1 #ifndef PROJECT_H_INCLUDED
2 #define PROJECT_H_INCLUDED
3 #define PROJECT_H_VERSION "$Id: project.h,v 1.24 2001/07/25 17:20:27 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 1.24 2001/07/25 17:20:27 oes
40 * Introduced http->user_agent
42 * Revision 1.23 2001/07/18 12:32:23 oes
43 * - Added ACTION_STRING_DEANIMATE
44 * - moved #define freez from jcc.h to project.h
46 * Revision 1.22 2001/07/15 17:51:41 jongfoster
47 * Renaming #define STATIC to STATIC_PCRE
49 * Revision 1.21 2001/07/13 14:03:19 oes
50 * - Reorganized regex header inclusion and #defines to
51 * comply to the scheme in configure.in
52 * - Added csp->content_type and its CT_* keys
53 * - Added ACTION_DEANIMATE
54 * - Removed all #ifdef PCRS
56 * Revision 1.20 2001/06/29 21:45:41 oes
57 * Indentation, CRLF->LF, Tab-> Space
59 * Revision 1.19 2001/06/29 13:33:36 oes
61 * - Introduced http_request.host_ip_addr_str
62 * - Introduced http_response.head_length
63 * - Introduced config.my_ip_addr_str, config.my_hostname,
64 * config.admin_address and config.proxy_info_url
65 * - Removed config.proxy_args_header and config.proxy_args_trailer,
66 * renamed config.proxy_args_invocation to config.proxy_args
67 * - Removed HTML snipplets and GIFs
68 * - Removed logentry from cancelled commit
70 * Revision 1.18 2001/06/09 10:57:39 jongfoster
71 * Adding definition of BUFFER_SIZE.
72 * Changing struct cgi_dispatcher to use "const" strings.
74 * Revision 1.17 2001/06/07 23:15:09 jongfoster
75 * Merging ACL and forward files into config file.
76 * Moving struct gateway members into struct forward_spec
77 * Removing config->proxy_args_gateways
78 * Cosmetic: Adding a few comments
80 * Revision 1.16 2001/06/04 18:31:58 swa
81 * files are now prefixed with either `confdir' or `logdir'.
82 * `make redhat-dist' replaces both entries confdir and logdir
85 * Revision 1.15 2001/06/04 11:28:53 swa
86 * redirect did not work due to missing /
88 * Revision 1.14 2001/06/03 11:03:48 oes
90 * added struct http_response,
91 * changed struct interceptors to struct cgi_dispatcher,
92 * moved HTML stuff to cgi.h
94 * Revision 1.13 2001/06/01 20:05:36 jongfoster
95 * Support for +image-blocker{}: added ACTION_IMAGE_BLOCKER
96 * constant, and removed csp->tinygif.
98 * Revision 1.12 2001/06/01 18:49:17 jongfoster
99 * Replaced "list_share" with "list" - the tiny memory gain was not
100 * worth the extra complexity.
102 * Revision 1.11 2001/06/01 10:32:47 oes
103 * Added constants for anchoring selection bitmap
105 * Revision 1.10 2001/05/31 21:33:53 jongfoster
106 * Changes for new actions file, replacing permissionsfile
107 * and parts of the config file. Also added support for
110 * Revision 1.9 2001/05/31 17:32:31 oes
112 * - Enhanced domain part globbing with infix and prefix asterisk
113 * matching and optional unanchored operation
115 * Revision 1.8 2001/05/29 20:09:15 joergs
116 * HTTP_REDIRECT_TEMPLATE fixed.
118 * Revision 1.7 2001/05/29 09:50:24 jongfoster
119 * Unified blocklist/imagelist/actionslist.
120 * File format is still under discussion, but the internal changes
123 * Also modified interceptor behaviour:
124 * - We now intercept all URLs beginning with one of the following
125 * prefixes (and *only* these prefixes):
127 * * http://ijbswa.sf.net/config/
128 * * http://ijbswa.sourceforge.net/config/
129 * - New interceptors "home page" - go to http://i.j.b/ to see it.
130 * - Internal changes so that intercepted and fast redirect pages
131 * are not replaced with an image.
132 * - Interceptors now have the option to send a binary page direct
133 * to the client. (i.e. ijb-send-banner uses this)
134 * - Implemented show-url-info interceptor. (Which is why I needed
135 * the above interceptors changes - a typical URL is
136 * "http://i.j.b/show-url-info?url=www.somesite.com/banner.gif".
137 * The previous mechanism would not have intercepted that, and
138 * if it had been intercepted then it then it would have replaced
141 * Revision 1.6 2001/05/27 22:17:04 oes
143 * - re_process_buffer no longer writes the modified buffer
144 * to the client, which was very ugly. It now returns the
145 * buffer, which it is then written by chat.
147 * - content_length now adjusts the Content-Length: header
148 * for modified documents rather than crunch()ing it.
149 * (Length info in csp->content_length, which is 0 for
150 * unmodified documents)
152 * - For this to work, sed() is called twice when filtering.
154 * Revision 1.5 2001/05/26 00:28:36 jongfoster
155 * Automatic reloading of config file.
156 * Removed obsolete SIGHUP support (Unix) and Reload menu option (Win32).
157 * Most of the global variables have been moved to a new
158 * struct configuration_spec, accessed through csp->config->globalname
159 * Most of the globals remaining are used by the Win32 GUI.
161 * Revision 1.4 2001/05/22 18:46:04 oes
163 * - Enabled filtering banners by size rather than URL
164 * by adding patterns that replace all standard banner
165 * sizes with the "Junkbuster" gif to the re_filterfile
167 * - Enabled filtering WebBugs by providing a pattern
168 * which kills all 1x1 images
170 * - Added support for PCRE_UNGREEDY behaviour to pcrs,
171 * which is selected by the (nonstandard and therefore
172 * capital) letter 'U' in the option string.
173 * It causes the quantifiers to be ungreedy by default.
174 * Appending a ? turns back to greedy (!).
176 * - Added a new interceptor ijb-send-banner, which
177 * sends back the "Junkbuster" gif. Without imagelist or
178 * MSIE detection support, or if tinygif = 1, or the
179 * URL isn't recognized as an imageurl, a lame HTML
180 * explanation is sent instead.
182 * - Added new feature, which permits blocking remote
183 * script redirects and firing back a local redirect
185 * The feature is conditionally compiled, i.e. it
186 * can be disabled with --disable-fast-redirects,
187 * plus it must be activated by a "fast-redirects"
188 * line in the config file, has its own log level
189 * and of course wants to be displayed by show-proxy-args
190 * Note: Boy, all the #ifdefs in 1001 locations and
191 * all the fumbling with configure.in and acconfig.h
192 * were *way* more work than the feature itself :-(
194 * - Because a generic redirect template was needed for
195 * this, tinygif = 3 now uses the same.
197 * - Moved GIFs, and other static HTTP response templates
202 * - Removed some >400 CRs again (Jon, you really worked
205 * Revision 1.3 2001/05/20 01:21:20 jongfoster
206 * Version 2.9.4 checkin.
207 * - Merged popupfile and cookiefile, and added control over PCRS
208 * filtering, in new "actionsfile".
209 * - Implemented LOG_LEVEL_FATAL, so that if there is a configuration
210 * file error you now get a message box (in the Win32 GUI) rather
211 * than the program exiting with no explanation.
212 * - Made killpopup use the PCRS MIME-type checking and HTTP-header
214 * - Removed tabs from "config"
215 * - Moved duplicated url parsing code in "loaders.c" to a new funcition.
216 * - Bumped up version number.
218 * Revision 1.2 2001/05/17 23:01:01 oes
219 * - Cleaned CRLF's from the sources and related files
221 * Revision 1.1.1.1 2001/05/15 13:59:03 oes
222 * Initial import of version 2.9.3 source tree
225 *********************************************************************/
228 /* Declare struct FILE for vars and funcs. */
231 /* Need time_t for file_list */
235 * Include appropriate regular expression libraries.
236 * Note that pcrs and pcre (native) are needed for cgi
237 * and are included anyway.
240 #if defined(REGEX_PCRE) || defined (REGEX_GNU)
242 #endif /* defined(REGEX_PCRE) || defined (REGEX_GNU) */
256 #if defined(REGEX_PCRE)
258 # include "pcreposix.h"
260 # include <pcreposix.h>
262 #endif /* defined(REGEX_PCRE) */
264 #if defined(REGEX_GNU)
265 # include "gnu_regex.h"
270 #endif /* def AMIGA */
276 #define freez(X) if(X) free(X); X = NULL
278 #define BUFFER_SIZE 5000
282 /* Default IP and port to listen on */
283 #define HADDR_DEFAULT "127.0.0.1"
284 #define HADDR_PORT 8000
287 /* Need this for struct client_state */
288 struct configuration_spec;
290 /* Generic linked list of strings */
291 struct list /* FIXME: Why not separate entries and header? */
293 char * str; /* valid in an entry */
294 struct list *last; /* valid in header */
310 char *host_ip_addr_str; /* NULL before connect_to() */
314 char *hostport; /* "host[:port]" */
316 char *user_agent; /* Client's User-Agent: header value */
319 /* Response generated by CGI, blocker, or error handler */
322 char *status; /* HTTP status (string)*/
323 struct list headers[1]; /* List of header lines */
324 char *head; /* Formatted http response head */
325 int head_length; /* Length of http response head */
326 char *body; /* HTTP document body */
327 int content_length; /* Length of body, REQUIRED if binary body*/
333 char *spec; /* The string which was parsed to produce this */
334 /* url_spec. Used for debugging or display only. */
336 /* Hostname matching: */
337 char *domain; /* Fully qalified domain name (FQDN) pattern. */
338 /* May contain "*". */
339 char *dbuf; /* Buffer with '\0'-delimited fqdn */
340 char **dvec; /* Domain ptr vector into dbuf */
341 int dcnt; /* How many domains in fqdn? */
342 int unanchored; /* Bitmap - flags are ANCHOR_LEFT and ANCHOR_RIGHT */
345 int port; /* The port number, or 0 to match all ports. */
348 char *path; /* The path prefix (if not using regex), or source */
350 int pathlen; /* ==strlen(path). Needed for prefix matching. */
352 regex_t *preg; /* Regex for matching path part */
356 /* Constants for host part matching in URLs */
357 #define ANCHOR_LEFT 1
358 #define ANCHOR_RIGHT 2
370 #define IOB_PEEK(CSP) ((CSP->iob->cur > CSP->iob->eod) ? (CSP->iob->eod - CSP->iob->cur) : 0)
371 #define IOB_RESET(CSP) if(CSP->iob->buf) free(CSP->iob->buf); memset(CSP->iob, '\0', sizeof(CSP->iob));
373 /* Keys for csp->content_type */
374 #define CT_TEXT 0x01U
377 #define ACTION_MASK_ALL (~0U)
379 #define ACTION_MOST_COMPATIBLE 0x0000U
381 #define ACTION_BLOCK 0x0001U
382 #define ACTION_DEANIMATE 0x2000U
383 #define ACTION_FAST_REDIRECTS 0x0002U
384 #define ACTION_FILTER 0x0004U
385 #define ACTION_HIDE_FORWARDED 0x0008U
386 #define ACTION_HIDE_FROM 0x0010U
387 #define ACTION_HIDE_REFERER 0x0020U /* sic - follow HTTP, not English */
388 #define ACTION_HIDE_USER_AGENT 0x0040U
389 #define ACTION_IMAGE 0x0080U
390 #define ACTION_IMAGE_BLOCKER 0x0100U
391 #define ACTION_NO_COOKIE_READ 0x0200U
392 #define ACTION_NO_COOKIE_SET 0x0400U
393 #define ACTION_NO_POPUPS 0x0800U
394 #define ACTION_VANILLA_WAFER 0x1000U
396 #define ACTION_STRING_DEANIMATE 0
397 #define ACTION_STRING_FROM 1
398 #define ACTION_STRING_IMAGE_BLOCKER 2
399 #define ACTION_STRING_REFERER 3
400 #define ACTION_STRING_USER_AGENT 4
401 #define ACTION_STRING_COUNT 5
404 #define ACTION_MULTI_ADD_HEADER 0
405 #define ACTION_MULTI_WAFER 1
406 #define ACTION_MULTI_COUNT 2
409 * This structure contains a list of actions to apply to a URL.
410 * It only contains positive instructions - no "-" options.
411 * It is not used to store the actions list itself, only for
412 * url_actions() to return the current values.
414 struct current_action_spec
416 unsigned flags; /* a bit set to "1" = add action */
418 /* For those actions that require parameters: */
420 /* each entry is valid if & only if corresponding entry in "add" set. */
421 char * string[ACTION_STRING_COUNT];
424 struct list multi[ACTION_MULTI_COUNT][1];
429 * This structure contains a set of changes to actions.
430 * It can contain both positive and negative instructions.
431 * It is used to store an entry in the actions list.
435 unsigned mask; /* a bit set to "0" = remove action */
436 unsigned add; /* a bit set to "1" = add action */
438 /* For those actions that require parameters: */
440 /* each entry is valid if & only if corresponding entry in "add" set. */
441 char * string[ACTION_STRING_COUNT];
443 /* Strings to remove. */
444 struct list multi_remove[ACTION_MULTI_COUNT][1];
446 /* If nonzero, remove *all* strings. */
447 int multi_remove_all[ACTION_MULTI_COUNT];
450 struct list multi_add[ACTION_MULTI_COUNT][1];
454 * This structure is used to store the actions list.
456 * It contains a URL pattern, and the chages to the actions.
457 * It is a linked list.
461 struct url_spec url[1];
463 struct action_spec action[1];
465 struct url_actions * next;
469 /* Constants defining bitmask for csp->accept_types */
471 #ifdef DETECT_MSIE_IMAGES
473 /* MSIE detected by user-agent string */
474 #define ACCEPT_TYPE_IS_MSIE 0x0001
477 * *If* this is MSIE, it wants an image. (Or this is a shift-reload, or
478 * it's got an image from this URL before... yuck!)
479 * Only meaningful if ACCEPT_TYPE_IS_MSIE set
481 #define ACCEPT_TYPE_MSIE_IMAGE 0x0002
484 * *If* this is MSIE, it wants a HTML document.
485 * Only meaningful if ACCEPT_TYPE_IS_MSIE set
487 #define ACCEPT_TYPE_MSIE_HTML 0x0004
489 #endif /* def DETECT_MSIE_IMAGES */
494 /* The proxy's configuration */
495 struct configuration_spec * config;
497 /* The actions to perform on the current request */
498 struct current_action_spec action[1];
500 /* socket to talk to client (web browser) */
503 /* socket to talk to server (web server or proxy) */
508 /* 1 if this URL was rejected, 0 otherwise. Allows actual stats inc to
509 * occur in main thread only for thread-safety.
512 #endif /* def STATISTICS */
516 #endif /* def FORCE_LOAD */
520 #endif /* def TOGGLE */
523 * Client PC's IP address, as reported by the accept()_ function.
524 * Both as string and number
530 /* Our IP address and hostname, i.e. the IP address that
531 the client used to reach us, and the associated hostname,
534 char *my_ip_addr_str;
538 /* The referer in this request, if one was specified. */
540 #endif /* def TRUST_FILES */
542 #if defined(DETECT_MSIE_IMAGES)
543 /* Types the client will accept.
544 * Bitmask - see ACCEPT_TYPE_XXX constants.
547 #endif /* defined(DETECT_MSIE_IMAGES) */
549 /* The URL that was requested */
550 struct http_request http[1];
552 /* An I/O buffer used for buffering data read from the client */
555 /* List of all headers for this request */
556 struct list headers[1];
558 /* List of all cookies for this request */
559 struct list cookie_list[1];
561 /* MIME-Type bitmap, see CT_* above */
562 unsigned char content_type;
564 /* The "X-Forwarded-For:" header sent by the client */
568 * Nonzero if this client is processing data.
569 * Set to zero when the thread associated with this structure dies.
573 /* files associated with this client */
574 struct file_list *actions_list;
576 struct file_list *rlist; /* pcrs job file */
577 size_t content_length; /* Length after content modification */
580 struct file_list *tlist; /* trustfile */
581 #endif /* def TRUST_FILES */
583 struct client_state *next;
591 char *(*parser)(const struct parsers *, char *, struct client_state *);
594 struct cgi_dispatcher
598 int (*handler)(struct client_state *csp, struct http_response *rsp, struct map *parameters);
599 const char *description;
605 * this is a pointer to the data structures associated with the file.
606 * Read-only once the structure has been created.
610 /* Normally NULL. When we are finished with file (i.e. when we have
611 * loaded a new one), set to a pointer to an unloader function.
612 * Unloader will be called by sweep() (called from main loop) when
613 * all clients using this file are done. This prevents threading
616 void (*unloader)(void *);
618 /* Used internally by sweep(). Do not access from elsewhere. */
621 #ifndef SPLIT_PROXY_ARGS
622 /* String to be displayed as part of show-proxy-args display.
623 * Read-only once the structure has been created.
626 #endif /* ndef SPLIT_PROXY_ARGS */
628 /* Following variables allow us to check if file has been changed.
629 * Read-only once the structure has been created.
634 /* Pointer to next entry in the linked list of all "file_list"s.
635 * This linked list is so that sweep() can navigate it.
636 * Since sweep() can remove items from the list, we must be careful
637 * to only access this value from main thread (when we know sweep
640 struct file_list *next;
647 struct url_spec url[1];
649 struct block_spec *next;
651 #endif /* def TRUST_FILES */
654 #define SOCKS_NONE 0 /* Don't use a SOCKS server */
655 #define SOCKS_4 40 /* original SOCKS 4 protocol */
656 #define SOCKS_4A 41 /* as modified for hosts w/o external DNS */
660 struct url_spec url[1];
662 /* Connection type - must be a SOCKS_xxx constant */
669 /* Parent HTTP proxy */
673 /* For the linked list */
674 struct forward_spec *next;
677 struct re_filterfile_spec
681 struct list patterns[1];
686 #define ACL_PERMIT 1 /* accept connection request */
687 #define ACL_DENY 2 /* reject connection request */
689 struct access_control_addr
696 struct access_control_list
698 struct access_control_addr src[1];
699 struct access_control_addr dst[1];
702 struct access_control_list *next;
704 #endif /* def ACL_FILES */
707 /* Maximum number of loaders (actions, re_filter, ...) */
711 * Data loaded from the configuration file.
713 * (Anomaly: toggle is still handled through a global, not this structure)
715 struct configuration_spec
724 const char *actions_file;
726 /* The administrator's email address */
729 /* A URL with info on this proxy */
730 char *proxy_info_url;
732 const char *re_filterfile;
735 const char * jarfile;
737 #endif /* def JAR_FILES */
740 * Port and IP to bind to.
741 * Defaults to HADDR_DEFAULT:HADDR_PORT == 127.0.0.1:8000
746 #ifndef SPLIT_PROXY_ARGS
747 const char *suppress_message;
748 #endif /* ndef SPLIT_PROXY_ARGS */
750 #ifndef SPLIT_PROXY_ARGS
751 /* suppress listing config files */
752 int suppress_blocklists;
753 #endif /* ndef SPLIT_PROXY_ARGS */
756 const char * trustfile;
758 struct list trust_info[1];
759 struct url_spec *trust_list[64];
760 #endif /* def TRUST_FILES */
763 struct access_control_list *acl;
764 #endif /* def ACL_FILES */
766 struct forward_spec *forward;
768 /* All options from the config file, HTML-formatted */
771 /* the configuration file object. */
772 struct file_list *config_file_list;
774 /* List of loaders */
775 int (*loaders[NLOADERS])(struct client_state *);
777 /* bool, nonzero if we need to bind() to the new port */
782 #define SZ(X) (sizeof(X) / sizeof(*X))
785 #define FORCE_PREFIX "/IJB-FORCE-LOAD"
786 #endif /* def FORCE_LOAD */
789 #define HOME_PAGE_URL "http://ijbswa.sourceforge.net"
790 #define REDIRECT_URL HOME_PAGE_URL "/redirect.php?v=" VERSION "&to="
791 #define CGI_PREFIX_HOST "i.j.b"
794 static const char CSUCCEED[] =
795 "HTTP/1.0 200 Connection established\n"
796 "Proxy-Agent: IJ/" VERSION "\n\n";
798 static const char CHEADER[] =
799 "HTTP/1.0 400 Invalid header received from browser\n\n";
805 #endif /* ndef PROJECT_H_INCLUDED */