Select Git revision
-
Tom Lane authored
varlena elements work now. Allow assignment to previously-nonexistent subscript position to extend array, but only for 1-D arrays and only if adjacent to existing positions (could do more if we had a way to represent nulls in arrays, but I don't want to tackle that now). Arrange for assignment of NULL to an array element in UPDATE to be a no-op, rather than setting the entire array to NULL as it used to. (Throwing an error would be a reasonable alternative, but it's never done that...) Update regress test accordingly.
Tom Lane authoredvarlena elements work now. Allow assignment to previously-nonexistent subscript position to extend array, but only for 1-D arrays and only if adjacent to existing positions (could do more if we had a way to represent nulls in arrays, but I don't want to tackle that now). Arrange for assignment of NULL to an array element in UPDATE to be a no-op, rather than setting the entire array to NULL as it used to. (Throwing an error would be a reasonable alternative, but it's never done that...) Update regress test accordingly.
execQual.c 47.36 KiB
/*-------------------------------------------------------------------------
*
* execQual.c
* Routines to evaluate qualification and targetlist expressions
*
* Portions Copyright (c) 1996-2000, PostgreSQL, Inc
* Portions Copyright (c) 1994, Regents of the University of California
*
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/executor/execQual.c,v 1.76 2000/07/23 01:35:58 tgl Exp $
*
*-------------------------------------------------------------------------
*/
/*
* INTERFACE ROUTINES
* ExecEvalExpr - evaluate an expression and return a datum
* ExecEvalExprSwitchContext - same, but switch into eval memory context
* ExecQual - return true/false if qualification is satisfied
* ExecProject - form a new tuple by projecting the given tuple
*
* NOTES
* ExecEvalExpr() and ExecEvalVar() are hotspots. making these faster
* will speed up the entire system. Unfortunately they are currently
* implemented recursively. Eliminating the recursion is bound to
* improve the speed of the executor.
*
* ExecProject() is used to make tuple projections. Rather then
* trying to speed it up, the execution plan should be pre-processed
* to facilitate attribute sharing between nodes wherever possible,
* instead of doing needless copying. -cim 5/31/91
*
*/
#include "postgres.h"
#include "access/heapam.h"
#include "catalog/pg_language.h"
#include "executor/execFlatten.h"
#include "executor/execdebug.h"
#include "executor/functions.h"
#include "executor/nodeSubplan.h"
#include "utils/array.h"
#include "utils/builtins.h"
#include "utils/fmgroids.h"
#include "utils/fcache2.h"
/* static function decls */
static Datum ExecEvalAggref(Aggref *aggref, ExprContext *econtext, bool *isNull);
static Datum ExecEvalArrayRef(ArrayRef *arrayRef, ExprContext *econtext,
bool *isNull, bool *isDone);
static Datum ExecEvalOper(Expr *opClause, ExprContext *econtext,
bool *isNull);
static Datum ExecEvalFunc(Expr *funcClause, ExprContext *econtext,
bool *isNull, bool *isDone);
static void ExecEvalFuncArgs(FunctionCachePtr fcache, ExprContext *econtext,
List *argList, FunctionCallInfo fcinfo,
bool *argIsDone);
static Datum ExecEvalNot(Expr *notclause, ExprContext *econtext, bool *isNull);
static Datum ExecEvalAnd(Expr *andExpr, ExprContext *econtext, bool *isNull);
static Datum ExecEvalOr(Expr *orExpr, ExprContext *econtext, bool *isNull);
static Datum ExecEvalVar(Var *variable, ExprContext *econtext, bool *isNull);
static Datum ExecMakeFunctionResult(Node *node, List *arguments,
ExprContext *econtext, bool *isNull, bool *isDone);
/*----------
* ExecEvalArrayRef
*
* This function takes an ArrayRef and returns the extracted Datum * if it's a simple reference, or the modified array value if it's
* an array assignment (i.e., array element or slice insertion).
*
* NOTE: if we get a NULL result from a subexpression, we return NULL when
* it's an array reference, or the unmodified source array when it's an
* array assignment. This may seem peculiar, but if we return NULL (as was
* done in versions up through 7.0) then an assignment like
* UPDATE table SET arrayfield[4] = NULL
* will result in setting the whole array to NULL, which is certainly not
* very desirable. By returning the source array we make the assignment
* into a no-op, instead. (Eventually we need to redesign arrays so that
* individual elements can be NULL, but for now, let's try to protect users
* from shooting themselves in the foot.)
*
* NOTE: we deliberately refrain from applying DatumGetArrayTypeP() here,
* even though that might seem natural, because this code needs to support
* both varlena arrays and fixed-length array types. DatumGetArrayTypeP()
* only works for the varlena kind. The routines we call in arrayfuncs.c
* have to know the difference (that's what they need refattrlength for).
*----------
*/
static Datum
ExecEvalArrayRef(ArrayRef *arrayRef,
ExprContext *econtext,
bool *isNull,
bool *isDone)
{
ArrayType *array_source;
ArrayType *resultArray;
bool isAssignment = (arrayRef->refassgnexpr != NULL);
List *elt;
int i = 0,
j = 0;
IntArray upper,
lower;
int *lIndex;
bool dummy;
*isNull = false;
if (arrayRef->refexpr != NULL)
{
array_source = (ArrayType *)
DatumGetPointer(ExecEvalExpr(arrayRef->refexpr,
econtext,
isNull,
isDone));
/*
* If refexpr yields NULL, result is always NULL, for now anyway.
* (This means you cannot assign to an element or slice of an array
* that's NULL; it'll just stay NULL.)
*/
if (*isNull)
return (Datum) NULL;
}
else
{
/*
* Empty refexpr indicates we are doing an INSERT into an array
* column. For now, we just take the refassgnexpr (which the
* parser will have ensured is an array value) and return it
* as-is, ignoring any subscripts that may have been supplied in
* the INSERT column list. This is a kluge, but it's not real
* clear what the semantics ought to be...
*/
array_source = NULL;
}
foreach(elt, arrayRef->refupperindexpr) {
if (i >= MAXDIM)
elog(ERROR, "ExecEvalArrayRef: can only handle %d dimensions",
MAXDIM);
upper.indx[i++] = DatumGetInt32(ExecEvalExpr((Node *) lfirst(elt),
econtext,
isNull,
&dummy));
/* If any index expr yields NULL, result is NULL or source array */
if (*isNull)
{
if (! isAssignment || array_source == NULL)
return (Datum) NULL;
*isNull = false;
return PointerGetDatum(array_source);
}
}
if (arrayRef->reflowerindexpr != NIL)
{
foreach(elt, arrayRef->reflowerindexpr)
{
if (j >= MAXDIM)
elog(ERROR, "ExecEvalArrayRef: can only handle %d dimensions",
MAXDIM);
lower.indx[j++] = DatumGetInt32(ExecEvalExpr((Node *) lfirst(elt),
econtext,
isNull,
&dummy));
/* If any index expr yields NULL, result is NULL or source array */
if (*isNull)
{
if (! isAssignment || array_source == NULL)
return (Datum) NULL;
*isNull = false;
return PointerGetDatum(array_source);
}
}
if (i != j)
elog(ERROR,
"ExecEvalArrayRef: upper and lower indices mismatch");
lIndex = lower.indx;
}
else
lIndex = NULL;
if (isAssignment)
{
Datum sourceData = ExecEvalExpr(arrayRef->refassgnexpr,
econtext,
isNull,
&dummy);
/*
* For now, can't cope with inserting NULL into an array,
* so make it a no-op per discussion above...
*/
if (*isNull)
{
if (array_source == NULL)
return (Datum) NULL;
*isNull = false;
return PointerGetDatum(array_source);
}
if (array_source == NULL)
return sourceData; /* XXX do something else? */
if (lIndex == NULL) resultArray = array_set(array_source, i,
upper.indx,
sourceData,
arrayRef->refelembyval,
arrayRef->refelemlength,
arrayRef->refattrlength,
isNull);
else
resultArray = array_set_slice(array_source, i,
upper.indx, lower.indx,
(ArrayType *) DatumGetPointer(sourceData),
arrayRef->refelembyval,
arrayRef->refelemlength,
arrayRef->refattrlength,
isNull);
return PointerGetDatum(resultArray);
}
if (lIndex == NULL)
return array_ref(array_source, i,
upper.indx,
arrayRef->refelembyval,
arrayRef->refelemlength,
arrayRef->refattrlength,
isNull);
else
{
resultArray = array_get_slice(array_source, i,
upper.indx, lower.indx,
arrayRef->refelembyval,
arrayRef->refelemlength,
arrayRef->refattrlength,
isNull);
return PointerGetDatum(resultArray);
}
}
/* ----------------------------------------------------------------
* ExecEvalAggref
*
* Returns a Datum whose value is the value of the precomputed
* aggregate found in the given expression context.
* ----------------------------------------------------------------
*/
static Datum
ExecEvalAggref(Aggref *aggref, ExprContext *econtext, bool *isNull)
{
if (econtext->ecxt_aggvalues == NULL) /* safety check */
elog(ERROR, "ExecEvalAggref: no aggregates in this expression context");
*isNull = econtext->ecxt_aggnulls[aggref->aggno];
return econtext->ecxt_aggvalues[aggref->aggno];
}
/* ----------------------------------------------------------------
* ExecEvalVar
*
* Returns a Datum whose value is the value of a range
* variable with respect to given expression context.
*
*
* As an entry condition, we expect that the datatype the
* plan expects to get (as told by our "variable" argument) is in
* fact the datatype of the attribute the plan says to fetch (as
* seen in the current context, identified by our "econtext"
* argument).
*
* If we fetch a Type A attribute and Caller treats it as if it
* were Type B, there will be undefined results (e.g. crash). * One way these might mismatch now is that we're accessing a
* catalog class and the type information in the pg_attribute
* class does not match the hardcoded pg_attribute information
* (in pg_attribute.h) for the class in question.
*
* We have an Assert to make sure this entry condition is met.
*
* ---------------------------------------------------------------- */
static Datum
ExecEvalVar(Var *variable, ExprContext *econtext, bool *isNull)
{
Datum result;
TupleTableSlot *slot;
AttrNumber attnum;
HeapTuple heapTuple;
TupleDesc tuple_type;
bool byval;
int16 len;
/*
* get the slot we want
*/
switch (variable->varno)
{
case INNER: /* get the tuple from the inner node */
slot = econtext->ecxt_innertuple;
break;
case OUTER: /* get the tuple from the outer node */
slot = econtext->ecxt_outertuple;
break;
default: /* get the tuple from the relation being
* scanned */
slot = econtext->ecxt_scantuple;
break;
}
/*
* extract tuple information from the slot
*/
heapTuple = slot->val;
tuple_type = slot->ttc_tupleDescriptor;
attnum = variable->varattno;
/* (See prolog for explanation of this Assert) */
Assert(attnum <= 0 ||
(attnum - 1 <= tuple_type->natts - 1 &&
tuple_type->attrs[attnum - 1] != NULL &&
variable->vartype == tuple_type->attrs[attnum - 1]->atttypid));
/*
* If the attribute number is invalid, then we are supposed to return
* the entire tuple, we give back a whole slot so that callers know
* what the tuple looks like.
*/
if (attnum == InvalidAttrNumber)
{
TupleTableSlot *tempSlot;
TupleDesc td;
HeapTuple tup;
tempSlot = makeNode(TupleTableSlot);
tempSlot->ttc_shouldFree = false;
tempSlot->ttc_descIsNew = true;
tempSlot->ttc_tupleDescriptor = (TupleDesc) NULL;
tempSlot->ttc_buffer = InvalidBuffer;
tempSlot->ttc_whichplan = -1;
tup = heap_copytuple(heapTuple);
td = CreateTupleDescCopy(tuple_type);
ExecSetSlotDescriptor(tempSlot, td);
ExecStoreTuple(tup, tempSlot, InvalidBuffer, true);
return PointerGetDatum(tempSlot);
}
result = heap_getattr(heapTuple, /* tuple containing attribute */
attnum, /* attribute number of desired
* attribute */
tuple_type, /* tuple descriptor of tuple */
isNull); /* return: is attribute null? */
/*
* return null if att is null
*/
if (*isNull)
return (Datum) 0;
/*
* get length and type information.. ??? what should we do about
* variable length attributes - variable length attributes have their
* length stored in the first 4 bytes of the memory pointed to by the
* returned value.. If we can determine that the type is a variable
* length type, we can do the right thing. -cim 9/15/89
*/
if (attnum < 0)
{
/*
* If this is a pseudo-att, we get the type and fake the length.
* There ought to be a routine to return the real lengths, so
* we'll mark this one ... XXX -mao
*/
len = heap_sysattrlen(attnum); /* XXX see -mao above */
byval = heap_sysattrbyval(attnum); /* XXX see -mao above */
}
else
{
len = tuple_type->attrs[attnum - 1]->attlen;
byval = tuple_type->attrs[attnum - 1]->attbyval ? true : false;
}
return result;
}
/* ----------------------------------------------------------------
* ExecEvalParam
*
* Returns the value of a parameter. A param node contains
* something like ($.name) and the expression context contains
* the current parameter bindings (name = "sam") (age = 34)...
* so our job is to replace the param node with the datum
* containing the appropriate information ("sam").
*
* Q: if we have a parameter ($.foo) without a binding, i.e.
* there is no (foo = xxx) in the parameter list info,
* is this a fatal error or should this be a "not available"
* (in which case we shoud return a Const node with the
* isnull flag) ? -cim 10/13/89
*
* Minor modification: Param nodes now have an extra field,
* `paramkind' which specifies the type of parameter
* (see params.h). So while searching the paramList for
* a paramname/value pair, we have also to check for `kind'.
*
* NOTE: The last entry in `paramList' is always an
* entry with kind == PARAM_INVALID. * ----------------------------------------------------------------
*/
Datum
ExecEvalParam(Param *expression, ExprContext *econtext, bool *isNull)
{
char *thisParameterName;
int thisParameterKind = expression->paramkind;
AttrNumber thisParameterId = expression->paramid;
int matchFound;
ParamListInfo paramList;
if (thisParameterKind == PARAM_EXEC)
{
ParamExecData *prm;
prm = &(econtext->ecxt_param_exec_vals[thisParameterId]);
if (prm->execPlan != NULL)
{
ExecSetParamPlan(prm->execPlan, econtext);
/* ExecSetParamPlan should have processed this param... */
Assert(prm->execPlan == NULL);
}
*isNull = prm->isnull;
return prm->value;
}
thisParameterName = expression->paramname;
paramList = econtext->ecxt_param_list_info;
*isNull = false;
/*
* search the list with the parameter info to find a matching name. An
* entry with an InvalidName denotes the last element in the array.
*/
matchFound = 0;
if (paramList != NULL)
{
/*
* search for an entry in 'paramList' that matches the
* `expression'.
*/
while (paramList->kind != PARAM_INVALID && !matchFound)
{
switch (thisParameterKind)
{
case PARAM_NAMED:
if (thisParameterKind == paramList->kind &&
strcmp(paramList->name, thisParameterName) == 0)
matchFound = 1;
break;
case PARAM_NUM:
if (thisParameterKind == paramList->kind &&
paramList->id == thisParameterId)
matchFound = 1;
break;
case PARAM_OLD:
case PARAM_NEW:
if (thisParameterKind == paramList->kind &&
paramList->id == thisParameterId)
{
matchFound = 1;
/*
* sanity check
*/
if (strcmp(paramList->name, thisParameterName) != 0)
{
elog(ERROR, "ExecEvalParam: new/old params with same id & diff names");
}
}
break;
default:
/*
* oops! this is not supposed to happen!
*/
elog(ERROR, "ExecEvalParam: invalid paramkind %d",
thisParameterKind);
}
if (!matchFound)
paramList++;
} /* while */
} /* if */
if (!matchFound)
{
/*
* ooops! we couldn't find this parameter in the parameter list.
* Signal an error
*/
elog(ERROR, "ExecEvalParam: Unknown value for parameter %s",
thisParameterName);
}
/*
* return the value.
*/
if (paramList->isnull)
{
*isNull = true;
return (Datum) 0;
}
if (expression->param_tlist != NIL)
{
HeapTuple tup;
Datum value;
List *tlist = expression->param_tlist;
TargetEntry *tle = (TargetEntry *) lfirst(tlist);
TupleTableSlot *slot = (TupleTableSlot *) paramList->value;
tup = slot->val;
value = ProjectAttribute(slot->ttc_tupleDescriptor,
tle, tup, isNull);
return value;
}
return paramList->value;
}
/* ----------------------------------------------------------------
* ExecEvalOper / ExecEvalFunc support routines
* ----------------------------------------------------------------
*/
/*
* GetAttributeByName
* GetAttributeByNum
*
* These are functions which return the value of the
* named attribute out of the tuple from the arg slot. User defined
* C functions which take a tuple as an argument are expected
* to use this. Ex: overpaid(EMP) might call GetAttributeByNum().
*
* XXX these two functions are misdeclared: they should be declared to
* return Datum. They are not used anywhere in the backend proper, and * exist only for use by user-defined functions. Should we change their
* definitions, at risk of breaking user code?
*/
char *
GetAttributeByNum(TupleTableSlot *slot,
AttrNumber attrno,
bool *isNull)
{
Datum retval;
if (!AttributeNumberIsValid(attrno))
elog(ERROR, "GetAttributeByNum: Invalid attribute number");
if (!AttrNumberIsForUserDefinedAttr(attrno))
elog(ERROR, "GetAttributeByNum: cannot access system attributes here");
if (isNull == (bool *) NULL)
elog(ERROR, "GetAttributeByNum: a NULL isNull flag was passed");
if (TupIsNull(slot))
{
*isNull = true;
return (char *) NULL;
}
retval = heap_getattr(slot->val,
attrno,
slot->ttc_tupleDescriptor,
isNull);
if (*isNull)
return (char *) NULL;
return (char *) retval;
}
char *
GetAttributeByName(TupleTableSlot *slot, char *attname, bool *isNull)
{
AttrNumber attrno;
TupleDesc tupdesc;
Datum retval;
int natts;
int i;
if (attname == NULL)
elog(ERROR, "GetAttributeByName: Invalid attribute name");
if (isNull == (bool *) NULL)
elog(ERROR, "GetAttributeByName: a NULL isNull flag was passed");
if (TupIsNull(slot))
{
*isNull = true;
return (char *) NULL;
}
tupdesc = slot->ttc_tupleDescriptor;
natts = slot->val->t_data->t_natts;
attrno = InvalidAttrNumber;
for (i = 0; i < tupdesc->natts; i++)
{
if (namestrcmp(&(tupdesc->attrs[i]->attname), attname) == 0)
{
attrno = tupdesc->attrs[i]->attnum;
break;
}
}
if (attrno == InvalidAttrNumber)
elog(ERROR, "GetAttributeByName: attribute %s not found", attname);
retval = heap_getattr(slot->val,
attrno,
tupdesc,
isNull);
if (*isNull)
return (char *) NULL;
return (char *) retval;
}
static void
ExecEvalFuncArgs(FunctionCachePtr fcache,
ExprContext *econtext,
List *argList,
FunctionCallInfo fcinfo,
bool *argIsDone)
{
int i;
List *arg;
i = 0;
foreach(arg, argList)
{
/*
* evaluate the expression, in general functions cannot take sets
* as arguments but we make an exception in the case of nested dot
* expressions. We have to watch out for this case here.
*/
fcinfo->arg[i] = ExecEvalExpr((Node *) lfirst(arg),
econtext,
&fcinfo->argnull[i],
argIsDone);
if (!(*argIsDone))
{
if (i != 0)
elog(ERROR, "functions can only take sets in their first argument");
fcache->setArg = fcinfo->arg[0];
fcache->hasSetArg = true;
}
i++;
}
}
/*
* ExecMakeFunctionResult
*/
static Datum
ExecMakeFunctionResult(Node *node,
List *arguments,
ExprContext *econtext,
bool *isNull,
bool *isDone)
{
FunctionCallInfoData fcinfo;
FunctionCachePtr fcache;
List *ftlist;
bool funcisset;
Datum result;
bool argDone;
MemSet(&fcinfo, 0, sizeof(fcinfo));
/*
* This is kind of ugly, Func nodes now have targetlists so that we
* know when and what to project out from postquel function results.
* ExecMakeFunctionResult becomes a little bit more of a dual personality
* as a result. */
if (IsA(node, Func))
{
fcache = ((Func *) node)->func_fcache;
ftlist = ((Func *) node)->func_tlist;
funcisset = (((Func *) node)->funcid == F_SETEVAL);
}
else
{
fcache = ((Oper *) node)->op_fcache;
ftlist = NIL;
funcisset = false;
}
fcinfo.flinfo = &fcache->func;
fcinfo.nargs = fcache->nargs;
/*
* arguments is a list of expressions to evaluate before passing to
* the function manager. We collect the results of evaluating the
* expressions into the FunctionCallInfo struct. Note we assume that
* fcache->nargs is the correct length of the arguments list!
*/
if (fcache->nargs > 0)
{
if (fcache->nargs > FUNC_MAX_ARGS)
elog(ERROR, "ExecMakeFunctionResult: too many arguments");
/*
* If the setArg in the fcache is set we have an argument
* returning a set of tuples (i.e. a nested dot expression). We
* don't want to evaluate the arguments again until the function
* is done. hasSetArg will always be false until we eval the args
* for the first time.
*/
if (fcache->hasSetArg && fcache->setArg != (Datum) 0)
{
fcinfo.arg[0] = fcache->setArg;
argDone = false;
}
else
ExecEvalFuncArgs(fcache, econtext, arguments, &fcinfo, &argDone);
if (fcache->hasSetArg && argDone)
{
/* can only get here if input is an empty set. */
*isNull = true;
*isDone = true;
return (Datum) 0;
}
}
/*
* If this function is really a set, we have to diddle with things. If
* the function has already been called at least once, then the setArg
* field of the fcache holds the OID of this set in pg_proc. (This is
* not quite legit, since the setArg field is really for functions
* which take sets of tuples as input - set functions take no inputs
* at all. But it's a nice place to stash this value, for now.)
*
* If this is the first call of the set's function, then the call to
* ExecEvalFuncArgs above just returned the OID of the pg_proc tuple
* which defines this set. So replace the existing funcid in the
* funcnode with the set's OID. Also, we want a new fcache which
* points to the right function, so get that, now that we have the
* right OID. Also zero out fcinfo.arg, since the real set doesn't take
* any arguments.
*/
if (funcisset)
{ if (fcache->setArg)
{
((Func *) node)->funcid = DatumGetObjectId(fcache->setArg);
}
else
{
((Func *) node)->funcid = DatumGetObjectId(fcinfo.arg[0]);
setFcache(node, DatumGetObjectId(fcinfo.arg[0]), NIL, econtext);
fcache = ((Func *) node)->func_fcache;
fcache->setArg = fcinfo.arg[0];
}
fcinfo.arg[0] = (Datum) 0;
}
/*
* now return the value gotten by calling the function manager,
* passing the function the evaluated parameter values.
*/
if (fcache->language == SQLlanguageId)
{
/*--------------------
* This loop handles the situation where we are iterating through
* all results in a nested dot function (whose argument function
* returns a set of tuples) and the current function finally
* finishes. We need to get the next argument in the set and start
* the function all over again. We might have to do it more than
* once, if the function produces no results for a particular argument.
* This is getting unclean.
*--------------------
*/
for (;;)
{
/*
* If function is strict, and there are any NULL arguments,
* skip calling the function (at least for this set of args).
*/
bool callit = true;
if (fcinfo.flinfo->fn_strict)
{
int i;
for (i = 0; i < fcinfo.nargs; i++)
{
if (fcinfo.argnull[i])
{
callit = false;
break;
}
}
}
if (callit)
{
result = postquel_function(&fcinfo, fcache, ftlist, isDone);
*isNull = fcinfo.isnull;
}
else
{
result = (Datum) 0;
*isNull = true;
*isDone = true;
}
if (!*isDone)
break; /* got a result from current argument */
if (!fcache->hasSetArg)
break; /* input not a set, so done */
/* OK, get the next argument... */ ExecEvalFuncArgs(fcache, econtext, arguments, &fcinfo, &argDone);
if (argDone)
{
/*
* End of arguments, so reset the setArg flag and say
* "Done"
*/
fcache->setArg = (Datum) 0;
fcache->hasSetArg = false;
*isDone = true;
*isNull = true;
result = (Datum) 0;
break;
}
/*
* If we reach here, loop around to run the function on the
* new argument.
*/
}
if (funcisset)
{
/*
* reset the funcid so that next call to this routine will
* still recognize this func as a set. Note that for now we
* assume that the set function in pg_proc must be a Postquel
* function - the funcid is not reset below for C functions.
*/
((Func *) node)->funcid = F_SETEVAL;
/*
* If we're done with the results of this function, get rid of
* its func cache.
*/
if (*isDone)
((Func *) node)->func_fcache = NULL;
}
}
else
{
/* A non-SQL function cannot return a set, at present. */
*isDone = true;
/*
* If function is strict, and there are any NULL arguments,
* skip calling the function and return NULL.
*/
if (fcinfo.flinfo->fn_strict)
{
int i;
for (i = 0; i < fcinfo.nargs; i++)
{
if (fcinfo.argnull[i])
{
*isNull = true;
return (Datum) 0;
}
}
}
result = FunctionCallInvoke(&fcinfo);
*isNull = fcinfo.isnull;
}
return result;
}
/* ----------------------------------------------------------------
* ExecEvalOper
* ExecEvalFunc
*
* Evaluate the functional result of a list of arguments by calling the
* function manager.
* ----------------------------------------------------------------
*/
/* ----------------------------------------------------------------
* ExecEvalOper
* ----------------------------------------------------------------
*/
static Datum
ExecEvalOper(Expr *opClause, ExprContext *econtext, bool *isNull)
{
Oper *op;
List *argList;
FunctionCachePtr fcache;
bool isDone;
/*
* we extract the oid of the function associated with the op and then
* pass the work onto ExecMakeFunctionResult which evaluates the
* arguments and returns the result of calling the function on the
* evaluated arguments.
*/
op = (Oper *) opClause->oper;
argList = opClause->args;
/*
* get the fcache from the Oper node. If it is NULL, then initialize
* it
*/
fcache = op->op_fcache;
if (fcache == NULL)
{
setFcache((Node *) op, op->opid, argList, econtext);
fcache = op->op_fcache;
}
/*
* call ExecMakeFunctionResult() with a dummy isDone that we ignore.
* We don't have operator whose arguments are sets.
*/
return ExecMakeFunctionResult((Node *) op, argList, econtext,
isNull, &isDone);
}
/* ----------------------------------------------------------------
* ExecEvalFunc
* ----------------------------------------------------------------
*/
static Datum
ExecEvalFunc(Expr *funcClause,
ExprContext *econtext,
bool *isNull,
bool *isDone)
{
Func *func;
List *argList;
FunctionCachePtr fcache;
/*
* we extract the oid of the function associated with the func node and
* then pass the work onto ExecMakeFunctionResult which evaluates the
* arguments and returns the result of calling the function on the * evaluated arguments.
*
* this is nearly identical to the ExecEvalOper code.
*/
func = (Func *) funcClause->oper;
argList = funcClause->args;
/*
* get the fcache from the Func node. If it is NULL, then initialize
* it
*/
fcache = func->func_fcache;
if (fcache == NULL)
{
setFcache((Node *) func, func->funcid, argList, econtext);
fcache = func->func_fcache;
}
return ExecMakeFunctionResult((Node *) func, argList, econtext,
isNull, isDone);
}
/* ----------------------------------------------------------------
* ExecEvalNot
* ExecEvalOr
* ExecEvalAnd
*
* Evaluate boolean expressions. Evaluation of 'or' is
* short-circuited when the first true (or null) value is found.
*
* The query planner reformulates clause expressions in the
* qualification to conjunctive normal form. If we ever get
* an AND to evaluate, we can be sure that it's not a top-level
* clause in the qualification, but appears lower (as a function
* argument, for example), or in the target list. Not that you
* need to know this, mind you...
* ----------------------------------------------------------------
*/
static Datum
ExecEvalNot(Expr *notclause, ExprContext *econtext, bool *isNull)
{
Node *clause;
Datum expr_value;
bool isDone;
clause = lfirst(notclause->args);
/*
* We don't iterate over sets in the quals, so pass in an isDone flag,
* but ignore it.
*/
expr_value = ExecEvalExpr(clause, econtext, isNull, &isDone);
/*
* if the expression evaluates to null, then we just cascade the null
* back to whoever called us.
*/
if (*isNull)
return expr_value;
/*
* evaluation of 'not' is simple.. expr is false, then return 'true'
* and vice versa.
*/
return BoolGetDatum(! DatumGetBool(expr_value));
}
/* ----------------------------------------------------------------
* ExecEvalOr
* ---------------------------------------------------------------- */
static Datum
ExecEvalOr(Expr *orExpr, ExprContext *econtext, bool *isNull)
{
List *clauses;
List *clause;
bool isDone;
bool AnyNull;
Datum clause_value;
clauses = orExpr->args;
AnyNull = false;
/*
* If any of the clauses is TRUE, the OR result is TRUE regardless of
* the states of the rest of the clauses, so we can stop evaluating
* and return TRUE immediately. If none are TRUE and one or more is
* NULL, we return NULL; otherwise we return FALSE. This makes sense
* when you interpret NULL as "don't know": if we have a TRUE then the
* OR is TRUE even if we aren't sure about some of the other inputs.
* If all the known inputs are FALSE, but we have one or more "don't
* knows", then we have to report that we "don't know" what the OR's
* result should be --- perhaps one of the "don't knows" would have
* been TRUE if we'd known its value. Only when all the inputs are
* known to be FALSE can we state confidently that the OR's result is
* FALSE.
*/
foreach(clause, clauses)
{
/*
* We don't iterate over sets in the quals, so pass in an isDone
* flag, but ignore it.
*/
clause_value = ExecEvalExpr((Node *) lfirst(clause),
econtext,
isNull,
&isDone);
/*
* if we have a non-null true result, then return it.
*/
if (*isNull)
AnyNull = true; /* remember we got a null */
else if (DatumGetBool(clause_value))
return clause_value;
}
/* AnyNull is true if at least one clause evaluated to NULL */
*isNull = AnyNull;
return BoolGetDatum(false);
}
/* ----------------------------------------------------------------
* ExecEvalAnd
* ----------------------------------------------------------------
*/
static Datum
ExecEvalAnd(Expr *andExpr, ExprContext *econtext, bool *isNull)
{
List *clauses;
List *clause;
bool isDone;
bool AnyNull;
Datum clause_value;
clauses = andExpr->args;
AnyNull = false;
/* * If any of the clauses is FALSE, the AND result is FALSE regardless
* of the states of the rest of the clauses, so we can stop evaluating
* and return FALSE immediately. If none are FALSE and one or more is
* NULL, we return NULL; otherwise we return TRUE. This makes sense
* when you interpret NULL as "don't know", using the same sort of
* reasoning as for OR, above.
*/
foreach(clause, clauses)
{
/*
* We don't iterate over sets in the quals, so pass in an isDone
* flag, but ignore it.
*/
clause_value = ExecEvalExpr((Node *) lfirst(clause),
econtext,
isNull,
&isDone);
/*
* if we have a non-null false result, then return it.
*/
if (*isNull)
AnyNull = true; /* remember we got a null */
else if (! DatumGetBool(clause_value))
return clause_value;
}
/* AnyNull is true if at least one clause evaluated to NULL */
*isNull = AnyNull;
return BoolGetDatum(!AnyNull);
}
/* ----------------------------------------------------------------
* ExecEvalCase
*
* Evaluate a CASE clause. Will have boolean expressions
* inside the WHEN clauses, and will have expressions
* for results.
* - thomas 1998-11-09
* ----------------------------------------------------------------
*/
static Datum
ExecEvalCase(CaseExpr *caseExpr, ExprContext *econtext, bool *isNull)
{
List *clauses;
List *clause;
Datum clause_value;
bool isDone;
clauses = caseExpr->args;
/*
* we evaluate each of the WHEN clauses in turn, as soon as one is
* true we return the corresponding result. If none are true then we
* return the value of the default clause, or NULL if there is none.
*/
foreach(clause, clauses)
{
CaseWhen *wclause = lfirst(clause);
/*
* We don't iterate over sets in the quals, so pass in an isDone
* flag, but ignore it.
*/
clause_value = ExecEvalExpr(wclause->expr,
econtext,
isNull,
&isDone);
/*
* if we have a true test, then we return the result, since the
* case statement is satisfied. A NULL result from the test is
* not considered true.
*/
if (DatumGetBool(clause_value) && !*isNull)
{
return ExecEvalExpr(wclause->result,
econtext,
isNull,
&isDone);
}
}
if (caseExpr->defresult)
{
return ExecEvalExpr(caseExpr->defresult,
econtext,
isNull,
&isDone);
}
*isNull = true;
return (Datum) 0;
}
/* ----------------------------------------------------------------
* ExecEvalExpr
*
* Recursively evaluate a targetlist or qualification expression.
*
* The caller should already have switched into the temporary
* memory context econtext->ecxt_per_tuple_memory. The convenience
* entry point ExecEvalExprSwitchContext() is provided for callers
* who don't prefer to do the switch in an outer loop. We do not
* do the switch here because it'd be a waste of cycles during
* recursive entries to ExecEvalExpr().
*
* This routine is an inner loop routine and must be as fast
* as possible.
* ----------------------------------------------------------------
*/
Datum
ExecEvalExpr(Node *expression,
ExprContext *econtext,
bool *isNull,
bool *isDone)
{
Datum retDatum;
/* Set default values for result flags: non-null, not a set result */
*isNull = false;
*isDone = true;
/* Is this still necessary? Doubtful... */
if (expression == NULL)
{
*isNull = true;
return (Datum) 0;
}
/*
* here we dispatch the work to the appropriate type of function given
* the type of our expression.
*/
switch (nodeTag(expression))
{
case T_Var:
retDatum = ExecEvalVar((Var *) expression, econtext, isNull);
break; case T_Const:
{
Const *con = (Const *) expression;
retDatum = con->constvalue;
*isNull = con->constisnull;
break;
}
case T_Param:
retDatum = ExecEvalParam((Param *) expression, econtext, isNull);
break;
case T_Iter:
retDatum = ExecEvalIter((Iter *) expression,
econtext,
isNull,
isDone);
break;
case T_Aggref:
retDatum = ExecEvalAggref((Aggref *) expression, econtext, isNull);
break;
case T_ArrayRef:
retDatum = ExecEvalArrayRef((ArrayRef *) expression,
econtext,
isNull,
isDone);
break;
case T_Expr:
{
Expr *expr = (Expr *) expression;
switch (expr->opType)
{
case OP_EXPR:
retDatum = ExecEvalOper(expr, econtext, isNull);
break;
case FUNC_EXPR:
retDatum = ExecEvalFunc(expr, econtext,
isNull, isDone);
break;
case OR_EXPR:
retDatum = ExecEvalOr(expr, econtext, isNull);
break;
case AND_EXPR:
retDatum = ExecEvalAnd(expr, econtext, isNull);
break;
case NOT_EXPR:
retDatum = ExecEvalNot(expr, econtext, isNull);
break;
case SUBPLAN_EXPR:
retDatum = ExecSubPlan((SubPlan *) expr->oper,
expr->args, econtext,
isNull);
break;
default:
elog(ERROR, "ExecEvalExpr: unknown expression type %d",
expr->opType);
retDatum = 0; /* keep compiler quiet */
break;
}
break;
}
case T_RelabelType:
retDatum = ExecEvalExpr(((RelabelType *) expression)->arg,
econtext,
isNull,
isDone);
break;
case T_CaseExpr:
retDatum = ExecEvalCase((CaseExpr *) expression, econtext, isNull);
break;
default:
elog(ERROR, "ExecEvalExpr: unknown expression type %d",
nodeTag(expression));
retDatum = 0; /* keep compiler quiet */
break;
}
return retDatum;
} /* ExecEvalExpr() */
/*
* Same as above, but get into the right allocation context explicitly.
*/
Datum
ExecEvalExprSwitchContext(Node *expression,
ExprContext *econtext,
bool *isNull,
bool *isDone)
{
Datum retDatum;
MemoryContext oldContext;
oldContext = MemoryContextSwitchTo(econtext->ecxt_per_tuple_memory);
retDatum = ExecEvalExpr(expression, econtext, isNull, isDone);
MemoryContextSwitchTo(oldContext);
return retDatum;
}
/* ----------------------------------------------------------------
* ExecQual / ExecTargetList / ExecProject
* ----------------------------------------------------------------
*/
/* ----------------------------------------------------------------
* ExecQual
*
* Evaluates a conjunctive boolean expression (qual list) and
* returns true iff none of the subexpressions are false.
* (We also return true if the list is empty.)
*
* If some of the subexpressions yield NULL but none yield FALSE,
* then the result of the conjunction is NULL (ie, unknown)
* according to three-valued boolean logic. In this case,
* we return the value specified by the "resultForNull" parameter.
*
* Callers evaluating WHERE clauses should pass resultForNull=FALSE,
* since SQL specifies that tuples with null WHERE results do not
* get selected. On the other hand, callers evaluating constraint
* conditions should pass resultForNull=TRUE, since SQL also specifies
* that NULL constraint conditions are not failures.
*
* NOTE: it would not be correct to use this routine to evaluate an
* AND subclause of a boolean expression; for that purpose, a NULL
* result must be returned as NULL so that it can be properly treated
* in the next higher operator (cf. ExecEvalAnd and ExecEvalOr).
* This routine is only used in contexts where a complete expression
* is being evaluated and we know that NULL can be treated the same
* as one boolean result or the other.
*
* ----------------------------------------------------------------
*/
bool
ExecQual(List *qual, ExprContext *econtext, bool resultForNull)
{
bool result;
MemoryContext oldContext;
List *qlist;
/*
* debugging stuff
*/
EV_printf("ExecQual: qual is ");
EV_nodeDisplay(qual);
EV_printf("\n");
IncrProcessed();
/*
* Run in short-lived per-tuple context while computing expressions.
*/
oldContext = MemoryContextSwitchTo(econtext->ecxt_per_tuple_memory);
/*
* Evaluate the qual conditions one at a time. If we find a FALSE
* result, we can stop evaluating and return FALSE --- the AND result
* must be FALSE. Also, if we find a NULL result when resultForNull
* is FALSE, we can stop and return FALSE --- the AND result must be
* FALSE or NULL in that case, and the caller doesn't care which.
*
* If we get to the end of the list, we can return TRUE. This will
* happen when the AND result is indeed TRUE, or when the AND result
* is NULL (one or more NULL subresult, with all the rest TRUE) and
* the caller has specified resultForNull = TRUE.
*/
result = true;
foreach(qlist, qual)
{
Node *clause = (Node *) lfirst(qlist);
Datum expr_value;
bool isNull;
bool isDone;
/*
* pass isDone, but ignore it. We don't iterate over multiple
* returns in the qualifications.
*/
expr_value = ExecEvalExpr(clause, econtext, &isNull, &isDone);
if (isNull)
{
if (resultForNull == false)
{
result = false; /* treat NULL as FALSE */
break;
}
}
else
{
if (! DatumGetBool(expr_value))
{
result = false; /* definitely FALSE */
break;
}
}
}
MemoryContextSwitchTo(oldContext);
return result;
}
int
ExecTargetListLength(List *targetlist)
{
int len;
List *tl; TargetEntry *curTle;
len = 0;
foreach(tl, targetlist)
{
curTle = lfirst(tl);
if (curTle->resdom != NULL)
len++;
else
len += curTle->fjoin->fj_nNodes;
}
return len;
}
/* ----------------------------------------------------------------
* ExecTargetList
*
* Evaluates a targetlist with respect to the current
* expression context and return a tuple.
* ----------------------------------------------------------------
*/
static HeapTuple
ExecTargetList(List *targetlist,
int nodomains,
TupleDesc targettype,
Datum *values,
ExprContext *econtext,
bool *isDone)
{
MemoryContext oldContext;
char nulls_array[64];
bool fjNullArray[64];
bool itemIsDoneArray[64];
char *null_head;
bool *fjIsNull;
bool *itemIsDone;
List *tl;
TargetEntry *tle;
Node *expr;
Resdom *resdom;
AttrNumber resind;
Datum constvalue;
HeapTuple newTuple;
bool isNull;
bool haveDoneIters;
static struct tupleDesc NullTupleDesc; /* we assume this inits to
* zeroes */
/*
* debugging stuff
*/
EV_printf("ExecTargetList: tl is ");
EV_nodeDisplay(targetlist);
EV_printf("\n");
/*
* Run in short-lived per-tuple context while computing expressions.
*/
oldContext = MemoryContextSwitchTo(econtext->ecxt_per_tuple_memory);
/*
* There used to be some klugy and demonstrably broken code here that
* special-cased the situation where targetlist == NIL. Now we just
* fall through and return an empty-but-valid tuple. We do, however,
* have to cope with the possibility that targettype is NULL ---
* heap_formtuple won't like that, so pass a dummy descriptor with
* natts = 0 to deal with it.
*/
if (targettype == NULL) targettype = &NullTupleDesc;
/*
* allocate an array of char's to hold the "null" information only if
* we have a really large targetlist. otherwise we use the stack.
*
* We also allocate a bool array that is used to hold fjoin result state,
* and another that holds the isDone status for each targetlist item.
*/
if (nodomains > 64)
{
null_head = (char *) palloc(nodomains + 1);
fjIsNull = (bool *) palloc(nodomains + 1);
itemIsDone = (bool *) palloc(nodomains + 1);
}
else
{
null_head = &nulls_array[0];
fjIsNull = &fjNullArray[0];
itemIsDone = &itemIsDoneArray[0];
}
/*
* evaluate all the expressions in the target list
*/
*isDone = true; /* until proven otherwise */
haveDoneIters = false; /* any isDone Iter exprs in tlist? */
foreach(tl, targetlist)
{
/*
* remember, a target list is a list of lists:
*
* ((<resdom | fjoin> expr) (<resdom | fjoin> expr) ...)
*
* tl is a pointer to successive cdr's of the targetlist tle is a
* pointer to the target list entry in tl
*/
tle = lfirst(tl);
if (tle->resdom != NULL)
{
expr = tle->expr;
resdom = tle->resdom;
resind = resdom->resno - 1;
constvalue = ExecEvalExpr(expr,
econtext,
&isNull,
&itemIsDone[resind]);
values[resind] = constvalue;
if (!isNull)
null_head[resind] = ' ';
else
null_head[resind] = 'n';
if (IsA(expr, Iter))
{
if (itemIsDone[resind])
haveDoneIters = true;
else
*isDone = false; /* we have undone Iters in the
* list */
}
}
else {
int curNode;
Resdom *fjRes;
List *fjTlist = (List *) tle->expr;
Fjoin *fjNode = tle->fjoin;
int nNodes = fjNode->fj_nNodes;
DatumPtr results = fjNode->fj_results;
ExecEvalFjoin(tle, econtext, fjIsNull, isDone);
/* this is probably wrong: */
if (*isDone)
{
newTuple = NULL;
goto exit;
}
/*
* get the result from the inner node
*/
fjRes = (Resdom *) fjNode->fj_innerNode;
resind = fjRes->resno - 1;
if (fjIsNull[0])
null_head[resind] = 'n';
else
{
null_head[resind] = ' ';
values[resind] = results[0];
}
/*
* Get results from all of the outer nodes
*/
for (curNode = 1;
curNode < nNodes;
curNode++, fjTlist = lnext(fjTlist))
{
#ifdef NOT_USED /* what is this?? */
Node *outernode = lfirst(fjTlist);
fjRes = (Resdom *) outernode->iterexpr;
#endif
resind = fjRes->resno - 1;
if (fjIsNull[curNode])
null_head[resind] = 'n';
else
{
null_head[resind] = ' ';
values[resind] = results[curNode];
}
}
}
}
if (haveDoneIters)
{
if (*isDone)
{
/*
* all Iters are done, so return a null indicating tlist set
* expansion is complete.
*/
newTuple = NULL;
goto exit;
}
else
{
/* * We have some done and some undone Iters. Restart the done
* ones so that we can deliver a tuple (if possible).
*
* XXX this code is a crock, because it only works for Iters at
* the top level of tlist expressions, and doesn't even work
* right for them: you should get all possible combinations of
* Iter results, but you won't unless the numbers of values
* returned by each are relatively prime. Should have a
* mechanism more like aggregate functions, where we make a
* list of all Iters contained in the tlist and cycle through
* their values in a methodical fashion. To do someday; can't
* get excited about fixing a Berkeley feature that's not in
* SQL92. (The only reason we're doing this much is that we
* have to be sure all the Iters are run to completion, or
* their subplan executors will have unreleased resources,
* e.g. pinned buffers...)
*/
foreach(tl, targetlist)
{
tle = lfirst(tl);
if (tle->resdom != NULL)
{
expr = tle->expr;
resdom = tle->resdom;
resind = resdom->resno - 1;
if (IsA(expr, Iter) &&itemIsDone[resind])
{
constvalue = ExecEvalExpr(expr,
econtext,
&isNull,
&itemIsDone[resind]);
if (itemIsDone[resind])
{
/*
* Oh dear, this Iter is returning an empty
* set. Guess we can't make a tuple after all.
*/
*isDone = true;
newTuple = NULL;
goto exit;
}
values[resind] = constvalue;
if (!isNull)
null_head[resind] = ' ';
else
null_head[resind] = 'n';
}
}
}
}
}
/*
* form the new result tuple (in the caller's memory context!)
*/
MemoryContextSwitchTo(oldContext);
newTuple = (HeapTuple) heap_formtuple(targettype, values, null_head);
exit:
/*
* free the status arrays if we palloc'd them
*/
if (nodomains > 64) {
pfree(null_head);
pfree(fjIsNull);
pfree(itemIsDone);
}
/* make sure we are in the right context if we did "goto exit" */
MemoryContextSwitchTo(oldContext);
return newTuple;
}
/* ----------------------------------------------------------------
* ExecProject
*
* projects a tuple based in projection info and stores
* it in the specified tuple table slot.
*
* Note: someday soon the executor can be extended to eliminate
* redundant projections by storing pointers to datums
* in the tuple table and then passing these around when
* possible. this should make things much quicker.
* -cim 6/3/91
* ----------------------------------------------------------------
*/
TupleTableSlot *
ExecProject(ProjectionInfo *projInfo, bool *isDone)
{
TupleTableSlot *slot;
List *targetlist;
int len;
TupleDesc tupType;
Datum *tupValue;
ExprContext *econtext;
HeapTuple newTuple;
/*
* sanity checks
*/
if (projInfo == NULL)
return (TupleTableSlot *) NULL;
/*
* get the projection info we want
*/
slot = projInfo->pi_slot;
targetlist = projInfo->pi_targetlist;
len = projInfo->pi_len;
tupType = slot->ttc_tupleDescriptor;
tupValue = projInfo->pi_tupValue;
econtext = projInfo->pi_exprContext;
/*
* form a new (result) tuple
*/
newTuple = ExecTargetList(targetlist,
len,
tupType,
tupValue,
econtext,
isDone);
/*
* store the tuple in the projection slot and return the slot.
*/
return (TupleTableSlot *)
ExecStoreTuple(newTuple,/* tuple to store */
slot, /* slot to store in */
InvalidBuffer, /* tuple has no buffer */ true);
}