src/whereexpr.c

   1 /*
   2 ** 2015-06-08
   3 **
   4 ** The author disclaims copyright to this source code.  In place of
   5 ** a legal notice, here is a blessing:
   6 **
   7 **    May you do good and not evil.
   8 **    May you find forgiveness for yourself and forgive others.
   9 **    May you share freely, never taking more than you give.
  10 **
  11 *************************************************************************
  12 ** This module contains C code that generates VDBE code used to process
  13 ** the WHERE clause of SQL statements.
  14 **
  15 ** This file was originally part of where.c but was split out to improve
  16 ** readability and editabiliity.  This file contains utility routines for
  17 ** analyzing Expr objects in the WHERE clause.
  18 */
  19 #include "sqliteInt.h"
  20 #include "whereInt.h"
  21
  22 /* Forward declarations */
  23 static void exprAnalyze(SrcList*, WhereClause*, int);
  24
  25 /*
  26 ** Deallocate all memory associated with a WhereOrInfo object.
  27 */
  28 static void whereOrInfoDelete(sqlite3 *db, WhereOrInfo *p){
  29   sqlite3WhereClauseClear(&p->wc);
  30   sqlite3DbFree(db, p);
  31 }
  32
  33 /*
  34 ** Deallocate all memory associated with a WhereAndInfo object.
  35 */
  36 static void whereAndInfoDelete(sqlite3 *db, WhereAndInfo *p){
  37   sqlite3WhereClauseClear(&p->wc);
  38   sqlite3DbFree(db, p);
  39 }
  40
  41 /*
  42 ** Add a single new WhereTerm entry to the WhereClause object pWC.
  43 ** The new WhereTerm object is constructed from Expr p and with wtFlags.
  44 ** The index in pWC->a[] of the new WhereTerm is returned on success.
  45 ** 0 is returned if the new WhereTerm could not be added due to a memory
  46 ** allocation error.  The memory allocation failure will be recorded in
  47 ** the db->mallocFailed flag so that higher-level functions can detect it.
  48 **
  49 ** This routine will increase the size of the pWC->a[] array as necessary.
  50 **
  51 ** If the wtFlags argument includes TERM_DYNAMIC, then responsibility
  52 ** for freeing the expression p is assumed by the WhereClause object pWC.
  53 ** This is true even if this routine fails to allocate a new WhereTerm.
  54 **
  55 ** WARNING:  This routine might reallocate the space used to store
  56 ** WhereTerms.  All pointers to WhereTerms should be invalidated after
  57 ** calling this routine.  Such pointers may be reinitialized by referencing
  58 ** the pWC->a[] array.
  59 */
  60 static int whereClauseInsert(WhereClause *pWC, Expr *p, u16 wtFlags){
  61   WhereTerm *pTerm;
  62   int idx;
  63   testcase( wtFlags & TERM_VIRTUAL );
  64   if( pWC->nTerm>=pWC->nSlot ){
  65     WhereTerm *pOld = pWC->a;
  66     sqlite3 *db = pWC->pWInfo->pParse->db;
  67     pWC->a = sqlite3WhereMalloc(pWC->pWInfo, sizeof(pWC->a[0])*pWC->nSlot*2 );
  68     if( pWC->a==0 ){
  69       if( wtFlags & TERM_DYNAMIC ){
  70         sqlite3ExprDelete(db, p);
  71       }
  72       pWC->a = pOld;
  73       return 0;
  74     }
  75     memcpy(pWC->a, pOld, sizeof(pWC->a[0])*pWC->nTerm);
  76     pWC->nSlot = pWC->nSlot*2;
  77   }
  78   pTerm = &pWC->a[idx = pWC->nTerm++];
  79   if( (wtFlags & TERM_VIRTUAL)==0 ) pWC->nBase = pWC->nTerm;
  80   if( p && ExprHasProperty(p, EP_Unlikely) ){
  81     pTerm->truthProb = sqlite3LogEst(p->iTable) - 270;
  82   }else{
  83     pTerm->truthProb = 1;
  84   }
  85   pTerm->pExpr = sqlite3ExprSkipCollateAndLikely(p);
  86   pTerm->wtFlags = wtFlags;
  87   pTerm->pWC = pWC;
  88   pTerm->iParent = -1;
  89   memset(&pTerm->eOperator, 0,
  90          sizeof(WhereTerm) - offsetof(WhereTerm,eOperator));
  91   return idx;
  92 }
  93
  94 /*
  95 ** Return TRUE if the given operator is one of the operators that is
  96 ** allowed for an indexable WHERE clause term.  The allowed operators are
  97 ** "=", "<", ">", "<=", ">=", "IN", "IS", and "IS NULL"
  98 */
  99 static int allowedOp(int op){
 100   assert( TK_GT>TK_EQ && TK_GT<TK_GE );
 101   assert( TK_LT>TK_EQ && TK_LT<TK_GE );
 102   assert( TK_LE>TK_EQ && TK_LE<TK_GE );
 103   assert( TK_GE==TK_EQ+4 );
 104   return op==TK_IN || (op>=TK_EQ && op<=TK_GE) || op==TK_ISNULL || op==TK_IS;
 105 }
 106
 107 /*
 108 ** Commute a comparison operator.  Expressions of the form "X op Y"
 109 ** are converted into "Y op X".
 110 */
 111 static u16 exprCommute(Parse *pParse, Expr *pExpr){
 112   if( pExpr->pLeft->op==TK_VECTOR
 113    || pExpr->pRight->op==TK_VECTOR
 114    || sqlite3BinaryCompareCollSeq(pParse, pExpr->pLeft, pExpr->pRight) !=
 115       sqlite3BinaryCompareCollSeq(pParse, pExpr->pRight, pExpr->pLeft)
 116   ){
 117     pExpr->flags ^= EP_Commuted;
 118   }
 119   SWAP(Expr*,pExpr->pRight,pExpr->pLeft);
 120   if( pExpr->op>=TK_GT ){
 121     assert( TK_LT==TK_GT+2 );
 122     assert( TK_GE==TK_LE+2 );
 123     assert( TK_GT>TK_EQ );
 124     assert( TK_GT<TK_LE );
 125     assert( pExpr->op>=TK_GT && pExpr->op<=TK_GE );
 126     pExpr->op = ((pExpr->op-TK_GT)^2)+TK_GT;
 127   }
 128   return 0;
 129 }
 130
 131 /*
 132 ** Translate from TK_xx operator to WO_xx bitmask.
 133 */
 134 static u16 operatorMask(int op){
 135   u16 c;
 136   assert( allowedOp(op) );
 137   if( op==TK_IN ){
 138     c = WO_IN;
 139   }else if( op==TK_ISNULL ){
 140     c = WO_ISNULL;
 141   }else if( op==TK_IS ){
 142     c = WO_IS;
 143   }else{
 144     assert( (WO_EQ<<(op-TK_EQ)) < 0x7fff );
 145     c = (u16)(WO_EQ<<(op-TK_EQ));
 146   }
 147   assert( op!=TK_ISNULL || c==WO_ISNULL );
 148   assert( op!=TK_IN || c==WO_IN );
 149   assert( op!=TK_EQ || c==WO_EQ );
 150   assert( op!=TK_LT || c==WO_LT );
 151   assert( op!=TK_LE || c==WO_LE );
 152   assert( op!=TK_GT || c==WO_GT );
 153   assert( op!=TK_GE || c==WO_GE );
 154   assert( op!=TK_IS || c==WO_IS );
 155   return c;
 156 }
 157
 158
 159 #ifndef SQLITE_OMIT_LIKE_OPTIMIZATION
 160 /*
 161 ** Check to see if the given expression is a LIKE or GLOB operator that
 162 ** can be optimized using inequality constraints.  Return TRUE if it is
 163 ** so and false if not.
 164 **
 165 ** In order for the operator to be optimizible, the RHS must be a string
 166 ** literal that does not begin with a wildcard.  The LHS must be a column
 167 ** that may only be NULL, a string, or a BLOB, never a number. (This means
 168 ** that virtual tables cannot participate in the LIKE optimization.)  The
 169 ** collating sequence for the column on the LHS must be appropriate for
 170 ** the operator.
 171 */
 172 static int isLikeOrGlob(
 173   Parse *pParse,    /* Parsing and code generating context */
 174   Expr *pExpr,      /* Test this expression */
 175   Expr **ppPrefix,  /* Pointer to TK_STRING expression with pattern prefix */
 176   int *pisComplete, /* True if the only wildcard is % in the last character */
 177   int *pnoCase      /* True if uppercase is equivalent to lowercase */
 178 ){
 179   const u8 *z = 0;           /* String on RHS of LIKE operator */
 180   Expr *pRight, *pLeft;      /* Right and left size of LIKE operator */
 181   ExprList *pList;           /* List of operands to the LIKE operator */
 182   u8 c;                      /* One character in z[] */
 183   int cnt;                   /* Number of non-wildcard prefix characters */
 184   u8 wc[4];                  /* Wildcard characters */
 185   sqlite3 *db = pParse->db;  /* Database connection */
 186   sqlite3_value *pVal = 0;
 187   int op;                    /* Opcode of pRight */
 188   int rc;                    /* Result code to return */
 189
 190   if( !sqlite3IsLikeFunction(db, pExpr, pnoCase, (char*)wc) ){
 191     return 0;
 192   }
 193 #ifdef SQLITE_EBCDIC
 194   if( *pnoCase ) return 0;
 195 #endif
 196   assert( ExprUseXList(pExpr) );
 197   pList = pExpr->x.pList;
 198   pLeft = pList->a[1].pExpr;
 199
 200   pRight = sqlite3ExprSkipCollate(pList->a[0].pExpr);
 201   op = pRight->op;
 202   if( op==TK_VARIABLE && (db->flags & SQLITE_EnableQPSG)==0 ){
 203     Vdbe *pReprepare = pParse->pReprepare;
 204     int iCol = pRight->iColumn;
 205     pVal = sqlite3VdbeGetBoundValue(pReprepare, iCol, SQLITE_AFF_BLOB);
 206     if( pVal && sqlite3_value_type(pVal)==SQLITE_TEXT ){
 207       z = sqlite3_value_text(pVal);
 208     }
 209     sqlite3VdbeSetVarmask(pParse->pVdbe, iCol);
 210     assert( pRight->op==TK_VARIABLE || pRight->op==TK_REGISTER );
 211   }else if( op==TK_STRING ){
 212     assert( !ExprHasProperty(pRight, EP_IntValue) );
 213      z = (u8*)pRight->u.zToken;
 214   }
 215   if( z ){
 216
 217     /* Count the number of prefix characters prior to the first wildcard */
 218     cnt = 0;
 219     while( (c=z[cnt])!=0 && c!=wc[0] && c!=wc[1] && c!=wc[2] ){
 220       cnt++;
 221       if( c==wc[3] && z[cnt]!=0 ) cnt++;
 222     }
 223
 224     /* The optimization is possible only if (1) the pattern does not begin
 225     ** with a wildcard and if (2) the non-wildcard prefix does not end with
 226     ** an (illegal 0xff) character, or (3) the pattern does not consist of
 227     ** a single escape character. The second condition is necessary so
 228     ** that we can increment the prefix key to find an upper bound for the
 229     ** range search. The third is because the caller assumes that the pattern
 230     ** consists of at least one character after all escapes have been
 231     ** removed.  */
 232     if( cnt!=0 && 255!=(u8)z[cnt-1] && (cnt>1 || z[0]!=wc[3]) ){
 233       Expr *pPrefix;
 234
 235       /* A "complete" match if the pattern ends with "*" or "%" */
 236       *pisComplete = c==wc[0] && z[cnt+1]==0;
 237
 238       /* Get the pattern prefix.  Remove all escapes from the prefix. */
 239       pPrefix = sqlite3Expr(db, TK_STRING, (char*)z);
 240       if( pPrefix ){
 241         int iFrom, iTo;
 242         char *zNew;
 243         assert( !ExprHasProperty(pPrefix, EP_IntValue) );
 244         zNew = pPrefix->u.zToken;
 245         zNew[cnt] = 0;
 246         for(iFrom=iTo=0; iFrom<cnt; iFrom++){
 247           if( zNew[iFrom]==wc[3] ) iFrom++;
 248           zNew[iTo++] = zNew[iFrom];
 249         }
 250         zNew[iTo] = 0;
 251         assert( iTo>0 );
 252
 253         /* If the LHS is not an ordinary column with TEXT affinity, then the
 254         ** pattern prefix boundaries (both the start and end boundaries) must
 255         ** not look like a number.  Otherwise the pattern might be treated as
 256         ** a number, which will invalidate the LIKE optimization.
 257         **
 258         ** Getting this right has been a persistent source of bugs in the
 259         ** LIKE optimization.  See, for example:
 260         **    2018-09-10 https://sqlite.org/src/info/c94369cae9b561b1
 261         **    2019-05-02 https://sqlite.org/src/info/b043a54c3de54b28
 262         **    2019-06-10 https://sqlite.org/src/info/fd76310a5e843e07
 263         **    2019-06-14 https://sqlite.org/src/info/ce8717f0885af975
 264         **    2019-09-03 https://sqlite.org/src/info/0f0428096f17252a
 265         */
 266         if( pLeft->op!=TK_COLUMN
 267          || sqlite3ExprAffinity(pLeft)!=SQLITE_AFF_TEXT
 268          || (ALWAYS( ExprUseYTab(pLeft) )
 269              && pLeft->y.pTab
 270              && IsVirtual(pLeft->y.pTab))  /* Might be numeric */
 271         ){
 272           int isNum;
 273           double rDummy;
 274           isNum = sqlite3AtoF(zNew, &rDummy, iTo, SQLITE_UTF8);
 275           if( isNum<=0 ){
 276             if( iTo==1 && zNew[0]=='-' ){
 277               isNum = +1;
 278             }else{
 279               zNew[iTo-1]++;
 280               isNum = sqlite3AtoF(zNew, &rDummy, iTo, SQLITE_UTF8);
 281               zNew[iTo-1]--;
 282             }
 283           }
 284           if( isNum>0 ){
 285             sqlite3ExprDelete(db, pPrefix);
 286             sqlite3ValueFree(pVal);
 287             return 0;
 288           }
 289         }
 290       }
 291       *ppPrefix = pPrefix;
 292
 293       /* If the RHS pattern is a bound parameter, make arrangements to
 294       ** reprepare the statement when that parameter is rebound */
 295       if( op==TK_VARIABLE ){
 296         Vdbe *v = pParse->pVdbe;
 297         sqlite3VdbeSetVarmask(v, pRight->iColumn);
 298         assert( !ExprHasProperty(pRight, EP_IntValue) );
 299         if( *pisComplete && pRight->u.zToken[1] ){
 300           /* If the rhs of the LIKE expression is a variable, and the current
 301           ** value of the variable means there is no need to invoke the LIKE
 302           ** function, then no OP_Variable will be added to the program.
 303           ** This causes problems for the sqlite3_bind_parameter_name()
 304           ** API. To work around them, add a dummy OP_Variable here.
 305           */
 306           int r1 = sqlite3GetTempReg(pParse);
 307           sqlite3ExprCodeTarget(pParse, pRight, r1);
 308           sqlite3VdbeChangeP3(v, sqlite3VdbeCurrentAddr(v)-1, 0);
 309           sqlite3ReleaseTempReg(pParse, r1);
 310         }
 311       }
 312     }else{
 313       z = 0;
 314     }
 315   }
 316
 317   rc = (z!=0);
 318   sqlite3ValueFree(pVal);
 319   return rc;
 320 }
 321 #endif /* SQLITE_OMIT_LIKE_OPTIMIZATION */
 322
 323
 324 #ifndef SQLITE_OMIT_VIRTUALTABLE
 325 /*
 326 ** Check to see if the pExpr expression is a form that needs to be passed
 327 ** to the xBestIndex method of virtual tables.  Forms of interest include:
 328 **
 329 **          Expression                   Virtual Table Operator
 330 **          -----------------------      ---------------------------------
 331 **      1.  column MATCH expr            SQLITE_INDEX_CONSTRAINT_MATCH
 332 **      2.  column GLOB expr             SQLITE_INDEX_CONSTRAINT_GLOB
 333 **      3.  column LIKE expr             SQLITE_INDEX_CONSTRAINT_LIKE
 334 **      4.  column REGEXP expr           SQLITE_INDEX_CONSTRAINT_REGEXP
 335 **      5.  column != expr               SQLITE_INDEX_CONSTRAINT_NE
 336 **      6.  expr != column               SQLITE_INDEX_CONSTRAINT_NE
 337 **      7.  column IS NOT expr           SQLITE_INDEX_CONSTRAINT_ISNOT
 338 **      8.  expr IS NOT column           SQLITE_INDEX_CONSTRAINT_ISNOT
 339 **      9.  column IS NOT NULL           SQLITE_INDEX_CONSTRAINT_ISNOTNULL
 340 **
 341 ** In every case, "column" must be a column of a virtual table.  If there
 342 ** is a match, set *ppLeft to the "column" expression, set *ppRight to the
 343 ** "expr" expression (even though in forms (6) and (8) the column is on the
 344 ** right and the expression is on the left).  Also set *peOp2 to the
 345 ** appropriate virtual table operator.  The return value is 1 or 2 if there
 346 ** is a match.  The usual return is 1, but if the RHS is also a column
 347 ** of virtual table in forms (5) or (7) then return 2.
 348 **
 349 ** If the expression matches none of the patterns above, return 0.
 350 */
 351 static int isAuxiliaryVtabOperator(
 352   sqlite3 *db,                    /* Parsing context */
 353   Expr *pExpr,                    /* Test this expression */
 354   unsigned char *peOp2,           /* OUT: 0 for MATCH, or else an op2 value */
 355   Expr **ppLeft,                  /* Column expression to left of MATCH/op2 */
 356   Expr **ppRight                  /* Expression to left of MATCH/op2 */
 357 ){
 358   if( pExpr->op==TK_FUNCTION ){
 359     static const struct Op2 {
 360       const char *zOp;
 361       unsigned char eOp2;
 362     } aOp[] = {
 363       { "match",  SQLITE_INDEX_CONSTRAINT_MATCH },
 364       { "glob",   SQLITE_INDEX_CONSTRAINT_GLOB },
 365       { "like",   SQLITE_INDEX_CONSTRAINT_LIKE },
 366       { "regexp", SQLITE_INDEX_CONSTRAINT_REGEXP }
 367     };
 368     ExprList *pList;
 369     Expr *pCol;                     /* Column reference */
 370     int i;
 371
 372     assert( ExprUseXList(pExpr) );
 373     pList = pExpr->x.pList;
 374     if( pList==0 || pList->nExpr!=2 ){
 375       return 0;
 376     }
 377
 378     /* Built-in operators MATCH, GLOB, LIKE, and REGEXP attach to a
 379     ** virtual table on their second argument, which is the same as
 380     ** the left-hand side operand in their in-fix form.
 381     **
 382     **       vtab_column MATCH expression
 383     **       MATCH(expression,vtab_column)
 384     */
 385     pCol = pList->a[1].pExpr;
 386     assert( pCol->op!=TK_COLUMN || ExprUseYTab(pCol) );
 387     testcase( pCol->op==TK_COLUMN && pCol->y.pTab==0 );
 388     if( ExprIsVtab(pCol) ){
 389       for(i=0; i<ArraySize(aOp); i++){
 390         assert( !ExprHasProperty(pExpr, EP_IntValue) );
 391         if( sqlite3StrICmp(pExpr->u.zToken, aOp[i].zOp)==0 ){
 392           *peOp2 = aOp[i].eOp2;
 393           *ppRight = pList->a[0].pExpr;
 394           *ppLeft = pCol;
 395           return 1;
 396         }
 397       }
 398     }
 399
 400     /* We can also match against the first column of overloaded
 401     ** functions where xFindFunction returns a value of at least
 402     ** SQLITE_INDEX_CONSTRAINT_FUNCTION.
 403     **
 404     **      OVERLOADED(vtab_column,expression)
 405     **
 406     ** Historically, xFindFunction expected to see lower-case function
 407     ** names.  But for this use case, xFindFunction is expected to deal
 408     ** with function names in an arbitrary case.
 409     */
 410     pCol = pList->a[0].pExpr;
 411     assert( pCol->op!=TK_COLUMN || ExprUseYTab(pCol) );
 412     testcase( pCol->op==TK_COLUMN && pCol->y.pTab==0 );
 413     if( ExprIsVtab(pCol) ){
 414       sqlite3_vtab *pVtab;
 415       sqlite3_module *pMod;
 416       void (*xNotUsed)(sqlite3_context*,int,sqlite3_value**);
 417       void *pNotUsed;
 418       pVtab = sqlite3GetVTable(db, pCol->y.pTab)->pVtab;
 419       assert( pVtab!=0 );
 420       assert( pVtab->pModule!=0 );
 421       assert( !ExprHasProperty(pExpr, EP_IntValue) );
 422       pMod = (sqlite3_module *)pVtab->pModule;
 423       if( pMod->xFindFunction!=0 ){
 424         i = pMod->xFindFunction(pVtab,2, pExpr->u.zToken, &xNotUsed, &pNotUsed);
 425         if( i>=SQLITE_INDEX_CONSTRAINT_FUNCTION ){
 426           *peOp2 = i;
 427           *ppRight = pList->a[1].pExpr;
 428           *ppLeft = pCol;
 429           return 1;
 430         }
 431       }
 432     }
 433   }else if( pExpr->op==TK_NE || pExpr->op==TK_ISNOT || pExpr->op==TK_NOTNULL ){
 434     int res = 0;
 435     Expr *pLeft = pExpr->pLeft;
 436     Expr *pRight = pExpr->pRight;
 437     assert( pLeft->op!=TK_COLUMN || ExprUseYTab(pLeft) );
 438     testcase( pLeft->op==TK_COLUMN && pLeft->y.pTab==0 );
 439     if( ExprIsVtab(pLeft) ){
 440       res++;
 441     }
 442     assert( pRight==0 || pRight->op!=TK_COLUMN || ExprUseYTab(pRight) );
 443     testcase( pRight && pRight->op==TK_COLUMN && pRight->y.pTab==0 );
 444     if( pRight && ExprIsVtab(pRight) ){
 445       res++;
 446       SWAP(Expr*, pLeft, pRight);
 447     }
 448     *ppLeft = pLeft;
 449     *ppRight = pRight;
 450     if( pExpr->op==TK_NE ) *peOp2 = SQLITE_INDEX_CONSTRAINT_NE;
 451     if( pExpr->op==TK_ISNOT ) *peOp2 = SQLITE_INDEX_CONSTRAINT_ISNOT;
 452     if( pExpr->op==TK_NOTNULL ) *peOp2 = SQLITE_INDEX_CONSTRAINT_ISNOTNULL;
 453     return res;
 454   }
 455   return 0;
 456 }
 457 #endif /* SQLITE_OMIT_VIRTUALTABLE */
 458
 459 /*
 460 ** If the pBase expression originated in the ON or USING clause of
 461 ** a join, then transfer the appropriate markings over to derived.
 462 */
 463 static void transferJoinMarkings(Expr *pDerived, Expr *pBase){
 464   if( pDerived && ExprHasProperty(pBase, EP_OuterON|EP_InnerON) ){
 465     pDerived->flags |= pBase->flags & (EP_OuterON|EP_InnerON);
 466     pDerived->w.iJoin = pBase->w.iJoin;
 467   }
 468 }
 469
 470 /*
 471 ** Mark term iChild as being a child of term iParent
 472 */
 473 static void markTermAsChild(WhereClause *pWC, int iChild, int iParent){
 474   pWC->a[iChild].iParent = iParent;
 475   pWC->a[iChild].truthProb = pWC->a[iParent].truthProb;
 476   pWC->a[iParent].nChild++;
 477 }
 478
 479 /*
 480 ** Return the N-th AND-connected subterm of pTerm.  Or if pTerm is not
 481 ** a conjunction, then return just pTerm when N==0.  If N is exceeds
 482 ** the number of available subterms, return NULL.
 483 */
 484 static WhereTerm *whereNthSubterm(WhereTerm *pTerm, int N){
 485   if( pTerm->eOperator!=WO_AND ){
 486     return N==0 ? pTerm : 0;
 487   }
 488   if( N<pTerm->u.pAndInfo->wc.nTerm ){
 489     return &pTerm->u.pAndInfo->wc.a[N];
 490   }
 491   return 0;
 492 }
 493
 494 /*
 495 ** Subterms pOne and pTwo are contained within WHERE clause pWC.  The
 496 ** two subterms are in disjunction - they are OR-ed together.
 497 **
 498 ** If these two terms are both of the form:  "A op B" with the same
 499 ** A and B values but different operators and if the operators are
 500 ** compatible (if one is = and the other is <, for example) then
 501 ** add a new virtual AND term to pWC that is the combination of the
 502 ** two.
 503 **
 504 ** Some examples:
 505 **
 506 **    x<y OR x=y    -->     x<=y
 507 **    x=y OR x=y    -->     x=y
 508 **    x<=y OR x<y   -->     x<=y
 509 **
 510 ** The following is NOT generated:
 511 **
 512 **    x<y OR x>y    -->     x!=y
 513 */
 514 static void whereCombineDisjuncts(
 515   SrcList *pSrc,         /* the FROM clause */
 516   WhereClause *pWC,      /* The complete WHERE clause */
 517   WhereTerm *pOne,       /* First disjunct */
 518   WhereTerm *pTwo        /* Second disjunct */
 519 ){
 520   u16 eOp = pOne->eOperator | pTwo->eOperator;
 521   sqlite3 *db;           /* Database connection (for malloc) */
 522   Expr *pNew;            /* New virtual expression */
 523   int op;                /* Operator for the combined expression */
 524   int idxNew;            /* Index in pWC of the next virtual term */
 525
 526   if( (pOne->wtFlags | pTwo->wtFlags) & TERM_VNULL ) return;
 527   if( (pOne->eOperator & (WO_EQ|WO_LT|WO_LE|WO_GT|WO_GE))==0 ) return;
 528   if( (pTwo->eOperator & (WO_EQ|WO_LT|WO_LE|WO_GT|WO_GE))==0 ) return;
 529   if( (eOp & (WO_EQ|WO_LT|WO_LE))!=eOp
 530    && (eOp & (WO_EQ|WO_GT|WO_GE))!=eOp ) return;
 531   assert( pOne->pExpr->pLeft!=0 && pOne->pExpr->pRight!=0 );
 532   assert( pTwo->pExpr->pLeft!=0 && pTwo->pExpr->pRight!=0 );
 533   if( sqlite3ExprCompare(0,pOne->pExpr->pLeft, pTwo->pExpr->pLeft, -1) ) return;
 534   if( sqlite3ExprCompare(0,pOne->pExpr->pRight, pTwo->pExpr->pRight,-1) )return;
 535   /* If we reach this point, it means the two subterms can be combined */
 536   if( (eOp & (eOp-1))!=0 ){
 537     if( eOp & (WO_LT|WO_LE) ){
 538       eOp = WO_LE;
 539     }else{
 540       assert( eOp & (WO_GT|WO_GE) );
 541       eOp = WO_GE;
 542     }
 543   }
 544   db = pWC->pWInfo->pParse->db;
 545   pNew = sqlite3ExprDup(db, pOne->pExpr, 0);
 546   if( pNew==0 ) return;
 547   for(op=TK_EQ; eOp!=(WO_EQ<<(op-TK_EQ)); op++){ assert( op<TK_GE ); }
 548   pNew->op = op;
 549   idxNew = whereClauseInsert(pWC, pNew, TERM_VIRTUAL|TERM_DYNAMIC);
 550   exprAnalyze(pSrc, pWC, idxNew);
 551 }
 552
 553 #if !defined(SQLITE_OMIT_OR_OPTIMIZATION) && !defined(SQLITE_OMIT_SUBQUERY)
 554 /*
 555 ** Analyze a term that consists of two or more OR-connected
 556 ** subterms.  So in:
 557 **
 558 **     ... WHERE  (a=5) AND (b=7 OR c=9 OR d=13) AND (d=13)
 559 **                          ^^^^^^^^^^^^^^^^^^^^
 560 **
 561 ** This routine analyzes terms such as the middle term in the above example.
 562 ** A WhereOrTerm object is computed and attached to the term under
 563 ** analysis, regardless of the outcome of the analysis.  Hence:
 564 **
 565 **     WhereTerm.wtFlags   |=  TERM_ORINFO
 566 **     WhereTerm.u.pOrInfo  =  a dynamically allocated WhereOrTerm object
 567 **
 568 ** The term being analyzed must have two or more of OR-connected subterms.
 569 ** A single subterm might be a set of AND-connected sub-subterms.
 570 ** Examples of terms under analysis:
 571 **
 572 **     (A)     t1.x=t2.y OR t1.x=t2.z OR t1.y=15 OR t1.z=t3.a+5
 573 **     (B)     x=expr1 OR expr2=x OR x=expr3
 574 **     (C)     t1.x=t2.y OR (t1.x=t2.z AND t1.y=15)
 575 **     (D)     x=expr1 OR (y>11 AND y<22 AND z LIKE '*hello*')
 576 **     (E)     (p.a=1 AND q.b=2 AND r.c=3) OR (p.x=4 AND q.y=5 AND r.z=6)
 577 **     (F)     x>A OR (x=A AND y>=B)
 578 **
 579 ** CASE 1:
 580 **
 581 ** If all subterms are of the form T.C=expr for some single column of C and
 582 ** a single table T (as shown in example B above) then create a new virtual
 583 ** term that is an equivalent IN expression.  In other words, if the term
 584 ** being analyzed is:
 585 **
 586 **      x = expr1  OR  expr2 = x  OR  x = expr3
 587 **
 588 ** then create a new virtual term like this:
 589 **
 590 **      x IN (expr1,expr2,expr3)
 591 **
 592 ** CASE 2:
 593 **
 594 ** If there are exactly two disjuncts and one side has x>A and the other side
 595 ** has x=A (for the same x and A) then add a new virtual conjunct term to the
 596 ** WHERE clause of the form "x>=A".  Example:
 597 **
 598 **      x>A OR (x=A AND y>B)    adds:    x>=A
 599 **
 600 ** The added conjunct can sometimes be helpful in query planning.
 601 **
 602 ** CASE 3:
 603 **
 604 ** If all subterms are indexable by a single table T, then set
 605 **
 606 **     WhereTerm.eOperator              =  WO_OR
 607 **     WhereTerm.u.pOrInfo->indexable  |=  the cursor number for table T
 608 **
 609 ** A subterm is "indexable" if it is of the form
 610 ** "T.C <op> <expr>" where C is any column of table T and
 611 ** <op> is one of "=", "<", "<=", ">", ">=", "IS NULL", or "IN".
 612 ** A subterm is also indexable if it is an AND of two or more
 613 ** subsubterms at least one of which is indexable.  Indexable AND
 614 ** subterms have their eOperator set to WO_AND and they have
 615 ** u.pAndInfo set to a dynamically allocated WhereAndTerm object.
 616 **
 617 ** From another point of view, "indexable" means that the subterm could
 618 ** potentially be used with an index if an appropriate index exists.
 619 ** This analysis does not consider whether or not the index exists; that
 620 ** is decided elsewhere.  This analysis only looks at whether subterms
 621 ** appropriate for indexing exist.
 622 **
 623 ** All examples A through E above satisfy case 3.  But if a term
 624 ** also satisfies case 1 (such as B) we know that the optimizer will
 625 ** always prefer case 1, so in that case we pretend that case 3 is not
 626 ** satisfied.
 627 **
 628 ** It might be the case that multiple tables are indexable.  For example,
 629 ** (E) above is indexable on tables P, Q, and R.
 630 **
 631 ** Terms that satisfy case 3 are candidates for lookup by using
 632 ** separate indices to find rowids for each subterm and composing
 633 ** the union of all rowids using a RowSet object.  This is similar
 634 ** to "bitmap indices" in other database engines.
 635 **
 636 ** OTHERWISE:
 637 **
 638 ** If none of cases 1, 2, or 3 apply, then leave the eOperator set to
 639 ** zero.  This term is not useful for search.
 640 */
 641 static void exprAnalyzeOrTerm(
 642   SrcList *pSrc,            /* the FROM clause */
 643   WhereClause *pWC,         /* the complete WHERE clause */
 644   int idxTerm               /* Index of the OR-term to be analyzed */
 645 ){
 646   WhereInfo *pWInfo = pWC->pWInfo;        /* WHERE clause processing context */
 647   Parse *pParse = pWInfo->pParse;         /* Parser context */
 648   sqlite3 *db = pParse->db;               /* Database connection */
 649   WhereTerm *pTerm = &pWC->a[idxTerm];    /* The term to be analyzed */
 650   Expr *pExpr = pTerm->pExpr;             /* The expression of the term */
 651   int i;                                  /* Loop counters */
 652   WhereClause *pOrWc;       /* Breakup of pTerm into subterms */
 653   WhereTerm *pOrTerm;       /* A Sub-term within the pOrWc */
 654   WhereOrInfo *pOrInfo;     /* Additional information associated with pTerm */
 655   Bitmask chngToIN;         /* Tables that might satisfy case 1 */
 656   Bitmask indexable;        /* Tables that are indexable, satisfying case 2 */
 657
 658   /*
 659   ** Break the OR clause into its separate subterms.  The subterms are
 660   ** stored in a WhereClause structure containing within the WhereOrInfo
 661   ** object that is attached to the original OR clause term.
 662   */
 663   assert( (pTerm->wtFlags & (TERM_DYNAMIC|TERM_ORINFO|TERM_ANDINFO))==0 );
 664   assert( pExpr->op==TK_OR );
 665   pTerm->u.pOrInfo = pOrInfo = sqlite3DbMallocZero(db, sizeof(*pOrInfo));
 666   if( pOrInfo==0 ) return;
 667   pTerm->wtFlags |= TERM_ORINFO;
 668   pOrWc = &pOrInfo->wc;
 669   memset(pOrWc->aStatic, 0, sizeof(pOrWc->aStatic));
 670   sqlite3WhereClauseInit(pOrWc, pWInfo);
 671   sqlite3WhereSplit(pOrWc, pExpr, TK_OR);
 672   sqlite3WhereExprAnalyze(pSrc, pOrWc);
 673   if( db->mallocFailed ) return;
 674   assert( pOrWc->nTerm>=2 );
 675
 676   /*
 677   ** Compute the set of tables that might satisfy cases 1 or 3.
 678   */
 679   indexable = ~(Bitmask)0;
 680   chngToIN = ~(Bitmask)0;
 681   for(i=pOrWc->nTerm-1, pOrTerm=pOrWc->a; i>=0 && indexable; i--, pOrTerm++){
 682     if( (pOrTerm->eOperator & WO_SINGLE)==0 ){
 683       WhereAndInfo *pAndInfo;
 684       assert( (pOrTerm->wtFlags & (TERM_ANDINFO|TERM_ORINFO))==0 );
 685       chngToIN = 0;
 686       pAndInfo = sqlite3DbMallocRawNN(db, sizeof(*pAndInfo));
 687       if( pAndInfo ){
 688         WhereClause *pAndWC;
 689         WhereTerm *pAndTerm;
 690         int j;
 691         Bitmask b = 0;
 692         pOrTerm->u.pAndInfo = pAndInfo;
 693         pOrTerm->wtFlags |= TERM_ANDINFO;
 694         pOrTerm->eOperator = WO_AND;
 695         pOrTerm->leftCursor = -1;
 696         pAndWC = &pAndInfo->wc;
 697         memset(pAndWC->aStatic, 0, sizeof(pAndWC->aStatic));
 698         sqlite3WhereClauseInit(pAndWC, pWC->pWInfo);
 699         sqlite3WhereSplit(pAndWC, pOrTerm->pExpr, TK_AND);
 700         sqlite3WhereExprAnalyze(pSrc, pAndWC);
 701         pAndWC->pOuter = pWC;
 702         if( !db->mallocFailed ){
 703           for(j=0, pAndTerm=pAndWC->a; j<pAndWC->nTerm; j++, pAndTerm++){
 704             assert( pAndTerm->pExpr );
 705             if( allowedOp(pAndTerm->pExpr->op)
 706              || pAndTerm->eOperator==WO_AUX
 707             ){
 708               b |= sqlite3WhereGetMask(&pWInfo->sMaskSet, pAndTerm->leftCursor);
 709             }
 710           }
 711         }
 712         indexable &= b;
 713       }
 714     }else if( pOrTerm->wtFlags & TERM_COPIED ){
 715       /* Skip this term for now.  We revisit it when we process the
 716       ** corresponding TERM_VIRTUAL term */
 717     }else{
 718       Bitmask b;
 719       b = sqlite3WhereGetMask(&pWInfo->sMaskSet, pOrTerm->leftCursor);
 720       if( pOrTerm->wtFlags & TERM_VIRTUAL ){
 721         WhereTerm *pOther = &pOrWc->a[pOrTerm->iParent];
 722         b |= sqlite3WhereGetMask(&pWInfo->sMaskSet, pOther->leftCursor);
 723       }
 724       indexable &= b;
 725       if( (pOrTerm->eOperator & WO_EQ)==0 ){
 726         chngToIN = 0;
 727       }else{
 728         chngToIN &= b;
 729       }
 730     }
 731   }
 732
 733   /*
 734   ** Record the set of tables that satisfy case 3.  The set might be
 735   ** empty.
 736   */
 737   pOrInfo->indexable = indexable;
 738   pTerm->eOperator = WO_OR;
 739   pTerm->leftCursor = -1;
 740   if( indexable ){
 741     pWC->hasOr = 1;
 742   }
 743
 744   /* For a two-way OR, attempt to implementation case 2.
 745   */
 746   if( indexable && pOrWc->nTerm==2 ){
 747     int iOne = 0;
 748     WhereTerm *pOne;
 749     while( (pOne = whereNthSubterm(&pOrWc->a[0],iOne++))!=0 ){
 750       int iTwo = 0;
 751       WhereTerm *pTwo;
 752       while( (pTwo = whereNthSubterm(&pOrWc->a[1],iTwo++))!=0 ){
 753         whereCombineDisjuncts(pSrc, pWC, pOne, pTwo);
 754       }
 755     }
 756   }
 757
 758   /*
 759   ** chngToIN holds a set of tables that *might* satisfy case 1.  But
 760   ** we have to do some additional checking to see if case 1 really
 761   ** is satisfied.
 762   **
 763   ** chngToIN will hold either 0, 1, or 2 bits.  The 0-bit case means
 764   ** that there is no possibility of transforming the OR clause into an
 765   ** IN operator because one or more terms in the OR clause contain
 766   ** something other than == on a column in the single table.  The 1-bit
 767   ** case means that every term of the OR clause is of the form
 768   ** "table.column=expr" for some single table.  The one bit that is set
 769   ** will correspond to the common table.  We still need to check to make
 770   ** sure the same column is used on all terms.  The 2-bit case is when
 771   ** the all terms are of the form "table1.column=table2.column".  It
 772   ** might be possible to form an IN operator with either table1.column
 773   ** or table2.column as the LHS if either is common to every term of
 774   ** the OR clause.
 775   **
 776   ** Note that terms of the form "table.column1=table.column2" (the
 777   ** same table on both sizes of the ==) cannot be optimized.
 778   */
 779   if( chngToIN ){
 780     int okToChngToIN = 0;     /* True if the conversion to IN is valid */
 781     int iColumn = -1;         /* Column index on lhs of IN operator */
 782     int iCursor = -1;         /* Table cursor common to all terms */
 783     int j = 0;                /* Loop counter */
 784
 785     /* Search for a table and column that appears on one side or the
 786     ** other of the == operator in every subterm.  That table and column
 787     ** will be recorded in iCursor and iColumn.  There might not be any
 788     ** such table and column.  Set okToChngToIN if an appropriate table
 789     ** and column is found but leave okToChngToIN false if not found.
 790     */
 791     for(j=0; j<2 && !okToChngToIN; j++){
 792       Expr *pLeft = 0;
 793       pOrTerm = pOrWc->a;
 794       for(i=pOrWc->nTerm-1; i>=0; i--, pOrTerm++){
 795         assert( pOrTerm->eOperator & WO_EQ );
 796         pOrTerm->wtFlags &= ~TERM_OK;
 797         if( pOrTerm->leftCursor==iCursor ){
 798           /* This is the 2-bit case and we are on the second iteration and
 799           ** current term is from the first iteration.  So skip this term. */
 800           assert( j==1 );
 801           continue;
 802         }
 803         if( (chngToIN & sqlite3WhereGetMask(&pWInfo->sMaskSet,
 804                                             pOrTerm->leftCursor))==0 ){
 805           /* This term must be of the form t1.a==t2.b where t2 is in the
 806           ** chngToIN set but t1 is not.  This term will be either preceded
 807           ** or follwed by an inverted copy (t2.b==t1.a).  Skip this term
 808           ** and use its inversion. */
 809           testcase( pOrTerm->wtFlags & TERM_COPIED );
 810           testcase( pOrTerm->wtFlags & TERM_VIRTUAL );
 811           assert( pOrTerm->wtFlags & (TERM_COPIED|TERM_VIRTUAL) );
 812           continue;
 813         }
 814         assert( (pOrTerm->eOperator & (WO_OR|WO_AND))==0 );
 815         iColumn = pOrTerm->u.x.leftColumn;
 816         iCursor = pOrTerm->leftCursor;
 817         pLeft = pOrTerm->pExpr->pLeft;
 818         break;
 819       }
 820       if( i<0 ){
 821         /* No candidate table+column was found.  This can only occur
 822         ** on the second iteration */
 823         assert( j==1 );
 824         assert( IsPowerOfTwo(chngToIN) );
 825         assert( chngToIN==sqlite3WhereGetMask(&pWInfo->sMaskSet, iCursor) );
 826         break;
 827       }
 828       testcase( j==1 );
 829
 830       /* We have found a candidate table and column.  Check to see if that
 831       ** table and column is common to every term in the OR clause */
 832       okToChngToIN = 1;
 833       for(; i>=0 && okToChngToIN; i--, pOrTerm++){
 834         assert( pOrTerm->eOperator & WO_EQ );
 835         assert( (pOrTerm->eOperator & (WO_OR|WO_AND))==0 );
 836         if( pOrTerm->leftCursor!=iCursor ){
 837           pOrTerm->wtFlags &= ~TERM_OK;
 838         }else if( pOrTerm->u.x.leftColumn!=iColumn || (iColumn==XN_EXPR
 839                && sqlite3ExprCompare(pParse, pOrTerm->pExpr->pLeft, pLeft, -1)
 840         )){
 841           okToChngToIN = 0;
 842         }else{
 843           int affLeft, affRight;
 844           /* If the right-hand side is also a column, then the affinities
 845           ** of both right and left sides must be such that no type
 846           ** conversions are required on the right.  (Ticket #2249)
 847           */
 848           affRight = sqlite3ExprAffinity(pOrTerm->pExpr->pRight);
 849           affLeft = sqlite3ExprAffinity(pOrTerm->pExpr->pLeft);
 850           if( affRight!=0 && affRight!=affLeft ){
 851             okToChngToIN = 0;
 852           }else{
 853             pOrTerm->wtFlags |= TERM_OK;
 854           }
 855         }
 856       }
 857     }
 858
 859     /* At this point, okToChngToIN is true if original pTerm satisfies
 860     ** case 1.  In that case, construct a new virtual term that is
 861     ** pTerm converted into an IN operator.
 862     */
 863     if( okToChngToIN ){
 864       Expr *pDup;            /* A transient duplicate expression */
 865       ExprList *pList = 0;   /* The RHS of the IN operator */
 866       Expr *pLeft = 0;       /* The LHS of the IN operator */
 867       Expr *pNew;            /* The complete IN operator */
 868
 869       for(i=pOrWc->nTerm-1, pOrTerm=pOrWc->a; i>=0; i--, pOrTerm++){
 870         if( (pOrTerm->wtFlags & TERM_OK)==0 ) continue;
 871         assert( pOrTerm->eOperator & WO_EQ );
 872         assert( (pOrTerm->eOperator & (WO_OR|WO_AND))==0 );
 873         assert( pOrTerm->leftCursor==iCursor );
 874         assert( pOrTerm->u.x.leftColumn==iColumn );
 875         pDup = sqlite3ExprDup(db, pOrTerm->pExpr->pRight, 0);
 876         pList = sqlite3ExprListAppend(pWInfo->pParse, pList, pDup);
 877         pLeft = pOrTerm->pExpr->pLeft;
 878       }
 879       assert( pLeft!=0 );
 880       pDup = sqlite3ExprDup(db, pLeft, 0);
 881       pNew = sqlite3PExpr(pParse, TK_IN, pDup, 0);
 882       if( pNew ){
 883         int idxNew;
 884         transferJoinMarkings(pNew, pExpr);
 885         assert( ExprUseXList(pNew) );
 886         pNew->x.pList = pList;
 887         idxNew = whereClauseInsert(pWC, pNew, TERM_VIRTUAL|TERM_DYNAMIC);
 888         testcase( idxNew==0 );
 889         exprAnalyze(pSrc, pWC, idxNew);
 890         /* pTerm = &pWC->a[idxTerm]; // would be needed if pTerm where reused */
 891         markTermAsChild(pWC, idxNew, idxTerm);
 892       }else{
 893         sqlite3ExprListDelete(db, pList);
 894       }
 895     }
 896   }
 897 }
 898 #endif /* !SQLITE_OMIT_OR_OPTIMIZATION && !SQLITE_OMIT_SUBQUERY */
 899
 900 /*
 901 ** We already know that pExpr is a binary operator where both operands are
 902 ** column references.  This routine checks to see if pExpr is an equivalence
 903 ** relation:
 904 **   1.  The SQLITE_Transitive optimization must be enabled
 905 **   2.  Must be either an == or an IS operator
 906 **   3.  Not originating in the ON clause of an OUTER JOIN
 907 **   4.  The affinities of A and B must be compatible
 908 **   5a. Both operands use the same collating sequence OR
 909 **   5b. The overall collating sequence is BINARY
 910 ** If this routine returns TRUE, that means that the RHS can be substituted
 911 ** for the LHS anyplace else in the WHERE clause where the LHS column occurs.
 912 ** This is an optimization.  No harm comes from returning 0.  But if 1 is
 913 ** returned when it should not be, then incorrect answers might result.
 914 */
 915 static int termIsEquivalence(Parse *pParse, Expr *pExpr){
 916   char aff1, aff2;
 917   CollSeq *pColl;
 918   if( !OptimizationEnabled(pParse->db, SQLITE_Transitive) ) return 0;
 919   if( pExpr->op!=TK_EQ && pExpr->op!=TK_IS ) return 0;
 920   if( ExprHasProperty(pExpr, EP_OuterON) ) return 0;
 921   aff1 = sqlite3ExprAffinity(pExpr->pLeft);
 922   aff2 = sqlite3ExprAffinity(pExpr->pRight);
 923   if( aff1!=aff2
 924    && (!sqlite3IsNumericAffinity(aff1) || !sqlite3IsNumericAffinity(aff2))
 925   ){
 926     return 0;
 927   }
 928   pColl = sqlite3ExprCompareCollSeq(pParse, pExpr);
 929   if( sqlite3IsBinary(pColl) ) return 1;
 930   return sqlite3ExprCollSeqMatch(pParse, pExpr->pLeft, pExpr->pRight);
 931 }
 932
 933 /*
 934 ** Recursively walk the expressions of a SELECT statement and generate
 935 ** a bitmask indicating which tables are used in that expression
 936 ** tree.
 937 */
 938 static Bitmask exprSelectUsage(WhereMaskSet *pMaskSet, Select *pS){
 939   Bitmask mask = 0;
 940   while( pS ){
 941     SrcList *pSrc = pS->pSrc;
 942     mask |= sqlite3WhereExprListUsage(pMaskSet, pS->pEList);
 943     mask |= sqlite3WhereExprListUsage(pMaskSet, pS->pGroupBy);
 944     mask |= sqlite3WhereExprListUsage(pMaskSet, pS->pOrderBy);
 945     mask |= sqlite3WhereExprUsage(pMaskSet, pS->pWhere);
 946     mask |= sqlite3WhereExprUsage(pMaskSet, pS->pHaving);
 947     if( ALWAYS(pSrc!=0) ){
 948       int i;
 949       for(i=0; i<pSrc->nSrc; i++){
 950         mask |= exprSelectUsage(pMaskSet, pSrc->a[i].pSelect);
 951         if( pSrc->a[i].fg.isUsing==0 ){
 952           mask |= sqlite3WhereExprUsage(pMaskSet, pSrc->a[i].u3.pOn);
 953         }
 954         if( pSrc->a[i].fg.isTabFunc ){
 955           mask |= sqlite3WhereExprListUsage(pMaskSet, pSrc->a[i].u1.pFuncArg);
 956         }
 957       }
 958     }
 959     pS = pS->pPrior;
 960   }
 961   return mask;
 962 }
 963
 964 /*
 965 ** Expression pExpr is one operand of a comparison operator that might
 966 ** be useful for indexing.  This routine checks to see if pExpr appears
 967 ** in any index.  Return TRUE (1) if pExpr is an indexed term and return
 968 ** FALSE (0) if not.  If TRUE is returned, also set aiCurCol[0] to the cursor
 969 ** number of the table that is indexed and aiCurCol[1] to the column number
 970 ** of the column that is indexed, or XN_EXPR (-2) if an expression is being
 971 ** indexed.
 972 **
 973 ** If pExpr is a TK_COLUMN column reference, then this routine always returns
 974 ** true even if that particular column is not indexed, because the column
 975 ** might be added to an automatic index later.
 976 */
 977 static SQLITE_NOINLINE int exprMightBeIndexed2(
 978   SrcList *pFrom,        /* The FROM clause */
 979   Bitmask mPrereq,       /* Bitmask of FROM clause terms referenced by pExpr */
 980   int *aiCurCol,         /* Write the referenced table cursor and column here */
 981   Expr *pExpr            /* An operand of a comparison operator */
 982 ){
 983   Index *pIdx;
 984   int i;
 985   int iCur;
 986   for(i=0; mPrereq>1; i++, mPrereq>>=1){}
 987   iCur = pFrom->a[i].iCursor;
 988   for(pIdx=pFrom->a[i].pTab->pIndex; pIdx; pIdx=pIdx->pNext){
 989     if( pIdx->aColExpr==0 ) continue;
 990     for(i=0; i<pIdx->nKeyCol; i++){
 991       if( pIdx->aiColumn[i]!=XN_EXPR ) continue;
 992       if( sqlite3ExprCompareSkip(pExpr, pIdx->aColExpr->a[i].pExpr, iCur)==0 ){
 993         aiCurCol[0] = iCur;
 994         aiCurCol[1] = XN_EXPR;
 995         return 1;
 996       }
 997     }
 998   }
 999   return 0;
1000 }
1001 static int exprMightBeIndexed(
1002   SrcList *pFrom,        /* The FROM clause */
1003   Bitmask mPrereq,       /* Bitmask of FROM clause terms referenced by pExpr */
1004   int *aiCurCol,         /* Write the referenced table cursor & column here */
1005   Expr *pExpr,           /* An operand of a comparison operator */
1006   int op                 /* The specific comparison operator */
1007 ){
1008   /* If this expression is a vector to the left or right of a
1009   ** inequality constraint (>, <, >= or <=), perform the processing
1010   ** on the first element of the vector.  */
1011   assert( TK_GT+1==TK_LE && TK_GT+2==TK_LT && TK_GT+3==TK_GE );
1012   assert( TK_IS<TK_GE && TK_ISNULL<TK_GE && TK_IN<TK_GE );
1013   assert( op<=TK_GE );
1014   if( pExpr->op==TK_VECTOR && (op>=TK_GT && ALWAYS(op<=TK_GE)) ){
1015     assert( ExprUseXList(pExpr) );
1016     pExpr = pExpr->x.pList->a[0].pExpr;
1017
1018   }
1019
1020   if( pExpr->op==TK_COLUMN ){
1021     aiCurCol[0] = pExpr->iTable;
1022     aiCurCol[1] = pExpr->iColumn;
1023     return 1;
1024   }
1025   if( mPrereq==0 ) return 0;                 /* No table references */
1026   if( (mPrereq&(mPrereq-1))!=0 ) return 0;   /* Refs more than one table */
1027   return exprMightBeIndexed2(pFrom,mPrereq,aiCurCol,pExpr);
1028 }
1029
1030
1031 /*
1032 ** The input to this routine is an WhereTerm structure with only the
1033 ** "pExpr" field filled in.  The job of this routine is to analyze the
1034 ** subexpression and populate all the other fields of the WhereTerm
1035 ** structure.
1036 **
1037 ** If the expression is of the form "<expr> <op> X" it gets commuted
1038 ** to the standard form of "X <op> <expr>".
1039 **
1040 ** If the expression is of the form "X <op> Y" where both X and Y are
1041 ** columns, then the original expression is unchanged and a new virtual
1042 ** term of the form "Y <op> X" is added to the WHERE clause and
1043 ** analyzed separately.  The original term is marked with TERM_COPIED
1044 ** and the new term is marked with TERM_DYNAMIC (because it's pExpr
1045 ** needs to be freed with the WhereClause) and TERM_VIRTUAL (because it
1046 ** is a commuted copy of a prior term.)  The original term has nChild=1
1047 ** and the copy has idxParent set to the index of the original term.
1048 */
1049 static void exprAnalyze(
1050   SrcList *pSrc,            /* the FROM clause */
1051   WhereClause *pWC,         /* the WHERE clause */
1052   int idxTerm               /* Index of the term to be analyzed */
1053 ){
1054   WhereInfo *pWInfo = pWC->pWInfo; /* WHERE clause processing context */
1055   WhereTerm *pTerm;                /* The term to be analyzed */
1056   WhereMaskSet *pMaskSet;          /* Set of table index masks */
1057   Expr *pExpr;                     /* The expression to be analyzed */
1058   Bitmask prereqLeft;              /* Prerequesites of the pExpr->pLeft */
1059   Bitmask prereqAll;               /* Prerequesites of pExpr */
1060   Bitmask extraRight = 0;          /* Extra dependencies on LEFT JOIN */
1061   Expr *pStr1 = 0;                 /* RHS of LIKE/GLOB operator */
1062   int isComplete = 0;              /* RHS of LIKE/GLOB ends with wildcard */
1063   int noCase = 0;                  /* uppercase equivalent to lowercase */
1064   int op;                          /* Top-level operator.  pExpr->op */
1065   Parse *pParse = pWInfo->pParse;  /* Parsing context */
1066   sqlite3 *db = pParse->db;        /* Database connection */
1067   unsigned char eOp2 = 0;          /* op2 value for LIKE/REGEXP/GLOB */
1068   int nLeft;                       /* Number of elements on left side vector */
1069
1070   if( db->mallocFailed ){
1071     return;
1072   }
1073   assert( pWC->nTerm > idxTerm );
1074   pTerm = &pWC->a[idxTerm];
1075   pMaskSet = &pWInfo->sMaskSet;
1076   pExpr = pTerm->pExpr;
1077   assert( pExpr!=0 ); /* Because malloc() has not failed */
1078   assert( pExpr->op!=TK_AS && pExpr->op!=TK_COLLATE );
1079   pMaskSet->bVarSelect = 0;
1080   prereqLeft = sqlite3WhereExprUsage(pMaskSet, pExpr->pLeft);
1081   op = pExpr->op;
1082   if( op==TK_IN ){
1083     assert( pExpr->pRight==0 );
1084     if( sqlite3ExprCheckIN(pParse, pExpr) ) return;
1085     if( ExprUseXSelect(pExpr) ){
1086       pTerm->prereqRight = exprSelectUsage(pMaskSet, pExpr->x.pSelect);
1087     }else{
1088       pTerm->prereqRight = sqlite3WhereExprListUsage(pMaskSet, pExpr->x.pList);
1089     }
1090     prereqAll = prereqLeft | pTerm->prereqRight;
1091   }else{
1092     pTerm->prereqRight = sqlite3WhereExprUsage(pMaskSet, pExpr->pRight);
1093     if( pExpr->pLeft==0
1094      || ExprHasProperty(pExpr, EP_xIsSelect|EP_IfNullRow)
1095      || pExpr->x.pList!=0
1096     ){
1097       prereqAll = sqlite3WhereExprUsageNN(pMaskSet, pExpr);
1098     }else{
1099       prereqAll = prereqLeft | pTerm->prereqRight;
1100     }
1101   }
1102   if( pMaskSet->bVarSelect ) pTerm->wtFlags |= TERM_VARSELECT;
1103
1104 #ifdef SQLITE_DEBUG
1105   if( prereqAll!=sqlite3WhereExprUsageNN(pMaskSet, pExpr) ){
1106     printf("\n*** Incorrect prereqAll computed for:\n");
1107     sqlite3TreeViewExpr(0,pExpr,0);
1108     assert( 0 );
1109   }
1110 #endif
1111
1112   if( ExprHasProperty(pExpr, EP_OuterON|EP_InnerON) ){
1113     Bitmask x = sqlite3WhereGetMask(pMaskSet, pExpr->w.iJoin);
1114     if( ExprHasProperty(pExpr, EP_OuterON) ){
1115       prereqAll |= x;
1116       extraRight = x-1;  /* ON clause terms may not be used with an index
1117                          ** on left table of a LEFT JOIN.  Ticket #3015 */
1118       if( (prereqAll>>1)>=x ){
1119         sqlite3ErrorMsg(pParse, "ON clause references tables to its right");
1120         return;
1121       }
1122     }else if( (prereqAll>>1)>=x ){
1123       /* The ON clause of an INNER JOIN references a table to its right.
1124       ** Most other SQL database engines raise an error.  But SQLite versions
1125       ** 3.0 through 3.38 just put the ON clause constraint into the WHERE
1126       ** clause and carried on.   Beginning with 3.39, raise an error only
1127       ** if there is a RIGHT or FULL JOIN in the query.  This makes SQLite
1128       ** more like other systems, and also preserves legacy. */
1129       if( ALWAYS(pSrc->nSrc>0) && (pSrc->a[0].fg.jointype & JT_LTORJ)!=0 ){
1130         sqlite3ErrorMsg(pParse, "ON clause references tables to its right");
1131         return;
1132       }
1133       ExprClearProperty(pExpr, EP_InnerON);
1134     }
1135   }
1136   pTerm->prereqAll = prereqAll;
1137   pTerm->leftCursor = -1;
1138   pTerm->iParent = -1;
1139   pTerm->eOperator = 0;
1140   if( allowedOp(op) ){
1141     int aiCurCol[2];
1142     Expr *pLeft = sqlite3ExprSkipCollate(pExpr->pLeft);
1143     Expr *pRight = sqlite3ExprSkipCollate(pExpr->pRight);
1144     u16 opMask = (pTerm->prereqRight & prereqLeft)==0 ? WO_ALL : WO_EQUIV;
1145
1146     if( pTerm->u.x.iField>0 ){
1147       assert( op==TK_IN );
1148       assert( pLeft->op==TK_VECTOR );
1149       assert( ExprUseXList(pLeft) );
1150       pLeft = pLeft->x.pList->a[pTerm->u.x.iField-1].pExpr;
1151     }
1152
1153     if( exprMightBeIndexed(pSrc, prereqLeft, aiCurCol, pLeft, op) ){
1154       pTerm->leftCursor = aiCurCol[0];
1155       assert( (pTerm->eOperator & (WO_OR|WO_AND))==0 );
1156       pTerm->u.x.leftColumn = aiCurCol[1];
1157       pTerm->eOperator = operatorMask(op) & opMask;
1158     }
1159     if( op==TK_IS ) pTerm->wtFlags |= TERM_IS;
1160     if( pRight
1161      && exprMightBeIndexed(pSrc, pTerm->prereqRight, aiCurCol, pRight, op)
1162      && !ExprHasProperty(pRight, EP_FixedCol)
1163     ){
1164       WhereTerm *pNew;
1165       Expr *pDup;
1166       u16 eExtraOp = 0;        /* Extra bits for pNew->eOperator */
1167       assert( pTerm->u.x.iField==0 );
1168       if( pTerm->leftCursor>=0 ){
1169         int idxNew;
1170         pDup = sqlite3ExprDup(db, pExpr, 0);
1171         if( db->mallocFailed ){
1172           sqlite3ExprDelete(db, pDup);
1173           return;
1174         }
1175         idxNew = whereClauseInsert(pWC, pDup, TERM_VIRTUAL|TERM_DYNAMIC);
1176         if( idxNew==0 ) return;
1177         pNew = &pWC->a[idxNew];
1178         markTermAsChild(pWC, idxNew, idxTerm);
1179         if( op==TK_IS ) pNew->wtFlags |= TERM_IS;
1180         pTerm = &pWC->a[idxTerm];
1181         pTerm->wtFlags |= TERM_COPIED;
1182
1183         if( termIsEquivalence(pParse, pDup) ){
1184           pTerm->eOperator |= WO_EQUIV;
1185           eExtraOp = WO_EQUIV;
1186         }
1187       }else{
1188         pDup = pExpr;
1189         pNew = pTerm;
1190       }
1191       pNew->wtFlags |= exprCommute(pParse, pDup);
1192       pNew->leftCursor = aiCurCol[0];
1193       assert( (pTerm->eOperator & (WO_OR|WO_AND))==0 );
1194       pNew->u.x.leftColumn = aiCurCol[1];
1195       testcase( (prereqLeft | extraRight) != prereqLeft );
1196       pNew->prereqRight = prereqLeft | extraRight;
1197       pNew->prereqAll = prereqAll;
1198       pNew->eOperator = (operatorMask(pDup->op) + eExtraOp) & opMask;
1199     }else
1200     if( op==TK_ISNULL
1201      && !ExprHasProperty(pExpr,EP_OuterON)
1202      && 0==sqlite3ExprCanBeNull(pLeft)
1203     ){
1204       assert( !ExprHasProperty(pExpr, EP_IntValue) );
1205       pExpr->op = TK_TRUEFALSE;
1206       pExpr->u.zToken = "false";
1207       ExprSetProperty(pExpr, EP_IsFalse);
1208       pTerm->prereqAll = 0;
1209       pTerm->eOperator = 0;
1210     }
1211   }
1212
1213 #ifndef SQLITE_OMIT_BETWEEN_OPTIMIZATION
1214   /* If a term is the BETWEEN operator, create two new virtual terms
1215   ** that define the range that the BETWEEN implements.  For example:
1216   **
1217   **      a BETWEEN b AND c
1218   **
1219   ** is converted into:
1220   **
1221   **      (a BETWEEN b AND c) AND (a>=b) AND (a<=c)
1222   **
1223   ** The two new terms are added onto the end of the WhereClause object.
1224   ** The new terms are "dynamic" and are children of the original BETWEEN
1225   ** term.  That means that if the BETWEEN term is coded, the children are
1226   ** skipped.  Or, if the children are satisfied by an index, the original
1227   ** BETWEEN term is skipped.
1228   */
1229   else if( pExpr->op==TK_BETWEEN && pWC->op==TK_AND ){
1230     ExprList *pList;
1231     int i;
1232     static const u8 ops[] = {TK_GE, TK_LE};
1233     assert( ExprUseXList(pExpr) );
1234     pList = pExpr->x.pList;
1235     assert( pList!=0 );
1236     assert( pList->nExpr==2 );
1237     for(i=0; i<2; i++){
1238       Expr *pNewExpr;
1239       int idxNew;
1240       pNewExpr = sqlite3PExpr(pParse, ops[i],
1241                              sqlite3ExprDup(db, pExpr->pLeft, 0),
1242                              sqlite3ExprDup(db, pList->a[i].pExpr, 0));
1243       transferJoinMarkings(pNewExpr, pExpr);
1244       idxNew = whereClauseInsert(pWC, pNewExpr, TERM_VIRTUAL|TERM_DYNAMIC);
1245       testcase( idxNew==0 );
1246       exprAnalyze(pSrc, pWC, idxNew);
1247       pTerm = &pWC->a[idxTerm];
1248       markTermAsChild(pWC, idxNew, idxTerm);
1249     }
1250   }
1251 #endif /* SQLITE_OMIT_BETWEEN_OPTIMIZATION */
1252
1253 #if !defined(SQLITE_OMIT_OR_OPTIMIZATION) && !defined(SQLITE_OMIT_SUBQUERY)
1254   /* Analyze a term that is composed of two or more subterms connected by
1255   ** an OR operator.
1256   */
1257   else if( pExpr->op==TK_OR ){
1258     assert( pWC->op==TK_AND );
1259     exprAnalyzeOrTerm(pSrc, pWC, idxTerm);
1260     pTerm = &pWC->a[idxTerm];
1261   }
1262 #endif /* SQLITE_OMIT_OR_OPTIMIZATION */
1263   /* The form "x IS NOT NULL" can sometimes be evaluated more efficiently
1264   ** as "x>NULL" if x is not an INTEGER PRIMARY KEY.  So construct a
1265   ** virtual term of that form.
1266   **
1267   ** The virtual term must be tagged with TERM_VNULL.
1268   */
1269   else if( pExpr->op==TK_NOTNULL ){
1270     if( pExpr->pLeft->op==TK_COLUMN
1271      && pExpr->pLeft->iColumn>=0
1272      && !ExprHasProperty(pExpr, EP_OuterON)
1273     ){
1274       Expr *pNewExpr;
1275       Expr *pLeft = pExpr->pLeft;
1276       int idxNew;
1277       WhereTerm *pNewTerm;
1278
1279       pNewExpr = sqlite3PExpr(pParse, TK_GT,
1280                               sqlite3ExprDup(db, pLeft, 0),
1281                               sqlite3ExprAlloc(db, TK_NULL, 0, 0));
1282
1283       idxNew = whereClauseInsert(pWC, pNewExpr,
1284                                 TERM_VIRTUAL|TERM_DYNAMIC|TERM_VNULL);
1285       if( idxNew ){
1286         pNewTerm = &pWC->a[idxNew];
1287         pNewTerm->prereqRight = 0;
1288         pNewTerm->leftCursor = pLeft->iTable;
1289         pNewTerm->u.x.leftColumn = pLeft->iColumn;
1290         pNewTerm->eOperator = WO_GT;
1291         markTermAsChild(pWC, idxNew, idxTerm);
1292         pTerm = &pWC->a[idxTerm];
1293         pTerm->wtFlags |= TERM_COPIED;
1294         pNewTerm->prereqAll = pTerm->prereqAll;
1295       }
1296     }
1297   }
1298
1299
1300 #ifndef SQLITE_OMIT_LIKE_OPTIMIZATION
1301   /* Add constraints to reduce the search space on a LIKE or GLOB
1302   ** operator.
1303   **
1304   ** A like pattern of the form "x LIKE 'aBc%'" is changed into constraints
1305   **
1306   **          x>='ABC' AND x<'abd' AND x LIKE 'aBc%'
1307   **
1308   ** The last character of the prefix "abc" is incremented to form the
1309   ** termination condition "abd".  If case is not significant (the default
1310   ** for LIKE) then the lower-bound is made all uppercase and the upper-
1311   ** bound is made all lowercase so that the bounds also work when comparing
1312   ** BLOBs.
1313   */
1314   else if( pExpr->op==TK_FUNCTION
1315    && pWC->op==TK_AND
1316    && isLikeOrGlob(pParse, pExpr, &pStr1, &isComplete, &noCase)
1317   ){
1318     Expr *pLeft;       /* LHS of LIKE/GLOB operator */
1319     Expr *pStr2;       /* Copy of pStr1 - RHS of LIKE/GLOB operator */
1320     Expr *pNewExpr1;
1321     Expr *pNewExpr2;
1322     int idxNew1;
1323     int idxNew2;
1324     const char *zCollSeqName;     /* Name of collating sequence */
1325     const u16 wtFlags = TERM_LIKEOPT | TERM_VIRTUAL | TERM_DYNAMIC;
1326
1327     assert( ExprUseXList(pExpr) );
1328     pLeft = pExpr->x.pList->a[1].pExpr;
1329     pStr2 = sqlite3ExprDup(db, pStr1, 0);
1330     assert( pStr1==0 || !ExprHasProperty(pStr1, EP_IntValue) );
1331     assert( pStr2==0 || !ExprHasProperty(pStr2, EP_IntValue) );
1332
1333
1334     /* Convert the lower bound to upper-case and the upper bound to
1335     ** lower-case (upper-case is less than lower-case in ASCII) so that
1336     ** the range constraints also work for BLOBs
1337     */
1338     if( noCase && !pParse->db->mallocFailed ){
1339       int i;
1340       char c;
1341       pTerm->wtFlags |= TERM_LIKE;
1342       for(i=0; (c = pStr1->u.zToken[i])!=0; i++){
1343         pStr1->u.zToken[i] = sqlite3Toupper(c);
1344         pStr2->u.zToken[i] = sqlite3Tolower(c);
1345       }
1346     }
1347
1348     if( !db->mallocFailed ){
1349       u8 c, *pC;       /* Last character before the first wildcard */
1350       pC = (u8*)&pStr2->u.zToken[sqlite3Strlen30(pStr2->u.zToken)-1];
1351       c = *pC;
1352       if( noCase ){
1353         /* The point is to increment the last character before the first
1354         ** wildcard.  But if we increment '@', that will push it into the
1355         ** alphabetic range where case conversions will mess up the
1356         ** inequality.  To avoid this, make sure to also run the full
1357         ** LIKE on all candidate expressions by clearing the isComplete flag
1358         */
1359         if( c=='A'-1 ) isComplete = 0;
1360         c = sqlite3UpperToLower[c];
1361       }
1362       *pC = c + 1;
1363     }
1364     zCollSeqName = noCase ? "NOCASE" : sqlite3StrBINARY;
1365     pNewExpr1 = sqlite3ExprDup(db, pLeft, 0);
1366     pNewExpr1 = sqlite3PExpr(pParse, TK_GE,
1367            sqlite3ExprAddCollateString(pParse,pNewExpr1,zCollSeqName),
1368            pStr1);
1369     transferJoinMarkings(pNewExpr1, pExpr);
1370     idxNew1 = whereClauseInsert(pWC, pNewExpr1, wtFlags);
1371     testcase( idxNew1==0 );
1372     exprAnalyze(pSrc, pWC, idxNew1);
1373     pNewExpr2 = sqlite3ExprDup(db, pLeft, 0);
1374     pNewExpr2 = sqlite3PExpr(pParse, TK_LT,
1375            sqlite3ExprAddCollateString(pParse,pNewExpr2,zCollSeqName),
1376            pStr2);
1377     transferJoinMarkings(pNewExpr2, pExpr);
1378     idxNew2 = whereClauseInsert(pWC, pNewExpr2, wtFlags);
1379     testcase( idxNew2==0 );
1380     exprAnalyze(pSrc, pWC, idxNew2);
1381     pTerm = &pWC->a[idxTerm];
1382     if( isComplete ){
1383       markTermAsChild(pWC, idxNew1, idxTerm);
1384       markTermAsChild(pWC, idxNew2, idxTerm);
1385     }
1386   }
1387 #endif /* SQLITE_OMIT_LIKE_OPTIMIZATION */
1388
1389   /* If there is a vector == or IS term - e.g. "(a, b) == (?, ?)" - create
1390   ** new terms for each component comparison - "a = ?" and "b = ?".  The
1391   ** new terms completely replace the original vector comparison, which is
1392   ** no longer used.
1393   **
1394   ** This is only required if at least one side of the comparison operation
1395   ** is not a sub-select.
1396   **
1397   ** tag-20220128a
1398   */
1399   if( (pExpr->op==TK_EQ || pExpr->op==TK_IS)
1400    && (nLeft = sqlite3ExprVectorSize(pExpr->pLeft))>1
1401    && sqlite3ExprVectorSize(pExpr->pRight)==nLeft
1402    && ( (pExpr->pLeft->flags & EP_xIsSelect)==0
1403      || (pExpr->pRight->flags & EP_xIsSelect)==0)
1404    && pWC->op==TK_AND
1405   ){
1406     int i;
1407     for(i=0; i<nLeft; i++){
1408       int idxNew;
1409       Expr *pNew;
1410       Expr *pLeft = sqlite3ExprForVectorField(pParse, pExpr->pLeft, i, nLeft);
1411       Expr *pRight = sqlite3ExprForVectorField(pParse, pExpr->pRight, i, nLeft);
1412
1413       pNew = sqlite3PExpr(pParse, pExpr->op, pLeft, pRight);
1414       transferJoinMarkings(pNew, pExpr);
1415       idxNew = whereClauseInsert(pWC, pNew, TERM_DYNAMIC|TERM_SLICE);
1416       exprAnalyze(pSrc, pWC, idxNew);
1417     }
1418     pTerm = &pWC->a[idxTerm];
1419     pTerm->wtFlags |= TERM_CODED|TERM_VIRTUAL;  /* Disable the original */
1420     pTerm->eOperator = WO_ROWVAL;
1421   }
1422
1423   /* If there is a vector IN term - e.g. "(a, b) IN (SELECT ...)" - create
1424   ** a virtual term for each vector component. The expression object
1425   ** used by each such virtual term is pExpr (the full vector IN(...)
1426   ** expression). The WhereTerm.u.x.iField variable identifies the index within
1427   ** the vector on the LHS that the virtual term represents.
1428   **
1429   ** This only works if the RHS is a simple SELECT (not a compound) that does
1430   ** not use window functions.
1431   */
1432   else if( pExpr->op==TK_IN
1433    && pTerm->u.x.iField==0
1434    && pExpr->pLeft->op==TK_VECTOR
1435    && ALWAYS( ExprUseXSelect(pExpr) )
1436    && pExpr->x.pSelect->pPrior==0
1437 #ifndef SQLITE_OMIT_WINDOWFUNC
1438    && pExpr->x.pSelect->pWin==0
1439 #endif
1440    && pWC->op==TK_AND
1441   ){
1442     int i;
1443     for(i=0; i<sqlite3ExprVectorSize(pExpr->pLeft); i++){
1444       int idxNew;
1445       idxNew = whereClauseInsert(pWC, pExpr, TERM_VIRTUAL|TERM_SLICE);
1446       pWC->a[idxNew].u.x.iField = i+1;
1447       exprAnalyze(pSrc, pWC, idxNew);
1448       markTermAsChild(pWC, idxNew, idxTerm);
1449     }
1450   }
1451
1452 #ifndef SQLITE_OMIT_VIRTUALTABLE
1453   /* Add a WO_AUX auxiliary term to the constraint set if the
1454   ** current expression is of the form "column OP expr" where OP
1455   ** is an operator that gets passed into virtual tables but which is
1456   ** not normally optimized for ordinary tables.  In other words, OP
1457   ** is one of MATCH, LIKE, GLOB, REGEXP, !=, IS, IS NOT, or NOT NULL.
1458   ** This information is used by the xBestIndex methods of
1459   ** virtual tables.  The native query optimizer does not attempt
1460   ** to do anything with MATCH functions.
1461   */
1462   else if( pWC->op==TK_AND ){
1463     Expr *pRight = 0, *pLeft = 0;
1464     int res = isAuxiliaryVtabOperator(db, pExpr, &eOp2, &pLeft, &pRight);
1465     while( res-- > 0 ){
1466       int idxNew;
1467       WhereTerm *pNewTerm;
1468       Bitmask prereqColumn, prereqExpr;
1469
1470       prereqExpr = sqlite3WhereExprUsage(pMaskSet, pRight);
1471       prereqColumn = sqlite3WhereExprUsage(pMaskSet, pLeft);
1472       if( (prereqExpr & prereqColumn)==0 ){
1473         Expr *pNewExpr;
1474         pNewExpr = sqlite3PExpr(pParse, TK_MATCH,
1475             0, sqlite3ExprDup(db, pRight, 0));
1476         if( ExprHasProperty(pExpr, EP_OuterON) && pNewExpr ){
1477           ExprSetProperty(pNewExpr, EP_OuterON);
1478           pNewExpr->w.iJoin = pExpr->w.iJoin;
1479         }
1480         idxNew = whereClauseInsert(pWC, pNewExpr, TERM_VIRTUAL|TERM_DYNAMIC);
1481         testcase( idxNew==0 );
1482         pNewTerm = &pWC->a[idxNew];
1483         pNewTerm->prereqRight = prereqExpr;
1484         pNewTerm->leftCursor = pLeft->iTable;
1485         pNewTerm->u.x.leftColumn = pLeft->iColumn;
1486         pNewTerm->eOperator = WO_AUX;
1487         pNewTerm->eMatchOp = eOp2;
1488         markTermAsChild(pWC, idxNew, idxTerm);
1489         pTerm = &pWC->a[idxTerm];
1490         pTerm->wtFlags |= TERM_COPIED;
1491         pNewTerm->prereqAll = pTerm->prereqAll;
1492       }
1493       SWAP(Expr*, pLeft, pRight);
1494     }
1495   }
1496 #endif /* SQLITE_OMIT_VIRTUALTABLE */
1497
1498   /* Prevent ON clause terms of a LEFT JOIN from being used to drive
1499   ** an index for tables to the left of the join.
1500   */
1501   testcase( pTerm!=&pWC->a[idxTerm] );
1502   pTerm = &pWC->a[idxTerm];
1503   pTerm->prereqRight |= extraRight;
1504 }
1505
1506 /***************************************************************************
1507 ** Routines with file scope above.  Interface to the rest of the where.c
1508 ** subsystem follows.
1509 ***************************************************************************/
1510
1511 /*
1512 ** This routine identifies subexpressions in the WHERE clause where
1513 ** each subexpression is separated by the AND operator or some other
1514 ** operator specified in the op parameter.  The WhereClause structure
1515 ** is filled with pointers to subexpressions.  For example:
1516 **
1517 **    WHERE  a=='hello' AND coalesce(b,11)<10 AND (c+12!=d OR c==22)
1518 **           \________/     \_______________/     \________________/
1519 **            slot[0]            slot[1]               slot[2]
1520 **
1521 ** The original WHERE clause in pExpr is unaltered.  All this routine
1522 ** does is make slot[] entries point to substructure within pExpr.
1523 **
1524 ** In the previous sentence and in the diagram, "slot[]" refers to
1525 ** the WhereClause.a[] array.  The slot[] array grows as needed to contain
1526 ** all terms of the WHERE clause.
1527 */
1528 void sqlite3WhereSplit(WhereClause *pWC, Expr *pExpr, u8 op){
1529   Expr *pE2 = sqlite3ExprSkipCollateAndLikely(pExpr);
1530   pWC->op = op;
1531   assert( pE2!=0 || pExpr==0 );
1532   if( pE2==0 ) return;
1533   if( pE2->op!=op ){
1534     whereClauseInsert(pWC, pExpr, 0);
1535   }else{
1536     sqlite3WhereSplit(pWC, pE2->pLeft, op);
1537     sqlite3WhereSplit(pWC, pE2->pRight, op);
1538   }
1539 }
1540
1541 /*
1542 ** Add either a LIMIT (if eMatchOp==SQLITE_INDEX_CONSTRAINT_LIMIT) or
1543 ** OFFSET (if eMatchOp==SQLITE_INDEX_CONSTRAINT_OFFSET) term to the
1544 ** where-clause passed as the first argument. The value for the term
1545 ** is found in register iReg.
1546 **
1547 ** In the common case where the value is a simple integer
1548 ** (example: "LIMIT 5 OFFSET 10") then the expression codes as a
1549 ** TK_INTEGER so that it will be available to sqlite3_vtab_rhs_value().
1550 ** If not, then it codes as a TK_REGISTER expression.
1551 */
1552 static void whereAddLimitExpr(
1553   WhereClause *pWC,   /* Add the constraint to this WHERE clause */
1554   int iReg,           /* Register that will hold value of the limit/offset */
1555   Expr *pExpr,        /* Expression that defines the limit/offset */
1556   int iCsr,           /* Cursor to which the constraint applies */
1557   int eMatchOp        /* SQLITE_INDEX_CONSTRAINT_LIMIT or _OFFSET */
1558 ){
1559   Parse *pParse = pWC->pWInfo->pParse;
1560   sqlite3 *db = pParse->db;
1561   Expr *pNew;
1562   int iVal = 0;
1563
1564   if( sqlite3ExprIsInteger(pExpr, &iVal) && iVal>=0 ){
1565     Expr *pVal = sqlite3Expr(db, TK_INTEGER, 0);
1566     if( pVal==0 ) return;
1567     ExprSetProperty(pVal, EP_IntValue);
1568     pVal->u.iValue = iVal;
1569     pNew = sqlite3PExpr(pParse, TK_MATCH, 0, pVal);
1570   }else{
1571     Expr *pVal = sqlite3Expr(db, TK_REGISTER, 0);
1572     if( pVal==0 ) return;
1573     pVal->iTable = iReg;
1574     pNew = sqlite3PExpr(pParse, TK_MATCH, 0, pVal);
1575   }
1576   if( pNew ){
1577     WhereTerm *pTerm;
1578     int idx;
1579     idx = whereClauseInsert(pWC, pNew, TERM_DYNAMIC|TERM_VIRTUAL);
1580     pTerm = &pWC->a[idx];
1581     pTerm->leftCursor = iCsr;
1582     pTerm->eOperator = WO_AUX;
1583     pTerm->eMatchOp = eMatchOp;
1584   }
1585 }
1586
1587 /*
1588 ** Possibly add terms corresponding to the LIMIT and OFFSET clauses of the
1589 ** SELECT statement passed as the second argument. These terms are only
1590 ** added if:
1591 **
1592 **   1. The SELECT statement has a LIMIT clause, and
1593 **   2. The SELECT statement is not an aggregate or DISTINCT query, and
1594 **   3. The SELECT statement has exactly one object in its from clause, and
1595 **      that object is a virtual table, and
1596 **   4. There are no terms in the WHERE clause that will not be passed
1597 **      to the virtual table xBestIndex method.
1598 **   5. The ORDER BY clause, if any, will be made available to the xBestIndex
1599 **      method.
1600 **
1601 ** LIMIT and OFFSET terms are ignored by most of the planner code. They
1602 ** exist only so that they may be passed to the xBestIndex method of the
1603 ** single virtual table in the FROM clause of the SELECT.
1604 */
1605 void sqlite3WhereAddLimit(WhereClause *pWC, Select *p){
1606   assert( p==0 || (p->pGroupBy==0 && (p->selFlags & SF_Aggregate)==0) );
1607   if( (p && p->pLimit)                                          /* 1 */
1608    && (p->selFlags & (SF_Distinct|SF_Aggregate))==0             /* 2 */
1609    && (p->pSrc->nSrc==1 && IsVirtual(p->pSrc->a[0].pTab))       /* 3 */
1610   ){
1611     ExprList *pOrderBy = p->pOrderBy;
1612     int iCsr = p->pSrc->a[0].iCursor;
1613     int ii;
1614
1615     /* Check condition (4). Return early if it is not met. */
1616     for(ii=0; ii<pWC->nTerm; ii++){
1617       if( pWC->a[ii].wtFlags & TERM_CODED ){
1618         /* This term is a vector operation that has been decomposed into
1619         ** other, subsequent terms.  It can be ignored. See tag-20220128a */
1620         assert( pWC->a[ii].wtFlags & TERM_VIRTUAL );
1621         assert( pWC->a[ii].eOperator==WO_ROWVAL );
1622         continue;
1623       }
1624       if( pWC->a[ii].leftCursor!=iCsr ) return;
1625     }
1626
1627     /* Check condition (5). Return early if it is not met. */
1628     if( pOrderBy ){
1629       for(ii=0; ii<pOrderBy->nExpr; ii++){
1630         Expr *pExpr = pOrderBy->a[ii].pExpr;
1631         if( pExpr->op!=TK_COLUMN ) return;
1632         if( pExpr->iTable!=iCsr ) return;
1633         if( pOrderBy->a[ii].fg.sortFlags & KEYINFO_ORDER_BIGNULL ) return;
1634       }
1635     }
1636
1637     /* All conditions are met. Add the terms to the where-clause object. */
1638     assert( p->pLimit->op==TK_LIMIT );
1639     whereAddLimitExpr(pWC, p->iLimit, p->pLimit->pLeft,
1640                       iCsr, SQLITE_INDEX_CONSTRAINT_LIMIT);
1641     if( p->iOffset>0 ){
1642       whereAddLimitExpr(pWC, p->iOffset, p->pLimit->pRight,
1643                         iCsr, SQLITE_INDEX_CONSTRAINT_OFFSET);
1644     }
1645   }
1646 }
1647
1648 /*
1649 ** Initialize a preallocated WhereClause structure.
1650 */
1651 void sqlite3WhereClauseInit(
1652   WhereClause *pWC,        /* The WhereClause to be initialized */
1653   WhereInfo *pWInfo        /* The WHERE processing context */
1654 ){
1655   pWC->pWInfo = pWInfo;
1656   pWC->hasOr = 0;
1657   pWC->pOuter = 0;
1658   pWC->nTerm = 0;
1659   pWC->nBase = 0;
1660   pWC->nSlot = ArraySize(pWC->aStatic);
1661   pWC->a = pWC->aStatic;
1662 }
1663
1664 /*
1665 ** Deallocate a WhereClause structure.  The WhereClause structure
1666 ** itself is not freed.  This routine is the inverse of
1667 ** sqlite3WhereClauseInit().
1668 */
1669 void sqlite3WhereClauseClear(WhereClause *pWC){
1670   sqlite3 *db = pWC->pWInfo->pParse->db;
1671   assert( pWC->nTerm>=pWC->nBase );
1672   if( pWC->nTerm>0 ){
1673     WhereTerm *a = pWC->a;
1674     WhereTerm *aLast = &pWC->a[pWC->nTerm-1];
1675 #ifdef SQLITE_DEBUG
1676     int i;
1677     /* Verify that every term past pWC->nBase is virtual */
1678     for(i=pWC->nBase; i<pWC->nTerm; i++){
1679       assert( (pWC->a[i].wtFlags & TERM_VIRTUAL)!=0 );
1680     }
1681 #endif
1682     while(1){
1683       assert( a->eMatchOp==0 || a->eOperator==WO_AUX );
1684       if( a->wtFlags & TERM_DYNAMIC ){
1685         sqlite3ExprDelete(db, a->pExpr);
1686       }
1687       if( a->wtFlags & (TERM_ORINFO|TERM_ANDINFO) ){
1688         if( a->wtFlags & TERM_ORINFO ){
1689           assert( (a->wtFlags & TERM_ANDINFO)==0 );
1690           whereOrInfoDelete(db, a->u.pOrInfo);
1691         }else{
1692           assert( (a->wtFlags & TERM_ANDINFO)!=0 );
1693           whereAndInfoDelete(db, a->u.pAndInfo);
1694         }
1695       }
1696       if( a==aLast ) break;
1697       a++;
1698     }
1699   }
1700 }
1701
1702
1703 /*
1704 ** These routines walk (recursively) an expression tree and generate
1705 ** a bitmask indicating which tables are used in that expression
1706 ** tree.
1707 **
1708 ** sqlite3WhereExprUsage(MaskSet, Expr) ->
1709 **
1710 **       Return a Bitmask of all tables referenced by Expr.  Expr can be
1711 **       be NULL, in which case 0 is returned.
1712 **
1713 ** sqlite3WhereExprUsageNN(MaskSet, Expr) ->
1714 **
1715 **       Same as sqlite3WhereExprUsage() except that Expr must not be
1716 **       NULL.  The "NN" suffix on the name stands for "Not Null".
1717 **
1718 ** sqlite3WhereExprListUsage(MaskSet, ExprList) ->
1719 **
1720 **       Return a Bitmask of all tables referenced by every expression
1721 **       in the expression list ExprList.  ExprList can be NULL, in which
1722 **       case 0 is returned.
1723 **
1724 ** sqlite3WhereExprUsageFull(MaskSet, ExprList) ->
1725 **
1726 **       Internal use only.  Called only by sqlite3WhereExprUsageNN() for
1727 **       complex expressions that require pushing register values onto
1728 **       the stack.  Many calls to sqlite3WhereExprUsageNN() do not need
1729 **       the more complex analysis done by this routine.  Hence, the
1730 **       computations done by this routine are broken out into a separate
1731 **       "no-inline" function to avoid the stack push overhead in the
1732 **       common case where it is not needed.
1733 */
1734 static SQLITE_NOINLINE Bitmask sqlite3WhereExprUsageFull(
1735   WhereMaskSet *pMaskSet,
1736   Expr *p
1737 ){
1738   Bitmask mask;
1739   mask = (p->op==TK_IF_NULL_ROW) ? sqlite3WhereGetMask(pMaskSet, p->iTable) : 0;
1740   if( p->pLeft ) mask |= sqlite3WhereExprUsageNN(pMaskSet, p->pLeft);
1741   if( p->pRight ){
1742     mask |= sqlite3WhereExprUsageNN(pMaskSet, p->pRight);
1743     assert( p->x.pList==0 );
1744   }else if( ExprUseXSelect(p) ){
1745     if( ExprHasProperty(p, EP_VarSelect) ) pMaskSet->bVarSelect = 1;
1746     mask |= exprSelectUsage(pMaskSet, p->x.pSelect);
1747   }else if( p->x.pList ){
1748     mask |= sqlite3WhereExprListUsage(pMaskSet, p->x.pList);
1749   }
1750 #ifndef SQLITE_OMIT_WINDOWFUNC
1751   if( (p->op==TK_FUNCTION || p->op==TK_AGG_FUNCTION) && ExprUseYWin(p) ){
1752     assert( p->y.pWin!=0 );
1753     mask |= sqlite3WhereExprListUsage(pMaskSet, p->y.pWin->pPartition);
1754     mask |= sqlite3WhereExprListUsage(pMaskSet, p->y.pWin->pOrderBy);
1755     mask |= sqlite3WhereExprUsage(pMaskSet, p->y.pWin->pFilter);
1756   }
1757 #endif
1758   return mask;
1759 }
1760 Bitmask sqlite3WhereExprUsageNN(WhereMaskSet *pMaskSet, Expr *p){
1761   if( p->op==TK_COLUMN && !ExprHasProperty(p, EP_FixedCol) ){
1762     return sqlite3WhereGetMask(pMaskSet, p->iTable);
1763   }else if( ExprHasProperty(p, EP_TokenOnly|EP_Leaf) ){
1764     assert( p->op!=TK_IF_NULL_ROW );
1765     return 0;
1766   }
1767   return sqlite3WhereExprUsageFull(pMaskSet, p);
1768 }
1769 Bitmask sqlite3WhereExprUsage(WhereMaskSet *pMaskSet, Expr *p){
1770   return p ? sqlite3WhereExprUsageNN(pMaskSet,p) : 0;
1771 }
1772 Bitmask sqlite3WhereExprListUsage(WhereMaskSet *pMaskSet, ExprList *pList){
1773   int i;
1774   Bitmask mask = 0;
1775   if( pList ){
1776     for(i=0; i<pList->nExpr; i++){
1777       mask |= sqlite3WhereExprUsage(pMaskSet, pList->a[i].pExpr);
1778     }
1779   }
1780   return mask;
1781 }
1782
1783
1784 /*
1785 ** Call exprAnalyze on all terms in a WHERE clause.
1786 **
1787 ** Note that exprAnalyze() might add new virtual terms onto the
1788 ** end of the WHERE clause.  We do not want to analyze these new
1789 ** virtual terms, so start analyzing at the end and work forward
1790 ** so that the added virtual terms are never processed.
1791 */
1792 void sqlite3WhereExprAnalyze(
1793   SrcList *pTabList,       /* the FROM clause */
1794   WhereClause *pWC         /* the WHERE clause to be analyzed */
1795 ){
1796   int i;
1797   for(i=pWC->nTerm-1; i>=0; i--){
1798     exprAnalyze(pTabList, pWC, i);
1799   }
1800 }
1801
1802 /*
1803 ** For table-valued-functions, transform the function arguments into
1804 ** new WHERE clause terms.
1805 **
1806 ** Each function argument translates into an equality constraint against
1807 ** a HIDDEN column in the table.
1808 */
1809 void sqlite3WhereTabFuncArgs(
1810   Parse *pParse,                    /* Parsing context */
1811   SrcItem *pItem,                   /* The FROM clause term to process */
1812   WhereClause *pWC                  /* Xfer function arguments to here */
1813 ){
1814   Table *pTab;
1815   int j, k;
1816   ExprList *pArgs;
1817   Expr *pColRef;
1818   Expr *pTerm;
1819   if( pItem->fg.isTabFunc==0 ) return;
1820   pTab = pItem->pTab;
1821   assert( pTab!=0 );
1822   pArgs = pItem->u1.pFuncArg;
1823   if( pArgs==0 ) return;
1824   for(j=k=0; j<pArgs->nExpr; j++){
1825     Expr *pRhs;
1826     u32 joinType;
1827     while( k<pTab->nCol && (pTab->aCol[k].colFlags & COLFLAG_HIDDEN)==0 ){k++;}
1828     if( k>=pTab->nCol ){
1829       sqlite3ErrorMsg(pParse, "too many arguments on %s() - max %d",
1830                       pTab->zName, j);
1831       return;
1832     }
1833     pColRef = sqlite3ExprAlloc(pParse->db, TK_COLUMN, 0, 0);
1834     if( pColRef==0 ) return;
1835     pColRef->iTable = pItem->iCursor;
1836     pColRef->iColumn = k++;
1837     assert( ExprUseYTab(pColRef) );
1838     pColRef->y.pTab = pTab;
1839     pItem->colUsed |= sqlite3ExprColUsed(pColRef);
1840     pRhs = sqlite3PExpr(pParse, TK_UPLUS,
1841         sqlite3ExprDup(pParse->db, pArgs->a[j].pExpr, 0), 0);
1842     pTerm = sqlite3PExpr(pParse, TK_EQ, pColRef, pRhs);
1843     if( pItem->fg.jointype & (JT_LEFT|JT_LTORJ) ){
1844       joinType = EP_OuterON;
1845     }else{
1846       joinType = EP_InnerON;
1847     }
1848     sqlite3SetJoinExpr(pTerm, pItem->iCursor, joinType);
1849     whereClauseInsert(pWC, pTerm, TERM_DYNAMIC);
1850   }
1851 }