coregrind/m_seqmatch.c

   1
   2 /*--------------------------------------------------------------------*/
   3 /*--- A simple sequence matching facility.                         ---*/
   4 /*---                                                 m_seqmatch.c ---*/
   5 /*--------------------------------------------------------------------*/
   6
   7 /*
   8    This file is part of Valgrind, a dynamic binary instrumentation
   9    framework.
  10
  11    Copyright (C) 2008-2017 OpenWorks Ltd
  12       info@open-works.co.uk
  13
  14    This program is free software; you can redistribute it and/or
  15    modify it under the terms of the GNU General Public License as
  16    published by the Free Software Foundation; either version 2 of the
  17    License, or (at your option) any later version.
  18
  19    This program is distributed in the hope that it will be useful, but
  20    WITHOUT ANY WARRANTY; without even the implied warranty of
  21    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  22    General Public License for more details.
  23
  24    You should have received a copy of the GNU General Public License
  25    along with this program; if not, see <http://www.gnu.org/licenses/>.
  26
  27    The GNU General Public License is contained in the file COPYING.
  28 */
  29
  30 #include "pub_core_basics.h"
  31 #include "pub_core_libcassert.h"
  32 #include "pub_core_libcbase.h"    // VG_(strlen)
  33 #include "pub_core_seqmatch.h"    // self
  34
  35 /* ---------------------------------------------------------------------
  36    A simple sequence matching facility
  37    ------------------------------------------------------------------ */
  38
  39 /* See detailed comment in include/pub_tool_seqmatch.h about this. */
  40 Bool VG_(generic_match) (
  41         Bool matchAll,
  42         const void* patt,  SizeT szbPatt,  UWord nPatt,  UWord ixPatt,
  43         const void* input, SizeT szbInput, UWord nInput, UWord ixInput,
  44         Bool (*pIsStar)(const void*),
  45         Bool (*pIsQuery)(const void*),
  46         Bool (*pattEQinp)(const void*,const void*,void*,UWord),
  47         void* inputCompleter, Bool (*haveInputInpC)(void*,UWord)
  48      )
  49 {
  50    /* This is the spec, written in my favourite formal specification
  51       language.  It specifies non-greedy matching of '*'s.
  52
  53       ma ('*':ps) (i:is) = ma ps (i:is) || ma ('*':ps) is
  54       ma ('*':ps) []     = ma ps []
  55
  56       ma ('?':ps) (i:is) = ma ps is
  57       ma ('?':ps) []     = False
  58
  59       ma (p:ps)   (i:is) = p == i && ma ps is
  60
  61       ma (p:ps)   []     = False
  62       ma []       (i:is) = False -- m-all, True for m-prefix
  63       ma []       []     = True
  64    */
  65    Bool  havePatt, haveInput;
  66    const HChar *currPatt, *currInput;
  67   tailcall:
  68    vg_assert(nPatt  < 1000000); /* arbitrary */
  69    vg_assert(inputCompleter || (nInput < 1000000)); /* arbitrary */
  70    vg_assert(ixPatt <= nPatt);
  71    vg_assert(inputCompleter || ixInput <= nInput);
  72
  73    havePatt  = ixPatt < nPatt;
  74    haveInput = inputCompleter ?
  75       (*haveInputInpC)(inputCompleter, ixInput)
  76       : ixInput < nInput;
  77
  78    /* No specific need to set NULL when !have{Patt,Input}, but guards
  79       against inadvertantly dereferencing an out of range pointer to
  80       the pattern or input arrays. */
  81    currPatt  = havePatt  ? ((const HChar*)patt) + szbPatt * ixPatt    : NULL;
  82    currInput = haveInput && !inputCompleter ?
  83       ((const HChar*)input) + szbInput * ixInput : NULL;
  84
  85    // Deal with the complex case first: wildcards.  Do frugal
  86    // matching.  When encountering a '*', first skip no characters
  87    // at all, and see if the rest of the match still works.  Only if
  88    // that fails do we then skip a character, and retry at the next
  89    // position.
  90    //
  91    // ma ('*':ps) (i:is) = ma ps (i:is) || ma ('*':ps) is
  92    //
  93    // If we're out of input, check the rest of the pattern matches
  94    // the empty input.  This really means can only be be empty or
  95    // composed entirely of '*'s.
  96    //
  97    // ma ('*':ps) []     = ma ps []
  98    //
  99    if (havePatt && pIsStar(currPatt)) {
 100       if (haveInput) {
 101          // ma ('*':ps) (i:is) = ma ps (i:is) || ma ('*':ps) is
 102          // we unavoidably have to make a real recursive call for the
 103          // first half of the OR, since this isn't straight tail-recursion.
 104          if (VG_(generic_match)( matchAll,
 105                                  patt, szbPatt, nPatt,  ixPatt+1,
 106                                  input,szbInput,nInput, ixInput+0,
 107                                  pIsStar,pIsQuery,pattEQinp,
 108                                  inputCompleter,haveInputInpC) ) {
 109             return True;
 110          }
 111          // but we can tail-recurse for the second call
 112          ixInput++; goto tailcall;
 113       } else {
 114          // ma ('*':ps) []     = ma ps []
 115          ixPatt++; goto tailcall;
 116       }
 117    }
 118
 119    // simpler cases now.  Deal with '?' wildcards.
 120    //
 121    // ma ('?':ps) (i:is) = ma ps is
 122    // ma ('?':ps) []     = False
 123    if (havePatt && pIsQuery(currPatt)) {
 124       if (haveInput) {
 125          ixPatt++; ixInput++; goto tailcall;
 126       } else {
 127          return False;
 128       }
 129    }
 130
 131    // obvious case with literal chars in the pattern
 132    //
 133    // ma (p:ps)   (i:is) = p == i && ma ps is
 134    if (havePatt && haveInput) {
 135       if (!pattEQinp(currPatt,currInput,inputCompleter,ixInput)) return False;
 136       ixPatt++; ixInput++; goto tailcall;
 137    }
 138
 139    // if we run out of input before we run out of pattern, we must fail
 140    // ma (p:ps)   []     = False
 141    if (havePatt && !haveInput) return False;
 142
 143    // if we run out of pattern before we run out of input, the
 144    // verdict depends on the matching mode.  If we are trying to
 145    // match exactly (the pattern must consume the entire input)
 146    // then the outcome is failure.  However, if we're merely attempting
 147    // to match some prefix of the input, then we have been successful.
 148    //
 149    // ma []       (i:is) = False -- m-all, True for m-prefix
 150    if (!havePatt && haveInput) {
 151       return matchAll ? False // match-all
 152                       : True; // match-prefix
 153    }
 154
 155    // finally, if both sequence and input are both completely
 156    // consumed, then we were successful, regardless of matching mode.
 157    if (!havePatt && !haveInput) return True;
 158
 159    // end of cases
 160    vg_assert(0);
 161 }
 162
 163
 164 /* And a parameterization of the above, to make it do
 165    string matching.
 166 */
 167 static Bool charIsStar  ( const void* pV ) { return *(const HChar*)pV == '*'; }
 168 static Bool charIsQuery ( const void* pV ) { return *(const HChar*)pV == '?'; }
 169 static Bool char_p_EQ_i ( const void* pV, const void* cV,
 170                           void* null_completer, UWord ixcV ) {
 171    HChar p = *(const HChar*)pV;
 172    HChar c = *(const HChar*)cV;
 173    vg_assert(p != '*' && p != '?');
 174    return p == c;
 175 }
 176 Bool VG_(string_match) ( const HChar* patt, const HChar* input )
 177 {
 178    return VG_(generic_match)(
 179              True/* match-all */,
 180              patt,  sizeof(HChar), VG_(strlen)(patt), 0,
 181              input, sizeof(HChar), VG_(strlen)(input), 0,
 182              charIsStar, charIsQuery, char_p_EQ_i,
 183              NULL, NULL
 184           );
 185 }
 186
 187
 188 // test cases for the matcher (in match-all mode)
 189 // typedef struct { char* patt; char* input; Bool xres; } Test;
 190 //
 191 //static Test tests[] =
 192 //  {
 193 //    { ""          ,""   , True },
 194 //    { "a"         ,""   , False },
 195 //    { "a"         ,"b"  , False },
 196 //    { "a"         ,"a"  , True },
 197 //    { "a"         ,"aa" , False },
 198 //    { "*"         ,""   , True },
 199 //    { "**"        ,""   , True },
 200 //    { "*"         ,"abc", True },
 201 //    { "*a"        ,"abc", False },
 202 //    { "*b"        ,"abc", False },
 203 //    { "*bc"       ,"abc", True },
 204 //    { "a*b"       ,"abc", False },
 205 //    { "a*c"       ,"abc", True },
 206 //    { "*c"        ,"abc", True },
 207 //    { "c*c"       ,"abc", False },
 208 //    { "abc*"      ,"abc", True },
 209 //    { "abc**"     ,"abc", True },
 210 //    { "**abc"     ,"abc", True },
 211 //    { "**a*b*c**" ,"abc", True },
 212 //    { "**a*b*d**" ,"abc", False },
 213 //    { "a?b"       ,"abc", False },
 214 //    { "a?c"       ,"abc", True },
 215 //    { "?"         ,""   , False },
 216 //    { "?"         ,"a"  , True },
 217 //    { "?"         ,"ab" , False },
 218 //    { "abcd"      ,"abc", False },
 219 //    { "ab"        ,"abc", False },
 220 //    { NULL        ,NULL , False }
 221 //  };
 222 //
 223 //int main ( void )
 224 //{
 225 //   Test* t;
 226 //   for (t = tests; t->patt; t++) {
 227 //     printf("%-10s %-6s  %s\n",
 228 //            t->patt, t->input,
 229 //            match_string_all((UChar*)t->patt,(UChar*)t->input,True)
 230 //            == t->xres
 231 //               ? "pass" : "FAIL" );
 232 //   }
 233 //   return 0;
 234 //}
 235
 236 /*--------------------------------------------------------------------*/
 237 /*--- end                                             m_seqmatch.c ---*/
 238 /*--------------------------------------------------------------------*/