_addr
675 struct access_control_list
677 struct access_control_addr src[1];
678 struct access_control_addr dst[1];
681 struct access_control_list *next;
683 #endif /* def ACL_FILES */
686 /* Maximum number of loaders (actions, re_filter, ...) */
690 * Data loaded from the configuration file.
692 * (Anomaly: toggle is still handled through a global, not this structure)
694 struct configuration_spec
703 const char *actions_file;
705 /* The administrator's email address */
708 /* A URL with info on this proxy */
709 char *proxy_info_url;
712 const char *re_filterfile;
713 #endif /* def PCRS */
716 const char * jarfile;
718 #endif /* def JAR_FILES */
721 * Port and IP to bind to.
722 * Defaults to HADDR_DEFAULT:HADDR_PORT == 127.0.0.1:8000
727 #ifndef SPLIT_PROXY_ARGS
728 const char *suppress_message;
729 #endif /* ndef SPLIT_PROXY_ARGS */
731 #ifndef SPLIT_PROXY_ARGS
732 /* suppress listing config files */
733 int suppress_blocklists;
734 #endif /* ndef SPLIT_PROXY_ARGS */
737 const char * trustfile;
739 struct list trust_info[1];
740 struct url_spec *trust_list[64];
741 #endif /* def TRUST_FILES */
744 struct access_control_list *acl;
745 #endif /* def ACL_FILES */
747 struct forward_spec *forward;
749 /* All options from the config file, HTML-formatted */
752 /* the configuration file object. */
753 struct file_list *config_file_list;
755 /* List of loaders */
756 int (*loaders[NLOADERS])(struct client_state *);
758 /* bool, nonzero if we need to bind() to the new port */
763 #define SZ(X) (sizeof(X) / sizeof(*X))
766 #define FORCE_PREFIX "/IJB-FORCE-LOAD"
767 #endif /* def FORCE_LOAD */
770 #define HOME_PAGE_URL "http://ijbswa.sourceforge.net"
771 #define REDIRECT_URL HOME_PAGE_URL "/redirect.php?v=" VERSION "&to="
772 #define CGI_PREFIX_HOST "i.j.b"
775 static const char CSUCCEED[] =
776 "HTTP/1.0 200 Connection established\n"
777 "Proxy-Agent: IJ/" VERSION "\n\n";
779 static const char CHEADER[] =
780 "HTTP/1.0 400 Invalid header received from browser\n\n";
786 #endif /* ndef _PROJECT_H */