1 #ifndef PROJECT_H_INCLUDED
2 #define PROJECT_H_INCLUDED
3 #define PROJECT_H_VERSION "$Id: project.h,v 1.32 2001/09/16 23:02:51 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 1.32 2001/09/16 23:02:51 jongfoster
42 * Revision 1.31 2001/09/16 13:20:29 jongfoster
43 * Rewrite of list library. Now has seperate header and list_entry
44 * structures. Also added a large sprinking of assert()s to the list
47 * Revision 1.30 2001/09/13 23:52:00 jongfoster
48 * Support for both static and dynamically generated CGI pages
50 * Revision 1.29 2001/09/13 23:29:43 jongfoster
51 * Defining FORWARD_SPEC_INITIALIZER
53 * Revision 1.28 2001/09/13 23:05:50 jongfoster
54 * Changing the string paramater to the header parsers a "const".
56 * Revision 1.27 2001/08/05 16:06:20 jongfoster
57 * Modifiying "struct map" so that there are now separate header and
58 * "map_entry" structures. This means that functions which modify a
59 * map no longer need to return a pointer to the modified map.
60 * Also, it no longer reverses the order of the entries (which may be
61 * important with some advanced template substitutions).
63 * Revision 1.26 2001/07/30 22:08:36 jongfoster
64 * Tidying up #defines:
65 * - All feature #defines are now of the form FEATURE_xxx
66 * - Permanently turned off WIN_GUI_EDIT
67 * - Permanently turned on WEBDAV and SPLIT_PROXY_ARGS
69 * Revision 1.25 2001/07/29 18:43:08 jongfoster
70 * Changing #ifdef _FILENAME_H to FILENAME_H_INCLUDED, to conform to
73 * Revision 1.24 2001/07/25 17:20:27 oes
74 * Introduced http->user_agent
76 * Revision 1.23 2001/07/18 12:32:23 oes
77 * - Added ACTION_STRING_DEANIMATE
78 * - moved #define freez from jcc.h to project.h
80 * Revision 1.22 2001/07/15 17:51:41 jongfoster
81 * Renaming #define STATIC to STATIC_PCRE
83 * Revision 1.21 2001/07/13 14:03:19 oes
84 * - Reorganized regex header inclusion and #defines to
85 * comply to the scheme in configure.in
86 * - Added csp->content_type and its CT_* keys
87 * - Added ACTION_DEANIMATE
88 * - Removed all #ifdef PCRS
90 * Revision 1.20 2001/06/29 21:45:41 oes
91 * Indentation, CRLF->LF, Tab-> Space
93 * Revision 1.19 2001/06/29 13:33:36 oes
95 * - Introduced http_request.host_ip_addr_str
96 * - Introduced http_response.head_length
97 * - Introduced config.my_ip_addr_str, config.my_hostname,
98 * config.admin_address and config.proxy_info_url
99 * - Removed config.proxy_args_header and config.proxy_args_trailer,
100 * renamed config.proxy_args_invocation to config.proxy_args
101 * - Removed HTML snipplets and GIFs
102 * - Removed logentry from cancelled commit
104 * Revision 1.18 2001/06/09 10:57:39 jongfoster
105 * Adding definition of BUFFER_SIZE.
106 * Changing struct cgi_dispatcher to use "const" strings.
108 * Revision 1.17 2001/06/07 23:15:09 jongfoster
109 * Merging ACL and forward files into config file.
110 * Moving struct gateway members into struct forward_spec
111 * Removing config->proxy_args_gateways
112 * Cosmetic: Adding a few comments
114 * Revision 1.16 2001/06/04 18:31:58 swa
115 * files are now prefixed with either `confdir' or `logdir'.
116 * `make redhat-dist' replaces both entries confdir and logdir
119 * Revision 1.15 2001/06/04 11:28:53 swa
120 * redirect did not work due to missing /
122 * Revision 1.14 2001/06/03 11:03:48 oes
124 * added struct http_response,
125 * changed struct interceptors to struct cgi_dispatcher,
126 * moved HTML stuff to cgi.h
128 * Revision 1.13 2001/06/01 20:05:36 jongfoster
129 * Support for +image-blocker{}: added ACTION_IMAGE_BLOCKER
130 * constant, and removed csp->tinygif.
132 * Revision 1.12 2001/06/01 18:49:17 jongfoster
133 * Replaced "list_share" with "list" - the tiny memory gain was not
134 * worth the extra complexity.
136 * Revision 1.11 2001/06/01 10:32:47 oes
137 * Added constants for anchoring selection bitmap
139 * Revision 1.10 2001/05/31 21:33:53 jongfoster
140 * Changes for new actions file, replacing permissionsfile
141 * and parts of the config file. Also added support for
144 * Revision 1.9 2001/05/31 17:32:31 oes
146 * - Enhanced domain part globbing with infix and prefix asterisk
147 * matching and optional unanchored operation
149 * Revision 1.8 2001/05/29 20:09:15 joergs
150 * HTTP_REDIRECT_TEMPLATE fixed.
152 * Revision 1.7 2001/05/29 09:50:24 jongfoster
153 * Unified blocklist/imagelist/actionslist.
154 * File format is still under discussion, but the internal changes
157 * Also modified interceptor behaviour:
158 * - We now intercept all URLs beginning with one of the following
159 * prefixes (and *only* these prefixes):
161 * * http://ijbswa.sf.net/config/
162 * * http://ijbswa.sourceforge.net/config/
163 * - New interceptors "home page" - go to http://i.j.b/ to see it.
164 * - Internal changes so that intercepted and fast redirect pages
165 * are not replaced with an image.
166 * - Interceptors now have the option to send a binary page direct
167 * to the client. (i.e. ijb-send-banner uses this)
168 * - Implemented show-url-info interceptor. (Which is why I needed
169 * the above interceptors changes - a typical URL is
170 * "http://i.j.b/show-url-info?url=www.somesite.com/banner.gif".
171 * The previous mechanism would not have intercepted that, and
172 * if it had been intercepted then it then it would have replaced
175 * Revision 1.6 2001/05/27 22:17:04 oes
177 * - re_process_buffer no longer writes the modified buffer
178 * to the client, which was very ugly. It now returns the
179 * buffer, which it is then written by chat.
181 * - content_length now adjusts the Content-Length: header
182 * for modified documents rather than crunch()ing it.
183 * (Length info in csp->content_length, which is 0 for
184 * unmodified documents)
186 * - For this to work, sed() is called twice when filtering.
188 * Revision 1.5 2001/05/26 00:28:36 jongfoster
189 * Automatic reloading of config file.
190 * Removed obsolete SIGHUP support (Unix) and Reload menu option (Win32).
191 * Most of the global variables have been moved to a new
192 * struct configuration_spec, accessed through csp->config->globalname
193 * Most of the globals remaining are used by the Win32 GUI.
195 * Revision 1.4 2001/05/22 18:46:04 oes
197 * - Enabled filtering banners by size rather than URL
198 * by adding patterns that replace all standard banner
199 * sizes with the "Junkbuster" gif to the re_filterfile
201 * - Enabled filtering WebBugs by providing a pattern
202 * which kills all 1x1 images
204 * - Added support for PCRE_UNGREEDY behaviour to pcrs,
205 * which is selected by the (nonstandard and therefore
206 * capital) letter 'U' in the option string.
207 * It causes the quantifiers to be ungreedy by default.
208 * Appending a ? turns back to greedy (!).
210 * - Added a new interceptor ijb-send-banner, which
211 * sends back the "Junkbuster" gif. Without imagelist or
212 * MSIE detection support, or if tinygif = 1, or the
213 * URL isn't recognized as an imageurl, a lame HTML
214 * explanation is sent instead.
216 * - Added new feature, which permits blocking remote
217 * script redirects and firing back a local redirect
219 * The feature is conditionally compiled, i.e. it
220 * can be disabled with --disable-fast-redirects,
221 * plus it must be activated by a "fast-redirects"
222 * line in the config file, has its own log level
223 * and of course wants to be displayed by show-proxy-args
224 * Note: Boy, all the #ifdefs in 1001 locations and
225 * all the fumbling with configure.in and acconfig.h
226 * were *way* more work than the feature itself :-(
228 * - Because a generic redirect template was needed for
229 * this, tinygif = 3 now uses the same.
231 * - Moved GIFs, and other static HTTP response templates
236 * - Removed some >400 CRs again (Jon, you really worked
239 * Revision 1.3 2001/05/20 01:21:20 jongfoster
240 * Version 2.9.4 checkin.
241 * - Merged popupfile and cookiefile, and added control over PCRS
242 * filtering, in new "actionsfile".
243 * - Implemented LOG_LEVEL_FATAL, so that if there is a configuration
244 * file error you now get a message box (in the Win32 GUI) rather
245 * than the program exiting with no explanation.
246 * - Made killpopup use the PCRS MIME-type checking and HTTP-header
248 * - Removed tabs from "config"
249 * - Moved duplicated url parsing code in "loaders.c" to a new funcition.
250 * - Bumped up version number.
252 * Revision 1.2 2001/05/17 23:01:01 oes
253 * - Cleaned CRLF's from the sources and related files
255 * Revision 1.1.1.1 2001/05/15 13:59:03 oes
256 * Initial import of version 2.9.3 source tree
259 *********************************************************************/
262 /* Declare struct FILE for vars and funcs. */
265 /* Need time_t for file_list */
269 * Include appropriate regular expression libraries.
270 * Note that pcrs and pcre (native) are needed for cgi
271 * and are included anyway.
274 #if defined(REGEX_PCRE) || defined (REGEX_GNU)
276 #endif /* defined(REGEX_PCRE) || defined (REGEX_GNU) */
290 #if defined(REGEX_PCRE)
292 # include "pcreposix.h"
294 # include <pcreposix.h>
296 #endif /* defined(REGEX_PCRE) */
298 #if defined(REGEX_GNU)
299 # include "gnu_regex.h"
304 #endif /* def AMIGA */
310 #define freez(X) { if(X) { free(X); X = NULL ; } }
312 #define BUFFER_SIZE 5000
316 /* Default IP and port to listen on */
317 #define HADDR_DEFAULT "127.0.0.1"
318 #define HADDR_PORT 8000
321 /* Forward defs for various structures */
323 /* Need this for struct client_state */
324 struct configuration_spec;
327 /* Generic linked list of strings */
332 struct list_entry *next;
337 struct list_entry *first;
338 struct list_entry *last;
342 /* A map from a string to another string */
348 struct map_entry *next;
353 struct map_entry *first;
354 struct map_entry *last;
363 char *host_ip_addr_str; /* NULL before connect_to() */
367 char *hostport; /* "host[:port]" */
369 char *user_agent; /* Client's User-Agent: header value */
372 /* Response generated by CGI, blocker, or error handler */
375 char *status; /* HTTP status (string)*/
376 struct list headers[1]; /* List of header lines */
377 char *head; /* Formatted http response head */
378 int head_length; /* Length of http response head */
379 char *body; /* HTTP document body */
380 int content_length; /* Length of body, REQUIRED if binary body */
381 int is_static; /* Nonzero if the content will never change and
382 * should be cached by the broser (e.g. images) */
388 char *spec; /* The string which was parsed to produce this */
389 /* url_spec. Used for debugging or display only. */
391 /* Hostname matching: */
392 char *domain; /* Fully qalified domain name (FQDN) pattern. */
393 /* May contain "*". */
394 char *dbuf; /* Buffer with '\0'-delimited fqdn */
395 char **dvec; /* Domain ptr vector into dbuf */
396 int dcnt; /* How many domains in fqdn? */
397 int unanchored; /* Bitmap - flags are ANCHOR_LEFT and ANCHOR_RIGHT */
400 int port; /* The port number, or 0 to match all ports. */
403 char *path; /* The path prefix (if not using regex), or source */
405 int pathlen; /* ==strlen(path). Needed for prefix matching. */
407 regex_t *preg; /* Regex for matching path part */
411 #define URL_SPEC_INITIALIZER { NULL, NULL, NULL, NULL, 0, 0, 0, NULL, 0, NULL }
412 #else /* ifndef REGEX */
413 #define URL_SPEC_INITIALIZER { NULL, NULL, NULL, NULL, 0, 0, 0, NULL, 0 }
414 #endif /* ndef REGEX */
416 /* Constants for host part matching in URLs */
417 #define ANCHOR_LEFT 1
418 #define ANCHOR_RIGHT 2
430 #define IOB_PEEK(CSP) ((CSP->iob->cur > CSP->iob->eod) ? (CSP->iob->eod - CSP->iob->cur) : 0)
431 #define IOB_RESET(CSP) if(CSP->iob->buf) free(CSP->iob->buf); memset(CSP->iob, '\0', sizeof(CSP->iob));
433 /* Keys for csp->content_type */
434 #define CT_TEXT 0x01U
437 #define ACTION_MASK_ALL (~0U)
439 #define ACTION_MOST_COMPATIBLE 0x0000U
441 #define ACTION_BLOCK 0x0001U
442 #define ACTION_DEANIMATE 0x2000U
443 #define ACTION_FAST_REDIRECTS 0x0002U
444 #define ACTION_FILTER 0x0004U
445 #define ACTION_HIDE_FORWARDED 0x0008U
446 #define ACTION_HIDE_FROM 0x0010U
447 #define ACTION_HIDE_REFERER 0x0020U /* sic - follow HTTP, not English */
448 #define ACTION_HIDE_USER_AGENT 0x0040U
449 #define ACTION_IMAGE 0x0080U
450 #define ACTION_IMAGE_BLOCKER 0x0100U
451 #define ACTION_NO_COOKIE_READ 0x0200U
452 #define ACTION_NO_COOKIE_SET 0x0400U
453 #define ACTION_NO_POPUPS 0x0800U
454 #define ACTION_VANILLA_WAFER 0x1000U
456 #define ACTION_STRING_DEANIMATE 0
457 #define ACTION_STRING_FROM 1
458 #define ACTION_STRING_IMAGE_BLOCKER 2
459 #define ACTION_STRING_REFERER 3
460 #define ACTION_STRING_USER_AGENT 4
461 #define ACTION_STRING_COUNT 5
464 #define ACTION_MULTI_ADD_HEADER 0
465 #define ACTION_MULTI_WAFER 1
466 #define ACTION_MULTI_COUNT 2
469 * This structure contains a list of actions to apply to a URL.
470 * It only contains positive instructions - no "-" options.
471 * It is not used to store the actions list itself, only for
472 * url_actions() to return the current values.
474 struct current_action_spec
476 unsigned flags; /* a bit set to "1" = add action */
478 /* For those actions that require parameters: */
480 /* each entry is valid if & only if corresponding entry in "add" set. */
481 char * string[ACTION_STRING_COUNT];
484 struct list multi[ACTION_MULTI_COUNT][1];
489 * This structure contains a set of changes to actions.
490 * It can contain both positive and negative instructions.
491 * It is used to store an entry in the actions list.
495 unsigned mask; /* a bit set to "0" = remove action */
496 unsigned add; /* a bit set to "1" = add action */
498 /* For those actions that require parameters: */
500 /* each entry is valid if & only if corresponding entry in "add" set. */
501 char * string[ACTION_STRING_COUNT];
503 /* Strings to remove. */
504 struct list multi_remove[ACTION_MULTI_COUNT][1];
506 /* If nonzero, remove *all* strings. */
507 int multi_remove_all[ACTION_MULTI_COUNT];
510 struct list multi_add[ACTION_MULTI_COUNT][1];
514 * This structure is used to store the actions list.
516 * It contains a URL pattern, and the chages to the actions.
517 * It is a linked list.
521 struct url_spec url[1];
523 struct action_spec action[1];
525 struct url_actions * next;
529 /* Constants defining bitmask for csp->accept_types */
531 #ifdef FEATURE_IMAGE_DETECT_MSIE
533 /* MSIE detected by user-agent string */
534 #define ACCEPT_TYPE_IS_MSIE 0x0001
537 * *If* this is MSIE, it wants an image. (Or this is a shift-reload, or
538 * it's got an image from this URL before... yuck!)
539 * Only meaningful if ACCEPT_TYPE_IS_MSIE set
541 #define ACCEPT_TYPE_MSIE_IMAGE 0x0002
544 * *If* this is MSIE, it wants a HTML document.
545 * Only meaningful if ACCEPT_TYPE_IS_MSIE set
547 #define ACCEPT_TYPE_MSIE_HTML 0x0004
549 #endif /* def FEATURE_IMAGE_DETECT_MSIE */
554 /* The proxy's configuration */
555 struct configuration_spec * config;
557 /* The actions to perform on the current request */
558 struct current_action_spec action[1];
560 /* socket to talk to client (web browser) */
563 /* socket to talk to server (web server or proxy) */
567 #ifdef FEATURE_STATISTICS
568 /* 1 if this URL was rejected, 0 otherwise. Allows actual stats inc to
569 * occur in main thread only for thread-safety.
572 #endif /* def FEATURE_STATISTICS */
574 #ifdef FEATURE_FORCE_LOAD
576 #endif /* def FEATURE_FORCE_LOAD */
578 #ifdef FEATURE_TOGGLE
580 #endif /* def FEATURE_TOGGLE */
583 * Client PC's IP address, as reported by the accept()_ function.
584 * Both as string and number
590 /* Our IP address and hostname, i.e. the IP address that
591 the client used to reach us, and the associated hostname,
594 char *my_ip_addr_str;
598 /* The referer in this request, if one was specified. */
600 #endif /* def FEATURE_TRUST */
602 #if defined(FEATURE_IMAGE_DETECT_MSIE)
603 /* Types the client will accept.
604 * Bitmask - see ACCEPT_TYPE_XXX constants.
607 #endif /* defined(FEATURE_IMAGE_DETECT_MSIE) */
609 /* The URL that was requested */
610 struct http_request http[1];
612 /* An I/O buffer used for buffering data read from the client */
615 /* List of all headers for this request */
616 struct list headers[1];
618 /* List of all cookies for this request */
619 struct list cookie_list[1];
621 /* MIME-Type bitmap, see CT_* above */
622 unsigned char content_type;
624 /* The "X-Forwarded-For:" header sent by the client */
628 * Nonzero if this client is processing data.
629 * Set to zero when the thread associated with this structure dies.
633 /* files associated with this client */
634 struct file_list *actions_list;
636 struct file_list *rlist; /* pcrs job file */
637 size_t content_length; /* Length after content modification */
640 struct file_list *tlist; /* trustfile */
641 #endif /* def FEATURE_TRUST */
643 struct client_state *next;
651 char *(*parser)(const struct parsers *, const char *, struct client_state *);
654 struct cgi_dispatcher
658 int (*handler)(struct client_state *csp, struct http_response *rsp, struct map *parameters);
659 const char *description;
665 * this is a pointer to the data structures associated with the file.
666 * Read-only once the structure has been created.
670 /* Normally NULL. When we are finished with file (i.e. when we have
671 * loaded a new one), set to a pointer to an unloader function.
672 * Unloader will be called by sweep() (called from main loop) when
673 * all clients using this file are done. This prevents threading
676 void (*unloader)(void *);
678 /* Used internally by sweep(). Do not access from elsewhere. */
681 /* Following variables allow us to check if file has been changed.
682 * Read-only once the structure has been created.
687 /* Pointer to next entry in the linked list of all "file_list"s.
688 * This linked list is so that sweep() can navigate it.
689 * Since sweep() can remove items from the list, we must be careful
690 * to only access this value from main thread (when we know sweep
693 struct file_list *next;
700 struct url_spec url[1];
702 struct block_spec *next;
704 #endif /* def FEATURE_TRUST */
707 #define SOCKS_NONE 0 /* Don't use a SOCKS server */
708 #define SOCKS_4 40 /* original SOCKS 4 protocol */
709 #define SOCKS_4A 41 /* as modified for hosts w/o external DNS */
713 struct url_spec url[1];
715 /* Connection type - must be a SOCKS_xxx constant */
722 /* Parent HTTP proxy */
726 /* For the linked list */
727 struct forward_spec *next;
729 #define FORWARD_SPEC_INITIALIZER { { URL_SPEC_INITIALIZER }, 0, NULL, 0, NULL, 0, NULL }
732 struct re_filterfile_spec
736 struct list patterns[1];
741 #define ACL_PERMIT 1 /* accept connection request */
742 #define ACL_DENY 2 /* reject connection request */
744 struct access_control_addr
751 struct access_control_list
753 struct access_control_addr src[1];
754 struct access_control_addr dst[1];
757 struct access_control_list *next;
759 #endif /* def FEATURE_ACL */
762 /* Maximum number of loaders (actions, re_filter, ...) */
766 * Data loaded from the configuration file.
768 * (Anomaly: toggle is still handled through a global, not this structure)
770 struct configuration_spec
779 const char *actions_file;
781 /* The administrator's email address */
784 /* A URL with info on this proxy */
785 char *proxy_info_url;
787 const char *re_filterfile;
789 #ifdef FEATURE_COOKIE_JAR
790 const char * jarfile;
792 #endif /* def FEATURE_COOKIE_JAR */
795 * Port and IP to bind to.
796 * Defaults to HADDR_DEFAULT:HADDR_PORT == 127.0.0.1:8000
802 const char * trustfile;
804 struct list trust_info[1];
805 struct url_spec *trust_list[64];
806 #endif /* def FEATURE_TRUST */
809 struct access_control_list *acl;
810 #endif /* def FEATURE_ACL */
812 struct forward_spec *forward;
814 /* All options from the config file, HTML-formatted */
817 /* the configuration file object. */
818 struct file_list *config_file_list;
820 /* List of loaders */
821 int (*loaders[NLOADERS])(struct client_state *);
823 /* bool, nonzero if we need to bind() to the new port */
828 #define SZ(X) (sizeof(X) / sizeof(*X))
830 #ifdef FEATURE_FORCE_LOAD
831 #define FORCE_PREFIX "/IJB-FORCE-LOAD"
832 #endif /* def FEATURE_FORCE_LOAD */
835 #define HOME_PAGE_URL "http://ijbswa.sourceforge.net"
836 #define REDIRECT_URL HOME_PAGE_URL "/redirect.php?v=" VERSION "&to="
837 #define CGI_PREFIX_HOST "i.j.b"
840 static const char CSUCCEED[] =
841 "HTTP/1.0 200 Connection established\n"
842 "Proxy-Agent: IJ/" VERSION "\n\n";
844 static const char CHEADER[] =
845 "HTTP/1.0 400 Invalid header received from browser\n\n";
851 #endif /* ndef PROJECT_H_INCLUDED */