src/backend/parser/parse_node.c

   1 /*-------------------------------------------------------------------------
   2  *
   3  * parse_node.c
   4  *        various routines that make nodes for querytrees
   5  *
   6  * Portions Copyright (c) 1996-2009, PostgreSQL Global Development Group
   7  * Portions Copyright (c) 1994, Regents of the University of California
   8  *
   9  *
  10  * IDENTIFICATION
  11  *        $PostgreSQL$
  12  *
  13  *-------------------------------------------------------------------------
  14  */
  15 #include "postgres.h"
  16
  17 #include "access/heapam.h"
  18 #include "catalog/pg_type.h"
  19 #include "mb/pg_wchar.h"
  20 #include "nodes/makefuncs.h"
  21 #include "nodes/nodeFuncs.h"
  22 #include "parser/parsetree.h"
  23 #include "parser/parse_coerce.h"
  24 #include "parser/parse_expr.h"
  25 #include "parser/parse_relation.h"
  26 #include "utils/builtins.h"
  27 #include "utils/int8.h"
  28 #include "utils/syscache.h"
  29 #include "utils/varbit.h"
  30
  31
  32 static void pcb_error_callback(void *arg);
  33
  34
  35 /*
  36  * make_parsestate
  37  *              Allocate and initialize a new ParseState.
  38  *
  39  * Caller should eventually release the ParseState via free_parsestate().
  40  */
  41 ParseState *
  42 make_parsestate(ParseState *parentParseState)
  43 {
  44         ParseState *pstate;
  45
  46         pstate = palloc0(sizeof(ParseState));
  47
  48         pstate->parentParseState = parentParseState;
  49
  50         /* Fill in fields that don't start at null/false/zero */
  51         pstate->p_next_resno = 1;
  52
  53         if (parentParseState)
  54         {
  55                 pstate->p_sourcetext = parentParseState->p_sourcetext;
  56                 pstate->p_variableparams = parentParseState->p_variableparams;
  57         }
  58
  59         return pstate;
  60 }
  61
  62 /*
  63  * free_parsestate
  64  *              Release a ParseState and any subsidiary resources.
  65  */
  66 void
  67 free_parsestate(ParseState *pstate)
  68 {
  69         /*
  70          * Check that we did not produce too many resnos; at the very least we
  71          * cannot allow more than 2^16, since that would exceed the range of a
  72          * AttrNumber. It seems safest to use MaxTupleAttributeNumber.
  73          */
  74         if (pstate->p_next_resno - 1 > MaxTupleAttributeNumber)
  75                 ereport(ERROR,
  76                                 (errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
  77                                  errmsg("target lists can have at most %d entries",
  78                                                 MaxTupleAttributeNumber)));
  79
  80         if (pstate->p_target_relation != NULL)
  81                 heap_close(pstate->p_target_relation, NoLock);
  82
  83         pfree(pstate);
  84 }
  85
  86
  87 /*
  88  * parser_errposition
  89  *              Report a parse-analysis-time cursor position, if possible.
  90  *
  91  * This is expected to be used within an ereport() call.  The return value
  92  * is a dummy (always 0, in fact).
  93  *
  94  * The locations stored in raw parsetrees are byte offsets into the source
  95  * string.      We have to convert them to 1-based character indexes for reporting
  96  * to clients.  (We do things this way to avoid unnecessary overhead in the
  97  * normal non-error case: computing character indexes would be much more
  98  * expensive than storing token offsets.)
  99  */
 100 int
 101 parser_errposition(ParseState *pstate, int location)
 102 {
 103         int                     pos;
 104
 105         /* No-op if location was not provided */
 106         if (location < 0)
 107                 return 0;
 108         /* Can't do anything if source text is not available */
 109         if (pstate == NULL || pstate->p_sourcetext == NULL)
 110                 return 0;
 111         /* Convert offset to character number */
 112         pos = pg_mbstrlen_with_len(pstate->p_sourcetext, location) + 1;
 113         /* And pass it to the ereport mechanism */
 114         return errposition(pos);
 115 }
 116
 117
 118 /*
 119  * setup_parser_errposition_callback
 120  *              Arrange for non-parser errors to report an error position
 121  *
 122  * Sometimes the parser calls functions that aren't part of the parser
 123  * subsystem and can't reasonably be passed a ParseState; yet we would
 124  * like any errors thrown in those functions to be tagged with a parse
 125  * error location.      Use this function to set up an error context stack
 126  * entry that will accomplish that.  Usage pattern:
 127  *
 128  *              declare a local variable "ParseCallbackState pcbstate"
 129  *              ...
 130  *              setup_parser_errposition_callback(&pcbstate, pstate, location);
 131  *              call function that might throw error;
 132  *              cancel_parser_errposition_callback(&pcbstate);
 133  */
 134 void
 135 setup_parser_errposition_callback(ParseCallbackState *pcbstate,
 136                                                                   ParseState *pstate, int location)
 137 {
 138         /* Setup error traceback support for ereport() */
 139         pcbstate->pstate = pstate;
 140         pcbstate->location = location;
 141         pcbstate->errcontext.callback = pcb_error_callback;
 142         pcbstate->errcontext.arg = (void *) pcbstate;
 143         pcbstate->errcontext.previous = error_context_stack;
 144         error_context_stack = &pcbstate->errcontext;
 145 }
 146
 147 /*
 148  * Cancel a previously-set-up errposition callback.
 149  */
 150 void
 151 cancel_parser_errposition_callback(ParseCallbackState *pcbstate)
 152 {
 153         /* Pop the error context stack */
 154         error_context_stack = pcbstate->errcontext.previous;
 155 }
 156
 157 /*
 158  * Error context callback for inserting parser error location.
 159  *
 160  * Note that this will be called for *any* error occurring while the
 161  * callback is installed.  We avoid inserting an irrelevant error location
 162  * if the error is a query cancel --- are there any other important cases?
 163  */
 164 static void
 165 pcb_error_callback(void *arg)
 166 {
 167         ParseCallbackState *pcbstate = (ParseCallbackState *) arg;
 168
 169         if (geterrcode() != ERRCODE_QUERY_CANCELED)
 170                 (void) parser_errposition(pcbstate->pstate, pcbstate->location);
 171 }
 172
 173
 174 /*
 175  * make_var
 176  *              Build a Var node for an attribute identified by RTE and attrno
 177  */
 178 Var *
 179 make_var(ParseState *pstate, RangeTblEntry *rte, int attrno, int location)
 180 {
 181         Var                *result;
 182         int                     vnum,
 183                                 sublevels_up;
 184         Oid                     vartypeid;
 185         int32           type_mod;
 186
 187         vnum = RTERangeTablePosn(pstate, rte, &sublevels_up);
 188         get_rte_attribute_type(rte, attrno, &vartypeid, &type_mod);
 189         result = makeVar(vnum, attrno, vartypeid, type_mod, sublevels_up);
 190         result->location = location;
 191         return result;
 192 }
 193
 194 /*
 195  * transformArrayType()
 196  *              Get the element type of an array type in preparation for subscripting
 197  */
 198 Oid
 199 transformArrayType(Oid arrayType)
 200 {
 201         Oid                     elementType;
 202         HeapTuple       type_tuple_array;
 203         Form_pg_type type_struct_array;
 204
 205         /* Get the type tuple for the array */
 206         type_tuple_array = SearchSysCache(TYPEOID,
 207                                                                           ObjectIdGetDatum(arrayType),
 208                                                                           0, 0, 0);
 209         if (!HeapTupleIsValid(type_tuple_array))
 210                 elog(ERROR, "cache lookup failed for type %u", arrayType);
 211         type_struct_array = (Form_pg_type) GETSTRUCT(type_tuple_array);
 212
 213         /* needn't check typisdefined since this will fail anyway */
 214
 215         elementType = type_struct_array->typelem;
 216         if (elementType == InvalidOid)
 217                 ereport(ERROR,
 218                                 (errcode(ERRCODE_DATATYPE_MISMATCH),
 219                                  errmsg("cannot subscript type %s because it is not an array",
 220                                                 format_type_be(arrayType))));
 221
 222         ReleaseSysCache(type_tuple_array);
 223
 224         return elementType;
 225 }
 226
 227 /*
 228  * transformArraySubscripts()
 229  *              Transform array subscripting.  This is used for both
 230  *              array fetch and array assignment.
 231  *
 232  * In an array fetch, we are given a source array value and we produce an
 233  * expression that represents the result of extracting a single array element
 234  * or an array slice.
 235  *
 236  * In an array assignment, we are given a destination array value plus a
 237  * source value that is to be assigned to a single element or a slice of
 238  * that array.  We produce an expression that represents the new array value
 239  * with the source data inserted into the right part of the array.
 240  *
 241  * pstate               Parse state
 242  * arrayBase    Already-transformed expression for the array as a whole
 243  * arrayType    OID of array's datatype (should match type of arrayBase)
 244  * elementType  OID of array's element type (fetch with transformArrayType,
 245  *                              or pass InvalidOid to do it here)
 246  * elementTypMod typmod to be applied to array elements (if storing) or of
 247  *                              the source array (if fetching)
 248  * indirection  Untransformed list of subscripts (must not be NIL)
 249  * assignFrom   NULL for array fetch, else transformed expression for source.
 250  */
 251 ArrayRef *
 252 transformArraySubscripts(ParseState *pstate,
 253                                                  Node *arrayBase,
 254                                                  Oid arrayType,
 255                                                  Oid elementType,
 256                                                  int32 elementTypMod,
 257                                                  List *indirection,
 258                                                  Node *assignFrom)
 259 {
 260         bool            isSlice = false;
 261         List       *upperIndexpr = NIL;
 262         List       *lowerIndexpr = NIL;
 263         ListCell   *idx;
 264         ArrayRef   *aref;
 265
 266         /* Caller may or may not have bothered to determine elementType */
 267         if (!OidIsValid(elementType))
 268                 elementType = transformArrayType(arrayType);
 269
 270         /*
 271          * A list containing only single subscripts refers to a single array
 272          * element.  If any of the items are double subscripts (lower:upper), then
 273          * the subscript expression means an array slice operation. In this case,
 274          * we supply a default lower bound of 1 for any items that contain only a
 275          * single subscript.  We have to prescan the indirection list to see if
 276          * there are any double subscripts.
 277          */
 278         foreach(idx, indirection)
 279         {
 280                 A_Indices  *ai = (A_Indices *) lfirst(idx);
 281
 282                 if (ai->lidx != NULL)
 283                 {
 284                         isSlice = true;
 285                         break;
 286                 }
 287         }
 288
 289         /*
 290          * Transform the subscript expressions.
 291          */
 292         foreach(idx, indirection)
 293         {
 294                 A_Indices  *ai = (A_Indices *) lfirst(idx);
 295                 Node       *subexpr;
 296
 297                 Assert(IsA(ai, A_Indices));
 298                 if (isSlice)
 299                 {
 300                         if (ai->lidx)
 301                         {
 302                                 subexpr = transformExpr(pstate, ai->lidx);
 303                                 /* If it's not int4 already, try to coerce */
 304                                 subexpr = coerce_to_target_type(pstate,
 305                                                                                                 subexpr, exprType(subexpr),
 306                                                                                                 INT4OID, -1,
 307                                                                                                 COERCION_ASSIGNMENT,
 308                                                                                                 COERCE_IMPLICIT_CAST,
 309                                                                                                 -1);
 310                                 if (subexpr == NULL)
 311                                         ereport(ERROR,
 312                                                         (errcode(ERRCODE_DATATYPE_MISMATCH),
 313                                                          errmsg("array subscript must have type integer"),
 314                                                 parser_errposition(pstate, exprLocation(ai->lidx))));
 315                         }
 316                         else
 317                         {
 318                                 /* Make a constant 1 */
 319                                 subexpr = (Node *) makeConst(INT4OID,
 320                                                                                          -1,
 321                                                                                          sizeof(int32),
 322                                                                                          Int32GetDatum(1),
 323                                                                                          false,
 324                                                                                          true);         /* pass by value */
 325                         }
 326                         lowerIndexpr = lappend(lowerIndexpr, subexpr);
 327                 }
 328                 subexpr = transformExpr(pstate, ai->uidx);
 329                 /* If it's not int4 already, try to coerce */
 330                 subexpr = coerce_to_target_type(pstate,
 331                                                                                 subexpr, exprType(subexpr),
 332                                                                                 INT4OID, -1,
 333                                                                                 COERCION_ASSIGNMENT,
 334                                                                                 COERCE_IMPLICIT_CAST,
 335                                                                                 -1);
 336                 if (subexpr == NULL)
 337                         ereport(ERROR,
 338                                         (errcode(ERRCODE_DATATYPE_MISMATCH),
 339                                          errmsg("array subscript must have type integer"),
 340                                          parser_errposition(pstate, exprLocation(ai->uidx))));
 341                 upperIndexpr = lappend(upperIndexpr, subexpr);
 342         }
 343
 344         /*
 345          * If doing an array store, coerce the source value to the right type.
 346          * (This should agree with the coercion done by transformAssignedExpr.)
 347          */
 348         if (assignFrom != NULL)
 349         {
 350                 Oid                     typesource = exprType(assignFrom);
 351                 Oid                     typeneeded = isSlice ? arrayType : elementType;
 352                 Node       *newFrom;
 353
 354                 newFrom = coerce_to_target_type(pstate,
 355                                                                                 assignFrom, typesource,
 356                                                                                 typeneeded, elementTypMod,
 357                                                                                 COERCION_ASSIGNMENT,
 358                                                                                 COERCE_IMPLICIT_CAST,
 359                                                                                 -1);
 360                 if (newFrom == NULL)
 361                         ereport(ERROR,
 362                                         (errcode(ERRCODE_DATATYPE_MISMATCH),
 363                                          errmsg("array assignment requires type %s"
 364                                                         " but expression is of type %s",
 365                                                         format_type_be(typeneeded),
 366                                                         format_type_be(typesource)),
 367                                  errhint("You will need to rewrite or cast the expression."),
 368                                          parser_errposition(pstate, exprLocation(assignFrom))));
 369                 assignFrom = newFrom;
 370         }
 371
 372         /*
 373          * Ready to build the ArrayRef node.
 374          */
 375         aref = makeNode(ArrayRef);
 376         aref->refarraytype = arrayType;
 377         aref->refelemtype = elementType;
 378         aref->reftypmod = elementTypMod;
 379         aref->refupperindexpr = upperIndexpr;
 380         aref->reflowerindexpr = lowerIndexpr;
 381         aref->refexpr = (Expr *) arrayBase;
 382         aref->refassgnexpr = (Expr *) assignFrom;
 383
 384         return aref;
 385 }
 386
 387 /*
 388  * make_const
 389  *
 390  *      Convert a Value node (as returned by the grammar) to a Const node
 391  *      of the "natural" type for the constant.  Note that this routine is
 392  *      only used when there is no explicit cast for the constant, so we
 393  *      have to guess what type is wanted.
 394  *
 395  *      For string literals we produce a constant of type UNKNOWN ---- whose
 396  *      representation is the same as cstring, but it indicates to later type
 397  *      resolution that we're not sure yet what type it should be considered.
 398  *      Explicit "NULL" constants are also typed as UNKNOWN.
 399  *
 400  *      For integers and floats we produce int4, int8, or numeric depending
 401  *      on the value of the number.  XXX We should produce int2 as well,
 402  *      but additional cleanup is needed before we can do that; there are
 403  *      too many examples that fail if we try.
 404  */
 405 Const *
 406 make_const(ParseState *pstate, Value *value, int location)
 407 {
 408         Const      *con;
 409         Datum           val;
 410         int64           val64;
 411         Oid                     typeid;
 412         int                     typelen;
 413         bool            typebyval;
 414         ParseCallbackState pcbstate;
 415
 416         switch (nodeTag(value))
 417         {
 418                 case T_Integer:
 419                         val = Int32GetDatum(intVal(value));
 420
 421                         typeid = INT4OID;
 422                         typelen = sizeof(int32);
 423                         typebyval = true;
 424                         break;
 425
 426                 case T_Float:
 427                         /* could be an oversize integer as well as a float ... */
 428                         if (scanint8(strVal(value), true, &val64))
 429                         {
 430                                 /*
 431                                  * It might actually fit in int32. Probably only INT_MIN can
 432                                  * occur, but we'll code the test generally just to be sure.
 433                                  */
 434                                 int32           val32 = (int32) val64;
 435
 436                                 if (val64 == (int64) val32)
 437                                 {
 438                                         val = Int32GetDatum(val32);
 439
 440                                         typeid = INT4OID;
 441                                         typelen = sizeof(int32);
 442                                         typebyval = true;
 443                                 }
 444                                 else
 445                                 {
 446                                         val = Int64GetDatum(val64);
 447
 448                                         typeid = INT8OID;
 449                                         typelen = sizeof(int64);
 450                                         typebyval = FLOAT8PASSBYVAL;            /* int8 and float8 alike */
 451                                 }
 452                         }
 453                         else
 454                         {
 455                                 /* arrange to report location if numeric_in() fails */
 456                                 setup_parser_errposition_callback(&pcbstate, pstate, location);
 457                                 val = DirectFunctionCall3(numeric_in,
 458                                                                                   CStringGetDatum(strVal(value)),
 459                                                                                   ObjectIdGetDatum(InvalidOid),
 460                                                                                   Int32GetDatum(-1));
 461                                 cancel_parser_errposition_callback(&pcbstate);
 462
 463                                 typeid = NUMERICOID;
 464                                 typelen = -1;   /* variable len */
 465                                 typebyval = false;
 466                         }
 467                         break;
 468
 469                 case T_String:
 470
 471                         /*
 472                          * We assume here that UNKNOWN's internal representation is the
 473                          * same as CSTRING
 474                          */
 475                         val = CStringGetDatum(strVal(value));
 476
 477                         typeid = UNKNOWNOID;    /* will be coerced later */
 478                         typelen = -2;           /* cstring-style varwidth type */
 479                         typebyval = false;
 480                         break;
 481
 482                 case T_BitString:
 483                         /* arrange to report location if bit_in() fails */
 484                         setup_parser_errposition_callback(&pcbstate, pstate, location);
 485                         val = DirectFunctionCall3(bit_in,
 486                                                                           CStringGetDatum(strVal(value)),
 487                                                                           ObjectIdGetDatum(InvalidOid),
 488                                                                           Int32GetDatum(-1));
 489                         cancel_parser_errposition_callback(&pcbstate);
 490                         typeid = BITOID;
 491                         typelen = -1;
 492                         typebyval = false;
 493                         break;
 494
 495                 case T_Null:
 496                         /* return a null const */
 497                         con = makeConst(UNKNOWNOID,
 498                                                         -1,
 499                                                         -2,
 500                                                         (Datum) 0,
 501                                                         true,
 502                                                         false);
 503                         con->location = location;
 504                         return con;
 505
 506                 default:
 507                         elog(ERROR, "unrecognized node type: %d", (int) nodeTag(value));
 508                         return NULL;            /* keep compiler quiet */
 509         }
 510
 511         con = makeConst(typeid,
 512                                         -1,                     /* typmod -1 is OK for all cases */
 513                                         typelen,
 514                                         val,
 515                                         false,
 516                                         typebyval);
 517         con->location = location;
 518
 519         return con;
 520 }