1 #ifndef PROJECT_H_INCLUDED
2 #define PROJECT_H_INCLUDED
3 #define PROJECT_H_VERSION "$Id: project.h,v 1.31 2001/09/16 13:20:29 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 1.31 2001/09/16 13:20:29 jongfoster
40 * Rewrite of list library. Now has seperate header and list_entry
41 * structures. Also added a large sprinking of assert()s to the list
44 * Revision 1.30 2001/09/13 23:52:00 jongfoster
45 * Support for both static and dynamically generated CGI pages
47 * Revision 1.29 2001/09/13 23:29:43 jongfoster
48 * Defining FORWARD_SPEC_INITIALIZER
50 * Revision 1.28 2001/09/13 23:05:50 jongfoster
51 * Changing the string paramater to the header parsers a "const".
53 * Revision 1.27 2001/08/05 16:06:20 jongfoster
54 * Modifiying "struct map" so that there are now separate header and
55 * "map_entry" structures. This means that functions which modify a
56 * map no longer need to return a pointer to the modified map.
57 * Also, it no longer reverses the order of the entries (which may be
58 * important with some advanced template substitutions).
60 * Revision 1.26 2001/07/30 22:08:36 jongfoster
61 * Tidying up #defines:
62 * - All feature #defines are now of the form FEATURE_xxx
63 * - Permanently turned off WIN_GUI_EDIT
64 * - Permanently turned on WEBDAV and SPLIT_PROXY_ARGS
66 * Revision 1.25 2001/07/29 18:43:08 jongfoster
67 * Changing #ifdef _FILENAME_H to FILENAME_H_INCLUDED, to conform to
70 * Revision 1.24 2001/07/25 17:20:27 oes
71 * Introduced http->user_agent
73 * Revision 1.23 2001/07/18 12:32:23 oes
74 * - Added ACTION_STRING_DEANIMATE
75 * - moved #define freez from jcc.h to project.h
77 * Revision 1.22 2001/07/15 17:51:41 jongfoster
78 * Renaming #define STATIC to STATIC_PCRE
80 * Revision 1.21 2001/07/13 14:03:19 oes
81 * - Reorganized regex header inclusion and #defines to
82 * comply to the scheme in configure.in
83 * - Added csp->content_type and its CT_* keys
84 * - Added ACTION_DEANIMATE
85 * - Removed all #ifdef PCRS
87 * Revision 1.20 2001/06/29 21:45:41 oes
88 * Indentation, CRLF->LF, Tab-> Space
90 * Revision 1.19 2001/06/29 13:33:36 oes
92 * - Introduced http_request.host_ip_addr_str
93 * - Introduced http_response.head_length
94 * - Introduced config.my_ip_addr_str, config.my_hostname,
95 * config.admin_address and config.proxy_info_url
96 * - Removed config.proxy_args_header and config.proxy_args_trailer,
97 * renamed config.proxy_args_invocation to config.proxy_args
98 * - Removed HTML snipplets and GIFs
99 * - Removed logentry from cancelled commit
101 * Revision 1.18 2001/06/09 10:57:39 jongfoster
102 * Adding definition of BUFFER_SIZE.
103 * Changing struct cgi_dispatcher to use "const" strings.
105 * Revision 1.17 2001/06/07 23:15:09 jongfoster
106 * Merging ACL and forward files into config file.
107 * Moving struct gateway members into struct forward_spec
108 * Removing config->proxy_args_gateways
109 * Cosmetic: Adding a few comments
111 * Revision 1.16 2001/06/04 18:31:58 swa
112 * files are now prefixed with either `confdir' or `logdir'.
113 * `make redhat-dist' replaces both entries confdir and logdir
116 * Revision 1.15 2001/06/04 11:28:53 swa
117 * redirect did not work due to missing /
119 * Revision 1.14 2001/06/03 11:03:48 oes
121 * added struct http_response,
122 * changed struct interceptors to struct cgi_dispatcher,
123 * moved HTML stuff to cgi.h
125 * Revision 1.13 2001/06/01 20:05:36 jongfoster
126 * Support for +image-blocker{}: added ACTION_IMAGE_BLOCKER
127 * constant, and removed csp->tinygif.
129 * Revision 1.12 2001/06/01 18:49:17 jongfoster
130 * Replaced "list_share" with "list" - the tiny memory gain was not
131 * worth the extra complexity.
133 * Revision 1.11 2001/06/01 10:32:47 oes
134 * Added constants for anchoring selection bitmap
136 * Revision 1.10 2001/05/31 21:33:53 jongfoster
137 * Changes for new actions file, replacing permissionsfile
138 * and parts of the config file. Also added support for
141 * Revision 1.9 2001/05/31 17:32:31 oes
143 * - Enhanced domain part globbing with infix and prefix asterisk
144 * matching and optional unanchored operation
146 * Revision 1.8 2001/05/29 20:09:15 joergs
147 * HTTP_REDIRECT_TEMPLATE fixed.
149 * Revision 1.7 2001/05/29 09:50:24 jongfoster
150 * Unified blocklist/imagelist/actionslist.
151 * File format is still under discussion, but the internal changes
154 * Also modified interceptor behaviour:
155 * - We now intercept all URLs beginning with one of the following
156 * prefixes (and *only* these prefixes):
158 * * http://ijbswa.sf.net/config/
159 * * http://ijbswa.sourceforge.net/config/
160 * - New interceptors "home page" - go to http://i.j.b/ to see it.
161 * - Internal changes so that intercepted and fast redirect pages
162 * are not replaced with an image.
163 * - Interceptors now have the option to send a binary page direct
164 * to the client. (i.e. ijb-send-banner uses this)
165 * - Implemented show-url-info interceptor. (Which is why I needed
166 * the above interceptors changes - a typical URL is
167 * "http://i.j.b/show-url-info?url=www.somesite.com/banner.gif".
168 * The previous mechanism would not have intercepted that, and
169 * if it had been intercepted then it then it would have replaced
172 * Revision 1.6 2001/05/27 22:17:04 oes
174 * - re_process_buffer no longer writes the modified buffer
175 * to the client, which was very ugly. It now returns the
176 * buffer, which it is then written by chat.
178 * - content_length now adjusts the Content-Length: header
179 * for modified documents rather than crunch()ing it.
180 * (Length info in csp->content_length, which is 0 for
181 * unmodified documents)
183 * - For this to work, sed() is called twice when filtering.
185 * Revision 1.5 2001/05/26 00:28:36 jongfoster
186 * Automatic reloading of config file.
187 * Removed obsolete SIGHUP support (Unix) and Reload menu option (Win32).
188 * Most of the global variables have been moved to a new
189 * struct configuration_spec, accessed through csp->config->globalname
190 * Most of the globals remaining are used by the Win32 GUI.
192 * Revision 1.4 2001/05/22 18:46:04 oes
194 * - Enabled filtering banners by size rather than URL
195 * by adding patterns that replace all standard banner
196 * sizes with the "Junkbuster" gif to the re_filterfile
198 * - Enabled filtering WebBugs by providing a pattern
199 * which kills all 1x1 images
201 * - Added support for PCRE_UNGREEDY behaviour to pcrs,
202 * which is selected by the (nonstandard and therefore
203 * capital) letter 'U' in the option string.
204 * It causes the quantifiers to be ungreedy by default.
205 * Appending a ? turns back to greedy (!).
207 * - Added a new interceptor ijb-send-banner, which
208 * sends back the "Junkbuster" gif. Without imagelist or
209 * MSIE detection support, or if tinygif = 1, or the
210 * URL isn't recognized as an imageurl, a lame HTML
211 * explanation is sent instead.
213 * - Added new feature, which permits blocking remote
214 * script redirects and firing back a local redirect
216 * The feature is conditionally compiled, i.e. it
217 * can be disabled with --disable-fast-redirects,
218 * plus it must be activated by a "fast-redirects"
219 * line in the config file, has its own log level
220 * and of course wants to be displayed by show-proxy-args
221 * Note: Boy, all the #ifdefs in 1001 locations and
222 * all the fumbling with configure.in and acconfig.h
223 * were *way* more work than the feature itself :-(
225 * - Because a generic redirect template was needed for
226 * this, tinygif = 3 now uses the same.
228 * - Moved GIFs, and other static HTTP response templates
233 * - Removed some >400 CRs again (Jon, you really worked
236 * Revision 1.3 2001/05/20 01:21:20 jongfoster
237 * Version 2.9.4 checkin.
238 * - Merged popupfile and cookiefile, and added control over PCRS
239 * filtering, in new "actionsfile".
240 * - Implemented LOG_LEVEL_FATAL, so that if there is a configuration
241 * file error you now get a message box (in the Win32 GUI) rather
242 * than the program exiting with no explanation.
243 * - Made killpopup use the PCRS MIME-type checking and HTTP-header
245 * - Removed tabs from "config"
246 * - Moved duplicated url parsing code in "loaders.c" to a new funcition.
247 * - Bumped up version number.
249 * Revision 1.2 2001/05/17 23:01:01 oes
250 * - Cleaned CRLF's from the sources and related files
252 * Revision 1.1.1.1 2001/05/15 13:59:03 oes
253 * Initial import of version 2.9.3 source tree
256 *********************************************************************/
259 /* Declare struct FILE for vars and funcs. */
262 /* Need time_t for file_list */
266 * Include appropriate regular expression libraries.
267 * Note that pcrs and pcre (native) are needed for cgi
268 * and are included anyway.
271 #if defined(REGEX_PCRE) || defined (REGEX_GNU)
273 #endif /* defined(REGEX_PCRE) || defined (REGEX_GNU) */
287 #if defined(REGEX_PCRE)
289 # include "pcreposix.h"
291 # include <pcreposix.h>
293 #endif /* defined(REGEX_PCRE) */
295 #if defined(REGEX_GNU)
296 # include "gnu_regex.h"
301 #endif /* def AMIGA */
307 #define freez(X) if(X) free(X); X = NULL
309 #define BUFFER_SIZE 5000
313 /* Default IP and port to listen on */
314 #define HADDR_DEFAULT "127.0.0.1"
315 #define HADDR_PORT 8000
318 /* Forward defs for various structures */
320 /* Need this for struct client_state */
321 struct configuration_spec;
324 /* Generic linked list of strings */
329 struct list_entry *next;
334 struct list_entry *first;
335 struct list_entry *last;
339 /* A map from a string to another string */
345 struct map_entry *next;
350 struct map_entry *first;
351 struct map_entry *last;
360 char *host_ip_addr_str; /* NULL before connect_to() */
364 char *hostport; /* "host[:port]" */
366 char *user_agent; /* Client's User-Agent: header value */
369 /* Response generated by CGI, blocker, or error handler */
372 char *status; /* HTTP status (string)*/
373 struct list headers[1]; /* List of header lines */
374 char *head; /* Formatted http response head */
375 int head_length; /* Length of http response head */
376 char *body; /* HTTP document body */
377 int content_length; /* Length of body, REQUIRED if binary body */
378 int is_static; /* Nonzero if the content will never change and
379 * should be cached by the broser (e.g. images) */
385 char *spec; /* The string which was parsed to produce this */
386 /* url_spec. Used for debugging or display only. */
388 /* Hostname matching: */
389 char *domain; /* Fully qalified domain name (FQDN) pattern. */
390 /* May contain "*". */
391 char *dbuf; /* Buffer with '\0'-delimited fqdn */
392 char **dvec; /* Domain ptr vector into dbuf */
393 int dcnt; /* How many domains in fqdn? */
394 int unanchored; /* Bitmap - flags are ANCHOR_LEFT and ANCHOR_RIGHT */
397 int port; /* The port number, or 0 to match all ports. */
400 char *path; /* The path prefix (if not using regex), or source */
402 int pathlen; /* ==strlen(path). Needed for prefix matching. */
404 regex_t *preg; /* Regex for matching path part */
408 #define URL_SPEC_INITIALIZER { NULL, NULL, NULL, NULL, 0, 0, 0, NULL, 0, NULL }
409 #else /* ifndef REGEX */
410 #define URL_SPEC_INITIALIZER { NULL, NULL, NULL, NULL, 0, 0, 0, NULL, 0 }
411 #endif /* ndef REGEX */
413 /* Constants for host part matching in URLs */
414 #define ANCHOR_LEFT 1
415 #define ANCHOR_RIGHT 2
427 #define IOB_PEEK(CSP) ((CSP->iob->cur > CSP->iob->eod) ? (CSP->iob->eod - CSP->iob->cur) : 0)
428 #define IOB_RESET(CSP) if(CSP->iob->buf) free(CSP->iob->buf); memset(CSP->iob, '\0', sizeof(CSP->iob));
430 /* Keys for csp->content_type */
431 #define CT_TEXT 0x01U
434 #define ACTION_MASK_ALL (~0U)
436 #define ACTION_MOST_COMPATIBLE 0x0000U
438 #define ACTION_BLOCK 0x0001U
439 #define ACTION_DEANIMATE 0x2000U
440 #define ACTION_FAST_REDIRECTS 0x0002U
441 #define ACTION_FILTER 0x0004U
442 #define ACTION_HIDE_FORWARDED 0x0008U
443 #define ACTION_HIDE_FROM 0x0010U
444 #define ACTION_HIDE_REFERER 0x0020U /* sic - follow HTTP, not English */
445 #define ACTION_HIDE_USER_AGENT 0x0040U
446 #define ACTION_IMAGE 0x0080U
447 #define ACTION_IMAGE_BLOCKER 0x0100U
448 #define ACTION_NO_COOKIE_READ 0x0200U
449 #define ACTION_NO_COOKIE_SET 0x0400U
450 #define ACTION_NO_POPUPS 0x0800U
451 #define ACTION_VANILLA_WAFER 0x1000U
453 #define ACTION_STRING_DEANIMATE 0
454 #define ACTION_STRING_FROM 1
455 #define ACTION_STRING_IMAGE_BLOCKER 2
456 #define ACTION_STRING_REFERER 3
457 #define ACTION_STRING_USER_AGENT 4
458 #define ACTION_STRING_COUNT 5
461 #define ACTION_MULTI_ADD_HEADER 0
462 #define ACTION_MULTI_WAFER 1
463 #define ACTION_MULTI_COUNT 2
466 * This structure contains a list of actions to apply to a URL.
467 * It only contains positive instructions - no "-" options.
468 * It is not used to store the actions list itself, only for
469 * url_actions() to return the current values.
471 struct current_action_spec
473 unsigned flags; /* a bit set to "1" = add action */
475 /* For those actions that require parameters: */
477 /* each entry is valid if & only if corresponding entry in "add" set. */
478 char * string[ACTION_STRING_COUNT];
481 struct list multi[ACTION_MULTI_COUNT][1];
486 * This structure contains a set of changes to actions.
487 * It can contain both positive and negative instructions.
488 * It is used to store an entry in the actions list.
492 unsigned mask; /* a bit set to "0" = remove action */
493 unsigned add; /* a bit set to "1" = add action */
495 /* For those actions that require parameters: */
497 /* each entry is valid if & only if corresponding entry in "add" set. */
498 char * string[ACTION_STRING_COUNT];
500 /* Strings to remove. */
501 struct list multi_remove[ACTION_MULTI_COUNT][1];
503 /* If nonzero, remove *all* strings. */
504 int multi_remove_all[ACTION_MULTI_COUNT];
507 struct list multi_add[ACTION_MULTI_COUNT][1];
511 * This structure is used to store the actions list.
513 * It contains a URL pattern, and the chages to the actions.
514 * It is a linked list.
518 struct url_spec url[1];
520 struct action_spec action[1];
522 struct url_actions * next;
526 /* Constants defining bitmask for csp->accept_types */
528 #ifdef FEATURE_IMAGE_DETECT_MSIE
530 /* MSIE detected by user-agent string */
531 #define ACCEPT_TYPE_IS_MSIE 0x0001
534 * *If* this is MSIE, it wants an image. (Or this is a shift-reload, or
535 * it's got an image from this URL before... yuck!)
536 * Only meaningful if ACCEPT_TYPE_IS_MSIE set
538 #define ACCEPT_TYPE_MSIE_IMAGE 0x0002
541 * *If* this is MSIE, it wants a HTML document.
542 * Only meaningful if ACCEPT_TYPE_IS_MSIE set
544 #define ACCEPT_TYPE_MSIE_HTML 0x0004
546 #endif /* def FEATURE_IMAGE_DETECT_MSIE */
551 /* The proxy's configuration */
552 struct configuration_spec * config;
554 /* The actions to perform on the current request */
555 struct current_action_spec action[1];
557 /* socket to talk to client (web browser) */
560 /* socket to talk to server (web server or proxy) */
564 #ifdef FEATURE_STATISTICS
565 /* 1 if this URL was rejected, 0 otherwise. Allows actual stats inc to
566 * occur in main thread only for thread-safety.
569 #endif /* def FEATURE_STATISTICS */
571 #ifdef FEATURE_FORCE_LOAD
573 #endif /* def FEATURE_FORCE_LOAD */
575 #ifdef FEATURE_TOGGLE
577 #endif /* def FEATURE_TOGGLE */
580 * Client PC's IP address, as reported by the accept()_ function.
581 * Both as string and number
587 /* Our IP address and hostname, i.e. the IP address that
588 the client used to reach us, and the associated hostname,
591 char *my_ip_addr_str;
595 /* The referer in this request, if one was specified. */
597 #endif /* def FEATURE_TRUST */
599 #if defined(FEATURE_IMAGE_DETECT_MSIE)
600 /* Types the client will accept.
601 * Bitmask - see ACCEPT_TYPE_XXX constants.
604 #endif /* defined(FEATURE_IMAGE_DETECT_MSIE) */
606 /* The URL that was requested */
607 struct http_request http[1];
609 /* An I/O buffer used for buffering data read from the client */
612 /* List of all headers for this request */
613 struct list headers[1];
615 /* List of all cookies for this request */
616 struct list cookie_list[1];
618 /* MIME-Type bitmap, see CT_* above */
619 unsigned char content_type;
621 /* The "X-Forwarded-For:" header sent by the client */
625 * Nonzero if this client is processing data.
626 * Set to zero when the thread associated with this structure dies.
630 /* files associated with this client */
631 struct file_list *actions_list;
633 struct file_list *rlist; /* pcrs job file */
634 size_t content_length; /* Length after content modification */
637 struct file_list *tlist; /* trustfile */
638 #endif /* def FEATURE_TRUST */
640 struct client_state *next;
648 char *(*parser)(const struct parsers *, const char *, struct client_state *);
651 struct cgi_dispatcher
655 int (*handler)(struct client_state *csp, struct http_response *rsp, struct map *parameters);
656 const char *description;
662 * this is a pointer to the data structures associated with the file.
663 * Read-only once the structure has been created.
667 /* Normally NULL. When we are finished with file (i.e. when we have
668 * loaded a new one), set to a pointer to an unloader function.
669 * Unloader will be called by sweep() (called from main loop) when
670 * all clients using this file are done. This prevents threading
673 void (*unloader)(void *);
675 /* Used internally by sweep(). Do not access from elsewhere. */
678 /* Following variables allow us to check if file has been changed.
679 * Read-only once the structure has been created.
684 /* Pointer to next entry in the linked list of all "file_list"s.
685 * This linked list is so that sweep() can navigate it.
686 * Since sweep() can remove items from the list, we must be careful
687 * to only access this value from main thread (when we know sweep
690 struct file_list *next;
697 struct url_spec url[1];
699 struct block_spec *next;
701 #endif /* def FEATURE_TRUST */
704 #define SOCKS_NONE 0 /* Don't use a SOCKS server */
705 #define SOCKS_4 40 /* original SOCKS 4 protocol */
706 #define SOCKS_4A 41 /* as modified for hosts w/o external DNS */
710 struct url_spec url[1];
712 /* Connection type - must be a SOCKS_xxx constant */
719 /* Parent HTTP proxy */
723 /* For the linked list */
724 struct forward_spec *next;
726 #define FORWARD_SPEC_INITIALIZER { { URL_SPEC_INITIALIZER }, 0, NULL, 0, NULL, 0, NULL }
729 struct re_filterfile_spec
733 struct list patterns[1];
738 #define ACL_PERMIT 1 /* accept connection request */
739 #define ACL_DENY 2 /* reject connection request */
741 struct access_control_addr
748 struct access_control_list
750 struct access_control_addr src[1];
751 struct access_control_addr dst[1];
754 struct access_control_list *next;
756 #endif /* def FEATURE_ACL */
759 /* Maximum number of loaders (actions, re_filter, ...) */
763 * Data loaded from the configuration file.
765 * (Anomaly: toggle is still handled through a global, not this structure)
767 struct configuration_spec
776 const char *actions_file;
778 /* The administrator's email address */
781 /* A URL with info on this proxy */
782 char *proxy_info_url;
784 const char *re_filterfile;
786 #ifdef FEATURE_COOKIE_JAR
787 const char * jarfile;
789 #endif /* def FEATURE_COOKIE_JAR */
792 * Port and IP to bind to.
793 * Defaults to HADDR_DEFAULT:HADDR_PORT == 127.0.0.1:8000
799 const char * trustfile;
801 struct list trust_info[1];
802 struct url_spec *trust_list[64];
803 #endif /* def FEATURE_TRUST */
806 struct access_control_list *acl;
807 #endif /* def FEATURE_ACL */
809 struct forward_spec *forward;
811 /* All options from the config file, HTML-formatted */
814 /* the configuration file object. */
815 struct file_list *config_file_list;
817 /* List of loaders */
818 int (*loaders[NLOADERS])(struct client_state *);
820 /* bool, nonzero if we need to bind() to the new port */
825 #define SZ(X) (sizeof(X) / sizeof(*X))
827 #ifdef FEATURE_FORCE_LOAD
828 #define FORCE_PREFIX "/IJB-FORCE-LOAD"
829 #endif /* def FEATURE_FORCE_LOAD */
832 #define HOME_PAGE_URL "http://ijbswa.sourceforge.net"
833 #define REDIRECT_URL HOME_PAGE_URL "/redirect.php?v=" VERSION "&to="
834 #define CGI_PREFIX_HOST "i.j.b"
837 static const char CSUCCEED[] =
838 "HTTP/1.0 200 Connection established\n"
839 "Proxy-Agent: IJ/" VERSION "\n\n";
841 static const char CHEADER[] =
842 "HTTP/1.0 400 Invalid header received from browser\n\n";
848 #endif /* ndef PROJECT_H_INCLUDED */