2 * IRC - Internet Relay Chat, common/match.c
3 * Copyright (C) 1990 Jarkko Oikarinen
5 * This program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 1, or (at your option)
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
15 * You should have received a copy of the GNU General Public License
16 * along with this program; if not, write to the Free Software
17 * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
24 #include "ircd_chattr.h"
28 * Written by Run (carlo@runaway.xs4all.nl), 25-10-96
31 * From: Carlo Wood <carlo@runaway.xs4all.nl>
32 * Message-Id: <199609021026.MAA02393@runaway.xs4all.nl>
33 * Subject: [C-Com] Analysis for `mmatch' (was: gline4 problem)
34 * To: coder-com@mail.undernet.org (coder committee)
35 * Date: Mon, 2 Sep 1996 12:26:01 +0200 (MET DST)
37 * We need a new function `mmatch(const char *old_mask, const char *new_mask)'
38 * which returns `true' likewise the current `match' (start with copying it),
39 * but which treats '*' and '?' in `new_mask' differently (not "\*" and "\?" !)
40 * as follows: a '*' in `new_mask' does not match a '?' in `old_mask' and
41 * a '?' in `new_mask' does not match a '\?' in `old_mask'.
42 * And ofcourse... a '*' in `new_mask' does not match a '\*' in `old_mask'...
43 * And last but not least, '\?' and '\*' in `new_mask' now become one character.
46 int mmatch(const char *old_mask, const char *new_mask)
48 const char *m = old_mask;
49 const char *n = new_mask;
70 for (m--; (m > old_mask) && (*m == '?'); m--)
72 if ((*m == '*') && (m > old_mask) && (m[-1] != '\\'))
78 /* Added to `mmatch' : Because '\?' and '\*' now is one character: */
79 if ((*na == '\\') && ((na[1] == '*') || (na[1] == '?')))
90 if ((*m == '\\') && ((m[1] == '*') || (m[1] == '?')))
98 /* Added to `mmatch' : Because '\?' and '\*' now is one character: */
99 if ((*n == '\\') && ((n[1] == '*') || (n[1] == '?')))
108 * This `if' has been changed compared to match() to do the following:
110 * old (m) new (n) boolean expression
111 * * any (*m == '*' && !mq) ||
112 * ? any except '*' (*m == '?' && !mq && (*n != '*' || nq)) ||
113 * any except * or ? same as m (!((*m == '*' || *m == '?') && !mq) &&
114 * ToLower(*m) == ToLower(*n) &&
115 * !((mq && !nq) || (!mq && nq)))
117 * Here `any' also includes \* and \? !
119 * After reworking the boolean expressions, we get:
120 * (Optimized to use boolean shortcircuits, with most frequently occuring
121 * cases upfront (which took 2 hours!)).
123 if ((*m == '*' && !mq) ||
124 ((!mq || nq) && ToLower(*m) == ToLower(*n)) ||
125 (*m == '?' && !mq && (*n != '*' || nq)))
138 /* Added to `mmatch' : Because '\?' and '\*' now is one character: */
139 if ((*na == '\\') && ((na[1] == '*') || (na[1] == '?')))
148 * Compare if a given string (name) matches the given
149 * mask (which can contain wild cards: '*' - match any
150 * number of chars, '?' - match any single character.
159 * Rewritten by Andrea Cocito (Nemesi), November 1998.
163 /****************** Nemesi's match() ***************/
165 int match(const char *mask, const char *string)
167 const char *m = mask, *s = string;
169 const char *bm, *bs; /* Will be reg anyway on a decent CPU/compiler */
171 /* Process the "head" of the mask, if any */
172 while ((ch = *m++) && (ch != '*'))
176 if (*m == '?' || *m == '*')
179 if (ToLower(*s) != ToLower(ch))
188 /* We got a star: quickly find if/where we match the next char */
190 bm = m; /* Next try rollback here */
199 continue; /* while */
201 if (*m == '?' || *m == '*')
204 goto break_while; /* C is structured ? */
208 return 0; /* mask ends with '*', we got it */
210 while (ToLower(*s++) != ch)
213 bs = s; /* Next try start from here */
215 /* Check the rest of the "chunk" */
223 if (*m == '?' || *m == '*')
226 if (ToLower(*s) != ToLower(ch))
248 * Collapse a pattern string into minimal components.
249 * This particular version is "in place", so that it changes the pattern
250 * which is to be reduced to a "minimal" size.
252 * (C) Carlo Wood - 6 Oct 1998
253 * Speedup rewrite by Andrea Cocito, December 1998.
254 * Note that this new optimized alghoritm can *only* work in place.
257 char *collapse(char *mask)
267 if ((*m == '*') && ((m[1] == '*') || (m[1] == '?')))
276 if (star && (*m != '?'))
282 if ((*m == '\\') && ((m[1] == '*') || (m[1] == '?')))
291 if ((*m == '\\') && ((m[1] == '*') || (m[1] == '?')))
301 ***************** Nemesi's matchcomp() / matchexec() **************
304 /* These functions allow the use of "compiled" masks, you compile a mask
305 * by means of matchcomp() that gets the plain text mask as input and writes
306 * its result in the memory locations addressed by the 3 parameters:
307 * - *cmask will contain the text of the compiled mask
308 * - *minlen will contain the lenght of the shortest string that can match
310 * - *charset will contain the minimal set of chars needed to match the mask
311 * You can pass NULL as *charset and it will be simply not returned, but you
312 * MUST pass valid pointers for *minlen and *cmask (wich must be big enough
313 * to contain the compiled mask text that is in the worst case as long as the
314 * text of the mask itself in plaintext format) and the return value of
315 * matchcomp() will be the number of chars actually written there (excluded
316 * the trailing zero). cmask can be == mask, matchcomp() can work in place.
317 * The {cmask, minlen} couple of values make the real compiled mask and
318 * need to be passed to the functions that use the compiled mask, if you pass
319 * the wrong minlen or something wrong in cmask to one of these expect a
320 * coredump. This means that when you record a compiled mask you must store
321 * *both* these values.
322 * Once compiled the mask can be used to match a string by means of
323 * matchexec(), it can be printed back to human-readable format by means
324 * of sprintmatch() or it can be compared to another compiled mask by means
325 * of mmexec() that will tell if it completely overrides that mask (a lot like
326 * what mmatch() does for plain text masks).
327 * You can gain a lot of speed in many situations avoiding to matchexec() when:
328 * - The maximum lenght of the field you are about to match() the mask to is
329 * shorter than minlen, in example when matching abc*def*ghil with a nick:
330 * It just cannot match since a nick is at most 9 chars long and the mask
331 * needs at least 10 chars (10 will be the value returned in minlen).
332 * - The charset allowed for the field you are about to match to doesn't
333 * "contain" the charset returned by matchcomp(), in example when you
334 * have *.* as mask it makes no sense to try to match it against a nick
335 * because, again, a nick can't contain a '.', you can check this with
336 * a simple (charset & NTL_IRCNK) in this case.
337 * - As a special case, since compiled masks are forced to lowercase,
338 * it would make no sense to use the NTL_LOWER and NTL_UPPER on a compiled
339 * mask, thus they are reused as follows: if the NTL_LOWER bit of charset
340 * is set it means that the mask contains only non-wilds chars (i.e. you can
341 * use strCasecmp() to match it or a direct hash lookup), if the NTL_UPPER
342 * bit is set it means that it contains only wild chars (and you can
343 * match it with strlen(field)>=minlen).
344 * Do these optimizations ONLY when the data you are about to pass to
345 * matchexec() are *known* to be invalid in advance, using strChattr()
346 * or strlen() on the text would be slower than calling matchexec() directly
348 * Internally a compiled mask contain in the *cmask area the text of
349 * the plain text form of the mask itself with applied the following hacks:
350 * - All characters are forced to lowercase (so that uppercase letters and
351 * specifically the symbols 'A' and 'Z' are reserved for special use)
352 * - All non-escaped stars '*' are replaced by the letter 'Z'
353 * - All non-escaped question marks '?' are replaced by the letter 'A'
354 * - All escape characters are removed, the wilds escaped by them are
355 * then passed by without the escape since they don't collide anymore
356 * with the real wilds (encoded as A/Z)
357 * - Finally the part of the mask that follows the last asterisk is
358 * reversed (byte order mirroring) and moved right after the first
360 * After all this a mask like: Head*CHUNK1*chu\*nK2*ch??k3*TaIl
361 * .... becomes: headZliatZchunk1Zchu*nk2ZchAAk3
362 * This can still be printed on a console, more or less understood by an
363 * human and handled with the usual str*() library functions.
364 * When you store somewhere the compiled mask you can avoid storing the
365 * textform of it since it can be "decompiled" by means of sprintmatch(),
366 * but at that time the following things are changed in the mask:
367 * - All chars have been forced to lowercase.
368 * - The mask is collapsed.
369 * The balance point of using compiled masks in terms of CPU is when you expect
370 * to use matchexec() instead of match() at least 20 times on the same mask
371 * or when you expect to use mmexec() instead of mmatch() 3 times.
377 * Compiles a mask into a form suitable for using in matchexec().
380 int matchcomp(char *cmask, int *minlen, int *charset, const char *mask)
382 const char *m = mask;
387 int l1, l2, lmin, loop, sign;
392 int chset2 = (NTL_LOWER | NTL_UPPER);
404 chset2 &= ~NTL_LOWER;
407 if ((*m == '?') || (*m == '*'))
415 chset2 &= ~NTL_LOWER;
420 chset &= IRCD_CharAttrTab[*b++ - CHAR_MIN];
421 chset2 &= ~NTL_UPPER;
425 *charset = (chset | chset2);
436 for (x1 = ls + 1, x2 = (b - 1); x1 < x2; x1++, x2--)
445 while ((lmin = (l1 < l2) ? l1 : l2))
448 for (loop = 0; loop < lmin; loop++)
456 l1 -= (sign < 0) ? 0 : lmin;
457 l2 -= (sign > 0) ? 0 : lmin;
470 * Executes a match with a mask previosuly compiled with matchcomp()
471 * Note 1: If the mask isn't correctly produced by matchcomp() I will core
472 * Note 2: 'min' MUST be the value returned by matchcomp on that mask,
473 * or.... I will core even faster :-)
474 * Note 3: This piece of code is not intended to be nice but efficient.
477 int matchexec(const char *string, const char *cmask, int minlen)
479 const char *s = string - 1;
480 const char *b = cmask - 1;
486 while ((ToLower(*++s) == *++b) && *s);
488 return ((*b != '\0') && ((*b++ != 'Z') || (*b != '\0')));
499 if ((trash = (s - string - minlen)) < 0)
503 while ((ToLower(*--s) == *++b) && *b && (ToLower(*--s) == *++b) && *b
504 && (ToLower(*--s) == *++b) && *b && (ToLower(*--s) == *++b) && *b);
517 while ((ToLower(*++s) != ch))
523 while ((ToLower(*++s) == *++b) && *b);
546 * Prints the human readable version of *cmask into *mask, (decompiles
548 * The area pointed by *mask MUST be big enough (the mask might be up to
549 * twice the size of its compiled form if it's made all of \? or \*, and
550 * this function can NOT work in place since it might enflate the mask)
551 * The printed mask is not identical to the one that was compiled to cmask,
552 * infact it is 1) forced to all lowercase, 2) collapsed, both things
553 * are supposed to NOT change it's meaning.
554 * It returns the number of chars actually written to *mask;
557 int matchdecomp(char *mask, const char *cmask)
560 const char *rcm = cmask;
561 const char *begtail, *endtail;
569 for (; (*rcm != 'Z'); rcm++, rtb++)
571 if ((*rcm == '?') || (*rcm == '*'))
573 if (!((*rtb = ((*rcm == 'A') ? '?' : *rcm))))
580 while (*rcm && (*rcm != 'Z'))
605 for (rcm = endtail; (--rcm) > begtail; *rtb++ = ((*rcm == 'A') ? '?' : *rcm))
606 if ((*rcm == '?') || (*rcm == '*'))
615 * Checks if a wider compiled mask (wcm/wminlen) completely overrides
616 * a more restrict one (rcm/rminlen), basically what mmatch() does for
617 * non-compiled masks, returns 0 if the override is true (like mmatch()).
618 * "the wider overrides the restrict" means that any string that matches
619 * the restrict one _will_ also match the wider one, always.
620 * In this we behave differently from mmatch() because in example we return
621 * true for " a?*cd overrides a*bcd " for wich the override happens for how
622 * we literally defined it, here mmatch() would have returned false.
623 * The original concepts and the base alghoritm are copied from mmatch()
624 * written by Run (Carlo Wood), this function is written by
625 * Nemesi (Andrea Cocito)
628 int mmexec(const char *wcm, int wminlen, const char *rcm, int rminlen)
630 const char *w, *r, *br, *bw, *rx, *rz;
633 /* First of all rm must have enough non-stars to 'contain' wm */
634 if ((trash = rminlen - wminlen) < 0)
640 /* Let's start the game, remember that '*' is mapped to 'Z', '?'
641 is mapped to 'A' and that head?*??*?chunk*???*tail becomes
642 headAAAAZliatAAAZchunk for compiled masks */
644 /* Match the head of wm with the head of rm */
645 for (; (*r) && (*r != 'Z') && ((*w == *r) || (*w == 'A')); r++, w++);
647 while (*w == 'A') /* Eat extra '?' before '*' in wm if got '*' in rm */
649 if (*w != 'Z') /* head1<any>.. can't match head2<any>.. */
650 return ((*w) || (*r)) ? 1 : 0; /* and head<nul> matches only head<nul> */
652 return 0; /* headZ<nul> matches head<anything> */
654 /* Does rm have any stars in it ? let's check */
655 for (rx = r; *r && (*r != 'Z'); r++);
658 /* rm has no stars and thus isn't a mask but it's just a flat
659 string: special handling occurs here, note that eat must be 0 here */
664 for (; r--, (*w) && ((*w == *r) || (*w == 'A')); w++);
665 if (*w != 'Z') /* headZliat1<any> fails on head<any>2tail */
666 return (*w) ? 1 : 0; /* but headZliat<nul> matches head<any>tail */
669 /* match the chunks */
671 { /* This loop can't break but only return */
673 for (bw = w++; (*w != *rx); rx++) /* Seek the 1st char of the chunk */
674 if (--trash < 0) /* See if we can trash one more char of rm */
675 return 1; /* If not we can only fail of course */
676 for (r = ++rx, w++; (*w) && ((*w == *r) || (*w == 'A')); r++, w++);
677 if (!*w) /* Did last loop match the rest of chunk ? */
678 return 0; /* ... Yes, end of wm, matched ! */
680 { /* ... No, hitted non-star */
681 w = bw; /* Rollback at beginning of chunk */
682 if (--trash < 0) /* Trashed the char where this try started */
683 return 1; /* if we can't trash more chars fail */
687 rx = r; /* Successfully matched a chunk, move rx */
688 } /* and go on with the next one */
692 /* rm has at least one '*' and thus is a 'real' mask */
693 rz = r++; /* rx = unused of head, rz = beg-tail */
695 /* Match the tail of wm (if any) against the tail of rm */
698 for (; (*w) && (*r != 'Z') && ((*w == *r) || (*w == 'A')); w++, r++);
699 if (*r == 'Z') /* extra '?' before tail are fluff, just flush 'em */
702 if (*w != 'Z') /* We aren't matching a chunk, can't rollback */
706 /* Match the chunks of wm against what remains of the head of rm */
710 for (bw++; (rx < rz) && (*bw != *rx); rx++) /* Seek the first */
711 if (--trash < 0) /* waste some trash reserve */
713 if (!(rx < rz)) /* head finished */
715 for (bw++, (br = ++rx);
716 (br < rz) && (*bw) && ((*bw == *br) || (*bw == 'A')); br++, bw++);
717 if (!(br < rz)) /* Note that we didn't use any 'eat' char yet, if */
718 while (*bw == 'A') /* there were eat-en chars the head would be over */
719 bw++, eat++; /* Happens only at end of head, and eat is still 0 */
726 { /* If we failed because we got the end of head */
727 trash -= (br - rx); /* it makes no sense to rollback, just trash */
728 if (--trash < 0) /* all the rest of the head wich isn't long */
729 return 1; /* enough for this chunk and go out of this */
730 break; /* loop, then we try with the chunks of rm */
742 /* Match the unused chunks of wm against the chunks of rm */
744 for (; *r && (*r != 'Z'); r++);
751 while (eat && *r) /* the '?' we had eated make us skip as many chars */
752 if (*r++ != 'Z') /* here, but can't skip stars or trailing zero */
754 for (bw++; (*r) && (*bw != *r); r++)
755 if ((*r != 'Z') && (--trash < 0))
759 for ((br = ++r), bw++;
760 (*br) && (*br != 'Z') && ((*bw == *br) || (*bw == 'A')); br++, bw++);
769 if ((!*br) || (*r == 'Z'))
770 { /* If we hit the end of rm or a star in it */
771 trash -= (br - r); /* makes no sense to rollback within this */
772 if (trash < 0) /* same chunk of br, skip it all and then */
773 return 1; /* either rollback or break this loop if */
774 if (!*br) /* it was the end of rm */
789 /* match the remaining chunks of wm against what remains of the tail of rm */
790 r = rz - eat - 1; /* can't have <nul> or 'Z'within the tail, so just move r */
794 for (bw++; (*bw != *r); r--)
799 for ((br = --r), bw++;
800 (*bw) && (br >= rx) && ((*bw == *br) || (*bw == 'A')); br--, bw++);
816 return 1; /* Auch... something left out ? Fail */
821 * Compiles an IP mask into an in_mask structure
822 * The given <mask> can either be:
823 * - An usual irc type mask, containing * and or ?
824 * - An ip number plus a /bitnumber part, that will only consider
825 * the first "bitnumber" bits of the IP (bitnumber must be in 0-31 range)
826 * - An ip numer plus a /ip.bit.mask.values that will consider
827 * only the bits marked as 1 in the ip.bit.mask.values
828 * In the last two cases both the ip number and the bitmask can specify
829 * less than 4 bytes, the missing bytes then default to zero, note that
830 * this is *different* from the way inet_aton() does and that this does
831 * NOT happen for normal IPmasks (not containing '/')
832 * If the returned value is zero the produced in_mask might match some IP,
833 * if it's nonzero it will never match anything (and the imask struct is
834 * set so that always fails).
836 * The returned structure contains 3 fields whose meaning is the following:
837 * im.mask = The bits considered significative in the IP
838 * im.bits = What these bits should look like to have a match
839 * im.fall = If zero means that the above information used as
840 * ((IP & im.mask) == im.bits) is enough to tell if the compiled
841 * mask matches the given IP, nonzero means that it is needed,
842 * in case they did match, to call also the usual text match
843 * functions, because the mask wasn't "completely compiled"
845 * They should be used like:
846 * matchcompIP(&im, mask);
847 * if ( ((IP & im.mask)!=im.bits)) || (im.fall&&match(mask,inet_ntoa(IP))) )
848 * { handle_non_match } else { handle_match };
850 * if ( match(mask, inet_ntoa(IP)) )
851 * { handle_non_match } else { handle_match };
853 * Note: This function could be smarter when dealing with complex masks,
854 * this implementation is quite lazy and understands only very simple
855 * cases, whatever contains a ? anywhere or contains a '*' that isn't
856 * part of a trailing '.*' will fallback to text-match, this could be
857 * avoided for masks like 12?3.5.6 12.*.3.4 1.*.*.2 72?72?72?72 and
858 * so on that "could" be completely compiled to IP masks.
859 * If you try to improve this be aware of the fact that ? and *
860 * could match both dots and digits and we _must_ always reject
861 * what doesn't match in textform (like leading zeros and so on),
862 * so it's a LOT more tricky than it might seem. By now most common
863 * cases are optimized.
866 int matchcompIP(struct in_mask *imask, const char *mask)
868 const char *m = mask;
869 unsigned int bits = 0;
870 unsigned int filt = 0;
881 if ((m[1] == '\\') || (m[1] == '*') || (m[1] == '?')
895 if (digits && !tmp) /* Leading zeros */
899 tmp += (*m - '0'); /* Can't overflow, INT_MAX > 2559 */
905 /* Intentional fallthrough */
907 if ((!shift) != (!*m))
909 /* Intentional fallthrough */
911 bits |= (tmp << shift);
923 if (digits && !tmp) /* Leading zeros */
927 tmp += (*m - '0'); /* Can't overflow, INT_MAX > 2559 */
937 if ((!shift) && (*m))
939 filt |= (tmp << shift);
953 if (filt && (!(shift < 16)) && (!(filt & 0xE0FFFFFF)))
954 filt = 0xFFFFFFFF << (32 - ((filt >> 24)));
959 /* Intentional fallthrough */
963 filt = (0xFFFFFFFF << (shift)) << 8;
996 /* If we get here there is some error and this can't ever match */
1000 break; /* This time break the loop :) */
1005 imask->bits.s_addr = htonl(bits);
1006 imask->mask.s_addr = htonl(filt);
1008 return ((bits & ~filt) ? -1 : 0);