1 #ifndef PROJECT_H_INCLUDED
2 #define PROJECT_H_INCLUDED
3 #define PROJECT_H_VERSION "$Id: project.h,v 1.25 2001/07/29 18:43:08 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 1.25 2001/07/29 18:43:08 jongfoster
40 * Changing #ifdef _FILENAME_H to FILENAME_H_INCLUDED, to conform to
43 * Revision 1.24 2001/07/25 17:20:27 oes
44 * Introduced http->user_agent
46 * Revision 1.23 2001/07/18 12:32:23 oes
47 * - Added ACTION_STRING_DEANIMATE
48 * - moved #define freez from jcc.h to project.h
50 * Revision 1.22 2001/07/15 17:51:41 jongfoster
51 * Renaming #define STATIC to STATIC_PCRE
53 * Revision 1.21 2001/07/13 14:03:19 oes
54 * - Reorganized regex header inclusion and #defines to
55 * comply to the scheme in configure.in
56 * - Added csp->content_type and its CT_* keys
57 * - Added ACTION_DEANIMATE
58 * - Removed all #ifdef PCRS
60 * Revision 1.20 2001/06/29 21:45:41 oes
61 * Indentation, CRLF->LF, Tab-> Space
63 * Revision 1.19 2001/06/29 13:33:36 oes
65 * - Introduced http_request.host_ip_addr_str
66 * - Introduced http_response.head_length
67 * - Introduced config.my_ip_addr_str, config.my_hostname,
68 * config.admin_address and config.proxy_info_url
69 * - Removed config.proxy_args_header and config.proxy_args_trailer,
70 * renamed config.proxy_args_invocation to config.proxy_args
71 * - Removed HTML snipplets and GIFs
72 * - Removed logentry from cancelled commit
74 * Revision 1.18 2001/06/09 10:57:39 jongfoster
75 * Adding definition of BUFFER_SIZE.
76 * Changing struct cgi_dispatcher to use "const" strings.
78 * Revision 1.17 2001/06/07 23:15:09 jongfoster
79 * Merging ACL and forward files into config file.
80 * Moving struct gateway members into struct forward_spec
81 * Removing config->proxy_args_gateways
82 * Cosmetic: Adding a few comments
84 * Revision 1.16 2001/06/04 18:31:58 swa
85 * files are now prefixed with either `confdir' or `logdir'.
86 * `make redhat-dist' replaces both entries confdir and logdir
89 * Revision 1.15 2001/06/04 11:28:53 swa
90 * redirect did not work due to missing /
92 * Revision 1.14 2001/06/03 11:03:48 oes
94 * added struct http_response,
95 * changed struct interceptors to struct cgi_dispatcher,
96 * moved HTML stuff to cgi.h
98 * Revision 1.13 2001/06/01 20:05:36 jongfoster
99 * Support for +image-blocker{}: added ACTION_IMAGE_BLOCKER
100 * constant, and removed csp->tinygif.
102 * Revision 1.12 2001/06/01 18:49:17 jongfoster
103 * Replaced "list_share" with "list" - the tiny memory gain was not
104 * worth the extra complexity.
106 * Revision 1.11 2001/06/01 10:32:47 oes
107 * Added constants for anchoring selection bitmap
109 * Revision 1.10 2001/05/31 21:33:53 jongfoster
110 * Changes for new actions file, replacing permissionsfile
111 * and parts of the config file. Also added support for
114 * Revision 1.9 2001/05/31 17:32:31 oes
116 * - Enhanced domain part globbing with infix and prefix asterisk
117 * matching and optional unanchored operation
119 * Revision 1.8 2001/05/29 20:09:15 joergs
120 * HTTP_REDIRECT_TEMPLATE fixed.
122 * Revision 1.7 2001/05/29 09:50:24 jongfoster
123 * Unified blocklist/imagelist/actionslist.
124 * File format is still under discussion, but the internal changes
127 * Also modified interceptor behaviour:
128 * - We now intercept all URLs beginning with one of the following
129 * prefixes (and *only* these prefixes):
131 * * http://ijbswa.sf.net/config/
132 * * http://ijbswa.sourceforge.net/config/
133 * - New interceptors "home page" - go to http://i.j.b/ to see it.
134 * - Internal changes so that intercepted and fast redirect pages
135 * are not replaced with an image.
136 * - Interceptors now have the option to send a binary page direct
137 * to the client. (i.e. ijb-send-banner uses this)
138 * - Implemented show-url-info interceptor. (Which is why I needed
139 * the above interceptors changes - a typical URL is
140 * "http://i.j.b/show-url-info?url=www.somesite.com/banner.gif".
141 * The previous mechanism would not have intercepted that, and
142 * if it had been intercepted then it then it would have replaced
145 * Revision 1.6 2001/05/27 22:17:04 oes
147 * - re_process_buffer no longer writes the modified buffer
148 * to the client, which was very ugly. It now returns the
149 * buffer, which it is then written by chat.
151 * - content_length now adjusts the Content-Length: header
152 * for modified documents rather than crunch()ing it.
153 * (Length info in csp->content_length, which is 0 for
154 * unmodified documents)
156 * - For this to work, sed() is called twice when filtering.
158 * Revision 1.5 2001/05/26 00:28:36 jongfoster
159 * Automatic reloading of config file.
160 * Removed obsolete SIGHUP support (Unix) and Reload menu option (Win32).
161 * Most of the global variables have been moved to a new
162 * struct configuration_spec, accessed through csp->config->globalname
163 * Most of the globals remaining are used by the Win32 GUI.
165 * Revision 1.4 2001/05/22 18:46:04 oes
167 * - Enabled filtering banners by size rather than URL
168 * by adding patterns that replace all standard banner
169 * sizes with the "Junkbuster" gif to the re_filterfile
171 * - Enabled filtering WebBugs by providing a pattern
172 * which kills all 1x1 images
174 * - Added support for PCRE_UNGREEDY behaviour to pcrs,
175 * which is selected by the (nonstandard and therefore
176 * capital) letter 'U' in the option string.
177 * It causes the quantifiers to be ungreedy by default.
178 * Appending a ? turns back to greedy (!).
180 * - Added a new interceptor ijb-send-banner, which
181 * sends back the "Junkbuster" gif. Without imagelist or
182 * MSIE detection support, or if tinygif = 1, or the
183 * URL isn't recognized as an imageurl, a lame HTML
184 * explanation is sent instead.
186 * - Added new feature, which permits blocking remote
187 * script redirects and firing back a local redirect
189 * The feature is conditionally compiled, i.e. it
190 * can be disabled with --disable-fast-redirects,
191 * plus it must be activated by a "fast-redirects"
192 * line in the config file, has its own log level
193 * and of course wants to be displayed by show-proxy-args
194 * Note: Boy, all the #ifdefs in 1001 locations and
195 * all the fumbling with configure.in and acconfig.h
196 * were *way* more work than the feature itself :-(
198 * - Because a generic redirect template was needed for
199 * this, tinygif = 3 now uses the same.
201 * - Moved GIFs, and other static HTTP response templates
206 * - Removed some >400 CRs again (Jon, you really worked
209 * Revision 1.3 2001/05/20 01:21:20 jongfoster
210 * Version 2.9.4 checkin.
211 * - Merged popupfile and cookiefile, and added control over PCRS
212 * filtering, in new "actionsfile".
213 * - Implemented LOG_LEVEL_FATAL, so that if there is a configuration
214 * file error you now get a message box (in the Win32 GUI) rather
215 * than the program exiting with no explanation.
216 * - Made killpopup use the PCRS MIME-type checking and HTTP-header
218 * - Removed tabs from "config"
219 * - Moved duplicated url parsing code in "loaders.c" to a new funcition.
220 * - Bumped up version number.
222 * Revision 1.2 2001/05/17 23:01:01 oes
223 * - Cleaned CRLF's from the sources and related files
225 * Revision 1.1.1.1 2001/05/15 13:59:03 oes
226 * Initial import of version 2.9.3 source tree
229 *********************************************************************/
232 /* Declare struct FILE for vars and funcs. */
235 /* Need time_t for file_list */
239 * Include appropriate regular expression libraries.
240 * Note that pcrs and pcre (native) are needed for cgi
241 * and are included anyway.
244 #if defined(REGEX_PCRE) || defined (REGEX_GNU)
246 #endif /* defined(REGEX_PCRE) || defined (REGEX_GNU) */
260 #if defined(REGEX_PCRE)
262 # include "pcreposix.h"
264 # include <pcreposix.h>
266 #endif /* defined(REGEX_PCRE) */
268 #if defined(REGEX_GNU)
269 # include "gnu_regex.h"
274 #endif /* def AMIGA */
280 #define freez(X) if(X) free(X); X = NULL
282 #define BUFFER_SIZE 5000
286 /* Default IP and port to listen on */
287 #define HADDR_DEFAULT "127.0.0.1"
288 #define HADDR_PORT 8000
291 /* Need this for struct client_state */
292 struct configuration_spec;
294 /* Generic linked list of strings */
295 struct list /* FIXME: Why not separate entries and header? */
297 char * str; /* valid in an entry */
298 struct list *last; /* valid in header */
314 char *host_ip_addr_str; /* NULL before connect_to() */
318 char *hostport; /* "host[:port]" */
320 char *user_agent; /* Client's User-Agent: header value */
323 /* Response generated by CGI, blocker, or error handler */
326 char *status; /* HTTP status (string)*/
327 struct list headers[1]; /* List of header lines */
328 char *head; /* Formatted http response head */
329 int head_length; /* Length of http response head */
330 char *body; /* HTTP document body */
331 int content_length; /* Length of body, REQUIRED if binary body*/
337 char *spec; /* The string which was parsed to produce this */
338 /* url_spec. Used for debugging or display only. */
340 /* Hostname matching: */
341 char *domain; /* Fully qalified domain name (FQDN) pattern. */
342 /* May contain "*". */
343 char *dbuf; /* Buffer with '\0'-delimited fqdn */
344 char **dvec; /* Domain ptr vector into dbuf */
345 int dcnt; /* How many domains in fqdn? */
346 int unanchored; /* Bitmap - flags are ANCHOR_LEFT and ANCHOR_RIGHT */
349 int port; /* The port number, or 0 to match all ports. */
352 char *path; /* The path prefix (if not using regex), or source */
354 int pathlen; /* ==strlen(path). Needed for prefix matching. */
356 regex_t *preg; /* Regex for matching path part */
360 /* Constants for host part matching in URLs */
361 #define ANCHOR_LEFT 1
362 #define ANCHOR_RIGHT 2
374 #define IOB_PEEK(CSP) ((CSP->iob->cur > CSP->iob->eod) ? (CSP->iob->eod - CSP->iob->cur) : 0)
375 #define IOB_RESET(CSP) if(CSP->iob->buf) free(CSP->iob->buf); memset(CSP->iob, '\0', sizeof(CSP->iob));
377 /* Keys for csp->content_type */
378 #define CT_TEXT 0x01U
381 #define ACTION_MASK_ALL (~0U)
383 #define ACTION_MOST_COMPATIBLE 0x0000U
385 #define ACTION_BLOCK 0x0001U
386 #define ACTION_DEANIMATE 0x2000U
387 #define ACTION_FAST_REDIRECTS 0x0002U
388 #define ACTION_FILTER 0x0004U
389 #define ACTION_HIDE_FORWARDED 0x0008U
390 #define ACTION_HIDE_FROM 0x0010U
391 #define ACTION_HIDE_REFERER 0x0020U /* sic - follow HTTP, not English */
392 #define ACTION_HIDE_USER_AGENT 0x0040U
393 #define ACTION_IMAGE 0x0080U
394 #define ACTION_IMAGE_BLOCKER 0x0100U
395 #define ACTION_NO_COOKIE_READ 0x0200U
396 #define ACTION_NO_COOKIE_SET 0x0400U
397 #define ACTION_NO_POPUPS 0x0800U
398 #define ACTION_VANILLA_WAFER 0x1000U
400 #define ACTION_STRING_DEANIMATE 0
401 #define ACTION_STRING_FROM 1
402 #define ACTION_STRING_IMAGE_BLOCKER 2
403 #define ACTION_STRING_REFERER 3
404 #define ACTION_STRING_USER_AGENT 4
405 #define ACTION_STRING_COUNT 5
408 #define ACTION_MULTI_ADD_HEADER 0
409 #define ACTION_MULTI_WAFER 1
410 #define ACTION_MULTI_COUNT 2
413 * This structure contains a list of actions to apply to a URL.
414 * It only contains positive instructions - no "-" options.
415 * It is not used to store the actions list itself, only for
416 * url_actions() to return the current values.
418 struct current_action_spec
420 unsigned flags; /* a bit set to "1" = add action */
422 /* For those actions that require parameters: */
424 /* each entry is valid if & only if corresponding entry in "add" set. */
425 char * string[ACTION_STRING_COUNT];
428 struct list multi[ACTION_MULTI_COUNT][1];
433 * This structure contains a set of changes to actions.
434 * It can contain both positive and negative instructions.
435 * It is used to store an entry in the actions list.
439 unsigned mask; /* a bit set to "0" = remove action */
440 unsigned add; /* a bit set to "1" = add action */
442 /* For those actions that require parameters: */
444 /* each entry is valid if & only if corresponding entry in "add" set. */
445 char * string[ACTION_STRING_COUNT];
447 /* Strings to remove. */
448 struct list multi_remove[ACTION_MULTI_COUNT][1];
450 /* If nonzero, remove *all* strings. */
451 int multi_remove_all[ACTION_MULTI_COUNT];
454 struct list multi_add[ACTION_MULTI_COUNT][1];
458 * This structure is used to store the actions list.
460 * It contains a URL pattern, and the chages to the actions.
461 * It is a linked list.
465 struct url_spec url[1];
467 struct action_spec action[1];
469 struct url_actions * next;
473 /* Constants defining bitmask for csp->accept_types */
475 #ifdef FEATURE_IMAGE_DETECT_MSIE
477 /* MSIE detected by user-agent string */
478 #define ACCEPT_TYPE_IS_MSIE 0x0001
481 * *If* this is MSIE, it wants an image. (Or this is a shift-reload, or
482 * it's got an image from this URL before... yuck!)
483 * Only meaningful if ACCEPT_TYPE_IS_MSIE set
485 #define ACCEPT_TYPE_MSIE_IMAGE 0x0002
488 * *If* this is MSIE, it wants a HTML document.
489 * Only meaningful if ACCEPT_TYPE_IS_MSIE set
491 #define ACCEPT_TYPE_MSIE_HTML 0x0004
493 #endif /* def FEATURE_IMAGE_DETECT_MSIE */
498 /* The proxy's configuration */
499 struct configuration_spec * config;
501 /* The actions to perform on the current request */
502 struct current_action_spec action[1];
504 /* socket to talk to client (web browser) */
507 /* socket to talk to server (web server or proxy) */
511 #ifdef FEATURE_STATISTICS
512 /* 1 if this URL was rejected, 0 otherwise. Allows actual stats inc to
513 * occur in main thread only for thread-safety.
516 #endif /* def FEATURE_STATISTICS */
518 #ifdef FEATURE_FORCE_LOAD
520 #endif /* def FEATURE_FORCE_LOAD */
522 #ifdef FEATURE_TOGGLE
524 #endif /* def FEATURE_TOGGLE */
527 * Client PC's IP address, as reported by the accept()_ function.
528 * Both as string and number
534 /* Our IP address and hostname, i.e. the IP address that
535 the client used to reach us, and the associated hostname,
538 char *my_ip_addr_str;
542 /* The referer in this request, if one was specified. */
544 #endif /* def FEATURE_TRUST */
546 #if defined(FEATURE_IMAGE_DETECT_MSIE)
547 /* Types the client will accept.
548 * Bitmask - see ACCEPT_TYPE_XXX constants.
551 #endif /* defined(FEATURE_IMAGE_DETECT_MSIE) */
553 /* The URL that was requested */
554 struct http_request http[1];
556 /* An I/O buffer used for buffering data read from the client */
559 /* List of all headers for this request */
560 struct list headers[1];
562 /* List of all cookies for this request */
563 struct list cookie_list[1];
565 /* MIME-Type bitmap, see CT_* above */
566 unsigned char content_type;
568 /* The "X-Forwarded-For:" header sent by the client */
572 * Nonzero if this client is processing data.
573 * Set to zero when the thread associated with this structure dies.
577 /* files associated with this client */
578 struct file_list *actions_list;
580 struct file_list *rlist; /* pcrs job file */
581 size_t content_length; /* Length after content modification */
584 struct file_list *tlist; /* trustfile */
585 #endif /* def FEATURE_TRUST */
587 struct client_state *next;
595 char *(*parser)(const struct parsers *, char *, struct client_state *);
598 struct cgi_dispatcher
602 int (*handler)(struct client_state *csp, struct http_response *rsp, struct map *parameters);
603 const char *description;
609 * this is a pointer to the data structures associated with the file.
610 * Read-only once the structure has been created.
614 /* Normally NULL. When we are finished with file (i.e. when we have
615 * loaded a new one), set to a pointer to an unloader function.
616 * Unloader will be called by sweep() (called from main loop) when
617 * all clients using this file are done. This prevents threading
620 void (*unloader)(void *);
622 /* Used internally by sweep(). Do not access from elsewhere. */
625 /* Following variables allow us to check if file has been changed.
626 * Read-only once the structure has been created.
631 /* Pointer to next entry in the linked list of all "file_list"s.
632 * This linked list is so that sweep() can navigate it.
633 * Since sweep() can remove items from the list, we must be careful
634 * to only access this value from main thread (when we know sweep
637 struct file_list *next;
644 struct url_spec url[1];
646 struct block_spec *next;
648 #endif /* def FEATURE_TRUST */
651 #define SOCKS_NONE 0 /* Don't use a SOCKS server */
652 #define SOCKS_4 40 /* original SOCKS 4 protocol */
653 #define SOCKS_4A 41 /* as modified for hosts w/o external DNS */
657 struct url_spec url[1];
659 /* Connection type - must be a SOCKS_xxx constant */
666 /* Parent HTTP proxy */
670 /* For the linked list */
671 struct forward_spec *next;
674 struct re_filterfile_spec
678 struct list patterns[1];
683 #define ACL_PERMIT 1 /* accept connection request */
684 #define ACL_DENY 2 /* reject connection request */
686 struct access_control_addr
693 struct access_control_list
695 struct access_control_addr src[1];
696 struct access_control_addr dst[1];
699 struct access_control_list *next;
701 #endif /* def FEATURE_ACL */
704 /* Maximum number of loaders (actions, re_filter, ...) */
708 * Data loaded from the configuration file.
710 * (Anomaly: toggle is still handled through a global, not this structure)
712 struct configuration_spec
721 const char *actions_file;
723 /* The administrator's email address */
726 /* A URL with info on this proxy */
727 char *proxy_info_url;
729 const char *re_filterfile;
731 #ifdef FEATURE_COOKIE_JAR
732 const char * jarfile;
734 #endif /* def FEATURE_COOKIE_JAR */
737 * Port and IP to bind to.
738 * Defaults to HADDR_DEFAULT:HADDR_PORT == 127.0.0.1:8000
744 const char * trustfile;
746 struct list trust_info[1];
747 struct url_spec *trust_list[64];
748 #endif /* def FEATURE_TRUST */
751 struct access_control_list *acl;
752 #endif /* def FEATURE_ACL */
754 struct forward_spec *forward;
756 /* All options from the config file, HTML-formatted */
759 /* the configuration file object. */
760 struct file_list *config_file_list;
762 /* List of loaders */
763 int (*loaders[NLOADERS])(struct client_state *);
765 /* bool, nonzero if we need to bind() to the new port */
770 #define SZ(X) (sizeof(X) / sizeof(*X))
772 #ifdef FEATURE_FORCE_LOAD
773 #define FORCE_PREFIX "/IJB-FORCE-LOAD"
774 #endif /* def FEATURE_FORCE_LOAD */
777 #define HOME_PAGE_URL "http://ijbswa.sourceforge.net"
778 #define REDIRECT_URL HOME_PAGE_URL "/redirect.php?v=" VERSION "&to="
779 #define CGI_PREFIX_HOST "i.j.b"
782 static const char CSUCCEED[] =
783 "HTTP/1.0 200 Connection established\n"
784 "Proxy-Agent: IJ/" VERSION "\n\n";
786 static const char CHEADER[] =
787 "HTTP/1.0 400 Invalid header received from browser\n\n";
793 #endif /* ndef PROJECT_H_INCLUDED */