postgresql/src/backend/parser/analyze.c

2789 lines
77 KiB
C
Raw Normal View History

/*-------------------------------------------------------------------------
*
* analyze.c
* transform the parse tree into a query tree
*
2002-06-20 22:29:54 +02:00
* Portions Copyright (c) 1996-2002, PostgreSQL Global Development Group
* Portions Copyright (c) 1994, Regents of the University of California
*
* $Header: /cvsroot/pgsql/src/backend/parser/analyze.c,v 1.261 2003/02/09 06:56:28 tgl Exp $
*
*-------------------------------------------------------------------------
*/
1997-11-26 02:14:33 +01:00
#include "postgres.h"
1999-07-16 07:00:38 +02:00
#include "access/heapam.h"
#include "catalog/catname.h"
#include "catalog/heap.h"
#include "catalog/index.h"
#include "catalog/namespace.h"
#include "catalog/pg_index.h"
1999-07-16 07:00:38 +02:00
#include "catalog/pg_type.h"
#include "commands/prepare.h"
#include "nodes/makefuncs.h"
#include "optimizer/clauses.h"
#include "parser/analyze.h"
#include "parser/gramparse.h"
#include "parser/parsetree.h"
#include "parser/parse_agg.h"
1997-11-26 02:14:33 +01:00
#include "parser/parse_clause.h"
#include "parser/parse_coerce.h"
#include "parser/parse_expr.h"
#include "parser/parse_oper.h"
#include "parser/parse_relation.h"
#include "parser/parse_target.h"
#include "parser/parse_type.h"
#include "parser/parse_expr.h"
#include "rewrite/rewriteManip.h"
#include "utils/builtins.h"
#include "utils/fmgroids.h"
#include "utils/lsyscache.h"
#include "utils/relcache.h"
#include "utils/syscache.h"
#include "mb/pg_wchar.h"
/* State shared by transformCreateSchemaStmt and its subroutines */
typedef struct
{
const char *stmtType; /* "CREATE SCHEMA" or "ALTER SCHEMA" */
char *schemaname; /* name of schema */
char *authid; /* owner of schema */
List *tables; /* CREATE TABLE items */
List *views; /* CREATE VIEW items */
List *grants; /* GRANT items */
2002-09-04 22:31:48 +02:00
List *fwconstraints; /* Forward referencing FOREIGN KEY
* constraints */
List *alters; /* Generated ALTER items (from the above) */
List *ixconstraints; /* index-creating constraints */
List *blist; /* "before list" of things to do before
* creating the schema */
List *alist; /* "after list" of things to do after
* creating the schema */
} CreateSchemaStmtContext;
/* State shared by transformCreateStmt and its subroutines */
typedef struct
{
const char *stmtType; /* "CREATE TABLE" or "ALTER TABLE" */
RangeVar *relation; /* relation to create */
List *inhRelations; /* relations to inherit from */
bool hasoids; /* does relation have an OID column? */
Oid relOid; /* OID of table, if ALTER TABLE case */
List *columns; /* ColumnDef items */
List *ckconstraints; /* CHECK constraints */
List *fkconstraints; /* FOREIGN KEY constraints */
List *ixconstraints; /* index-creating constraints */
List *blist; /* "before list" of things to do before
* creating the table */
List *alist; /* "after list" of things to do after
* creating the table */
IndexStmt *pkey; /* PRIMARY KEY index, if any */
} CreateStmtContext;
static Query *transformStmt(ParseState *pstate, Node *stmt,
2002-09-04 22:31:48 +02:00
List **extras_before, List **extras_after);
static Query *transformDeleteStmt(ParseState *pstate, DeleteStmt *stmt);
static Query *transformInsertStmt(ParseState *pstate, InsertStmt *stmt,
2002-09-04 22:31:48 +02:00
List **extras_before, List **extras_after);
static Query *transformIndexStmt(ParseState *pstate, IndexStmt *stmt);
static Query *transformRuleStmt(ParseState *query, RuleStmt *stmt,
2002-09-04 22:31:48 +02:00
List **extras_before, List **extras_after);
1998-01-09 21:06:08 +01:00
static Query *transformSelectStmt(ParseState *pstate, SelectStmt *stmt);
static Query *transformSetOperationStmt(ParseState *pstate, SelectStmt *stmt);
static Node *transformSetOperationTree(ParseState *pstate, SelectStmt *stmt);
1998-01-09 21:06:08 +01:00
static Query *transformUpdateStmt(ParseState *pstate, UpdateStmt *stmt);
static Query *transformPrepareStmt(ParseState *pstate, PrepareStmt *stmt);
static Query *transformExecuteStmt(ParseState *pstate, ExecuteStmt *stmt);
static Query *transformCreateStmt(ParseState *pstate, CreateStmt *stmt,
2002-09-04 22:31:48 +02:00
List **extras_before, List **extras_after);
static Query *transformAlterTableStmt(ParseState *pstate, AlterTableStmt *stmt,
List **extras_before, List **extras_after);
static void transformColumnDefinition(ParseState *pstate,
CreateStmtContext *cxt,
ColumnDef *column);
static void transformTableConstraint(ParseState *pstate,
CreateStmtContext *cxt,
Constraint *constraint);
static void transformIndexConstraints(ParseState *pstate,
CreateStmtContext *cxt);
static void transformFKConstraints(ParseState *pstate,
CreateStmtContext *cxt);
static void applyColumnNames(List *dst, List *src);
static List *getSetColTypes(ParseState *pstate, Node *node);
1999-05-25 18:15:34 +02:00
static void transformForUpdate(Query *qry, List *forUpdate);
static void transformConstraintAttrs(List *constraintList);
static void transformColumnType(ParseState *pstate, ColumnDef *column);
static bool relationHasPrimaryKey(Oid relationOid);
static void release_pstate_resources(ParseState *pstate);
static FromExpr *makeFromExpr(List *fromlist, Node *quals);
1996-10-30 03:02:41 +01:00
/*
* parse_analyze -
* analyze a raw parse tree and transform it to Query form.
*
* The result is a List of Query nodes (we need a list since some commands
* produce multiple Queries). Optimizable statements require considerable
* transformation, while many utility-type statements are simply hung off
* a dummy CMD_UTILITY Query node.
*/
List *
parse_analyze(Node *parseTree, ParseState *parentParseState)
{
1999-05-25 18:15:34 +02:00
List *result = NIL;
ParseState *pstate = make_parsestate(parentParseState);
2002-09-04 22:31:48 +02:00
/* Lists to return extra commands from transformation */
List *extras_before = NIL;
List *extras_after = NIL;
Query *query;
List *listscan;
query = transformStmt(pstate, parseTree, &extras_before, &extras_after);
release_pstate_resources(pstate);
while (extras_before != NIL)
{
result = nconc(result, parse_analyze(lfirst(extras_before), pstate));
extras_before = lnext(extras_before);
}
1998-01-20 06:05:08 +01:00
result = lappend(result, query);
1998-10-28 17:06:54 +01:00
while (extras_after != NIL)
{
result = nconc(result, parse_analyze(lfirst(extras_after), pstate));
extras_after = lnext(extras_after);
}
/*
2002-09-04 22:31:48 +02:00
* Make sure that only the original query is marked original. We have
* to do this explicitly since recursive calls of parse_analyze will
* have marked some of the added-on queries as "original".
*/
foreach(listscan, result)
{
2002-09-04 22:31:48 +02:00
Query *q = lfirst(listscan);
q->querySource = (q == query ? QSRC_ORIGINAL : QSRC_PARSER);
}
pfree(pstate);
return result;
}
static void
release_pstate_resources(ParseState *pstate)
{
if (pstate->p_target_relation != NULL)
heap_close(pstate->p_target_relation, NoLock);
pstate->p_target_relation = NULL;
pstate->p_target_rangetblentry = NULL;
}
/*
* transformStmt -
* transform a Parse tree into a Query tree.
*/
static Query *
transformStmt(ParseState *pstate, Node *parseTree,
2002-09-04 22:31:48 +02:00
List **extras_before, List **extras_after)
{
Query *result = NULL;
switch (nodeTag(parseTree))
{
/*
* Non-optimizable statements
*/
case T_CreateStmt:
result = transformCreateStmt(pstate, (CreateStmt *) parseTree,
2002-09-04 22:31:48 +02:00
extras_before, extras_after);
break;
case T_IndexStmt:
result = transformIndexStmt(pstate, (IndexStmt *) parseTree);
break;
case T_RuleStmt:
result = transformRuleStmt(pstate, (RuleStmt *) parseTree,
2002-09-04 22:31:48 +02:00
extras_before, extras_after);
break;
case T_ViewStmt:
{
ViewStmt *n = (ViewStmt *) parseTree;
n->query = transformStmt(pstate, (Node *) n->query,
2002-09-04 22:31:48 +02:00
extras_before, extras_after);
/*
* If a list of column names was given, run through and
* insert these into the actual query tree. - thomas
* 2000-03-08
*
* Outer loop is over targetlist to make it easier to skip
* junk targetlist entries.
*/
if (n->aliases != NIL)
{
List *aliaslist = n->aliases;
List *targetList;
foreach(targetList, n->query->targetList)
{
TargetEntry *te = (TargetEntry *) lfirst(targetList);
Resdom *rd;
Assert(IsA(te, TargetEntry));
rd = te->resdom;
Assert(IsA(rd, Resdom));
/* junk columns don't get aliases */
if (rd->resjunk)
continue;
rd->resname = pstrdup(strVal(lfirst(aliaslist)));
aliaslist = lnext(aliaslist);
if (aliaslist == NIL)
break; /* done assigning aliases */
}
if (aliaslist != NIL)
elog(ERROR, "CREATE VIEW specifies more column names than columns");
}
result = makeNode(Query);
result->commandType = CMD_UTILITY;
result->utilityStmt = (Node *) n;
}
break;
case T_ExplainStmt:
{
ExplainStmt *n = (ExplainStmt *) parseTree;
result = makeNode(Query);
result->commandType = CMD_UTILITY;
n->query = transformStmt(pstate, (Node *) n->query,
2002-09-04 22:31:48 +02:00
extras_before, extras_after);
result->utilityStmt = (Node *) parseTree;
}
break;
case T_AlterTableStmt:
result = transformAlterTableStmt(pstate, (AlterTableStmt *) parseTree,
2002-09-04 22:31:48 +02:00
extras_before, extras_after);
break;
case T_PrepareStmt:
result = transformPrepareStmt(pstate, (PrepareStmt *) parseTree);
break;
case T_ExecuteStmt:
result = transformExecuteStmt(pstate, (ExecuteStmt *) parseTree);
break;
/*
* Optimizable statements
*/
1998-01-09 21:06:08 +01:00
case T_InsertStmt:
result = transformInsertStmt(pstate, (InsertStmt *) parseTree,
2002-09-04 22:31:48 +02:00
extras_before, extras_after);
break;
case T_DeleteStmt:
result = transformDeleteStmt(pstate, (DeleteStmt *) parseTree);
break;
1998-01-09 21:06:08 +01:00
case T_UpdateStmt:
result = transformUpdateStmt(pstate, (UpdateStmt *) parseTree);
break;
1998-01-09 21:06:08 +01:00
case T_SelectStmt:
if (((SelectStmt *) parseTree)->op == SETOP_NONE)
result = transformSelectStmt(pstate,
(SelectStmt *) parseTree);
else
result = transformSetOperationStmt(pstate,
2001-03-22 05:01:46 +01:00
(SelectStmt *) parseTree);
break;
default:
/*
* other statements don't require any transformation-- just
1998-01-20 06:05:08 +01:00
* return the original parsetree, yea!
*/
result = makeNode(Query);
result->commandType = CMD_UTILITY;
result->utilityStmt = (Node *) parseTree;
break;
}
return result;
}
/*
* transformDeleteStmt -
* transforms a Delete Statement
*/
static Query *
transformDeleteStmt(ParseState *pstate, DeleteStmt *stmt)
{
Query *qry = makeNode(Query);
Node *qual;
qry->commandType = CMD_DELETE;
/* set up range table with just the result rel */
qry->resultRelation = setTargetTable(pstate, stmt->relation,
2002-09-04 22:31:48 +02:00
interpretInhOption(stmt->relation->inhOpt),
true);
qry->distinctClause = NIL;
/* fix where clause */
qual = transformWhereClause(pstate, stmt->whereClause);
/* done building the range table and jointree */
qry->rtable = pstate->p_rtable;
qry->jointree = makeFromExpr(pstate->p_joinlist, qual);
qry->hasSubLinks = pstate->p_hasSubLinks;
1998-01-20 06:05:08 +01:00
qry->hasAggs = pstate->p_hasAggs;
if (pstate->p_hasAggs)
parseCheckAggregates(pstate, qry);
return qry;
}
/*
* transformInsertStmt -
* transform an Insert Statement
*/
static Query *
transformInsertStmt(ParseState *pstate, InsertStmt *stmt,
List **extras_before, List **extras_after)
{
Query *qry = makeNode(Query);
List *sub_rtable;
List *sub_namespace;
1997-10-12 09:09:20 +02:00
List *icolumns;
List *attrnos;
List *attnos;
List *tl;
qry->commandType = CMD_INSERT;
pstate->p_is_insert = true;
/*
* If a non-nil rangetable/namespace was passed in, and we are doing
* INSERT/SELECT, arrange to pass the rangetable/namespace down to the
2001-03-22 05:01:46 +01:00
* SELECT. This can only happen if we are inside a CREATE RULE, and
* in that case we want the rule's OLD and NEW rtable entries to
* appear as part of the SELECT's rtable, not as outer references for
2001-03-22 05:01:46 +01:00
* it. (Kluge!) The SELECT's joinlist is not affected however. We
* must do this before adding the target table to the INSERT's rtable.
*/
if (stmt->selectStmt)
{
sub_rtable = pstate->p_rtable;
pstate->p_rtable = NIL;
sub_namespace = pstate->p_namespace;
pstate->p_namespace = NIL;
}
else
{
sub_rtable = NIL; /* not used, but keep compiler quiet */
sub_namespace = NIL;
}
/*
* Must get write lock on INSERT target table before scanning SELECT,
* else we will grab the wrong kind of initial lock if the target
2001-03-22 05:01:46 +01:00
* table is also mentioned in the SELECT part. Note that the target
* table is not added to the joinlist or namespace.
*/
qry->resultRelation = setTargetTable(pstate, stmt->relation,
false, false);
/*
* Is it INSERT ... SELECT or INSERT ... VALUES?
1999-05-25 18:15:34 +02:00
*/
if (stmt->selectStmt)
{
ParseState *sub_pstate = make_parsestate(pstate->parentParseState);
2001-03-22 05:01:46 +01:00
Query *selectQuery;
RangeTblEntry *rte;
RangeTblRef *rtr;
1999-05-25 18:15:34 +02:00
/*
* Process the source SELECT.
*
2001-03-22 05:01:46 +01:00
* It is important that this be handled just like a standalone
* SELECT; otherwise the behavior of SELECT within INSERT might be
* different from a stand-alone SELECT. (Indeed, Postgres up
* through 6.5 had bugs of just that nature...)
*/
sub_pstate->p_rtable = sub_rtable;
sub_pstate->p_namespace = sub_namespace;
/*
* Note: we are not expecting that extras_before and extras_after
2002-09-04 22:31:48 +02:00
* are going to be used by the transformation of the SELECT
* statement.
*/
selectQuery = transformStmt(sub_pstate, stmt->selectStmt,
2002-09-04 22:31:48 +02:00
extras_before, extras_after);
release_pstate_resources(sub_pstate);
pfree(sub_pstate);
Assert(IsA(selectQuery, Query));
Assert(selectQuery->commandType == CMD_SELECT);
if (selectQuery->into || selectQuery->isPortal)
elog(ERROR, "INSERT ... SELECT may not specify INTO");
2001-03-22 05:01:46 +01:00
/*
2001-03-22 05:01:46 +01:00
* Make the source be a subquery in the INSERT's rangetable, and
* add it to the INSERT's joinlist.
*/
rte = addRangeTableEntryForSubquery(pstate,
selectQuery,
makeAlias("*SELECT*", NIL),
true);
rtr = makeNode(RangeTblRef);
/* assume new rte is at end */
rtr->rtindex = length(pstate->p_rtable);
Assert(rte == rt_fetch(rtr->rtindex, pstate->p_rtable));
pstate->p_joinlist = lappend(pstate->p_joinlist, rtr);
2001-03-22 05:01:46 +01:00
/*
2001-03-22 05:01:46 +01:00
* Generate a targetlist for the INSERT that selects all the
* non-resjunk columns from the subquery. (We need this to be
* separate from the subquery's tlist because we may add columns,
* insert datatype coercions, etc.)
*
* HACK: constants in the INSERT's targetlist are copied up as-is
2001-03-22 05:01:46 +01:00
* rather than being referenced as subquery outputs. This is
* mainly to ensure that when we try to coerce them to the target
* column's datatype, the right things happen for UNKNOWN
* constants. Otherwise this fails: INSERT INTO foo SELECT 'bar',
* ... FROM baz
*/
qry->targetList = NIL;
foreach(tl, selectQuery->targetList)
{
TargetEntry *tle = (TargetEntry *) lfirst(tl);
Resdom *resnode = tle->resdom;
Expr *expr;
if (resnode->resjunk)
continue;
if (tle->expr && IsA(tle->expr, Const))
expr = tle->expr;
else
expr = (Expr *) makeVar(rtr->rtindex,
resnode->resno,
resnode->restype,
resnode->restypmod,
0);
resnode = copyObject(resnode);
resnode->resno = (AttrNumber) pstate->p_last_resno++;
qry->targetList = lappend(qry->targetList,
makeTargetEntry(resnode, expr));
}
}
else
{
/*
2001-03-22 05:01:46 +01:00
* For INSERT ... VALUES, transform the given list of values to
* form a targetlist for the INSERT.
*/
qry->targetList = transformTargetList(pstate, stmt->targetList);
}
1998-01-11 04:41:57 +01:00
/*
* Now we are done with SELECT-like processing, and can get on with
* transforming the target list to match the INSERT target columns.
*/
/* Prepare to assign non-conflicting resnos to resjunk attributes */
if (pstate->p_last_resno <= pstate->p_target_relation->rd_rel->relnatts)
pstate->p_last_resno = pstate->p_target_relation->rd_rel->relnatts + 1;
/* Validate stmt->cols list, or build default list if no list given */
icolumns = checkInsertTargets(pstate, stmt->cols, &attrnos);
/*
* Prepare columns for assignment to target table.
*/
attnos = attrnos;
/* cannot use foreach here because of possible lremove */
tl = qry->targetList;
while (tl)
{
TargetEntry *tle = (TargetEntry *) lfirst(tl);
2002-09-04 22:31:48 +02:00
ResTarget *col;
/* must advance tl before lremove possibly pfree's it */
tl = lnext(tl);
if (icolumns == NIL || attnos == NIL)
elog(ERROR, "INSERT has more expressions than target columns");
col = (ResTarget *) lfirst(icolumns);
Assert(IsA(col, ResTarget));
2002-04-05 13:56:55 +02:00
/*
* When the value is to be set to the column default we can simply
* drop the TLE now and handle it later on using methods for missing
2002-04-05 13:56:55 +02:00
* columns.
*/
if (IsA(tle, InsertDefault))
2002-04-05 13:56:55 +02:00
{
qry->targetList = lremove(tle, qry->targetList);
/* Note: the stmt->cols list is not adjusted to match */
2002-04-05 13:56:55 +02:00
}
else
{
/* Normal case */
Assert(!tle->resdom->resjunk);
updateTargetListEntry(pstate, tle, col->name, lfirsti(attnos),
col->indirection);
2002-04-05 13:56:55 +02:00
}
icolumns = lnext(icolumns);
attnos = lnext(attnos);
}
/*
* Ensure that the targetlist has the same number of entries that
* were present in the columns list. Don't do the check unless
* an explicit columns list was given, though.
2002-09-04 22:31:48 +02:00
* statements.
*/
if (stmt->cols != NIL && (icolumns != NIL || attnos != NIL))
elog(ERROR, "INSERT has more target columns than expressions");
/* done building the range table and jointree */
qry->rtable = pstate->p_rtable;
qry->jointree = makeFromExpr(pstate->p_joinlist, NULL);
1999-01-21 17:08:55 +01:00
qry->hasSubLinks = pstate->p_hasSubLinks;
qry->hasAggs = pstate->p_hasAggs;
if (pstate->p_hasAggs)
parseCheckAggregates(pstate, qry);
return qry;
}
1998-01-20 06:05:08 +01:00
/*
* makeObjectName()
*
* Create a name for an implicitly created index, sequence, constraint, etc.
*
* The parameters are: the original table name, the original field name, and
* a "type" string (such as "seq" or "pkey"). The field name and/or type
* can be NULL if not relevant.
*
* The result is a palloc'd string.
*
* The basic result we want is "name1_name2_type", omitting "_name2" or
* "_type" when those parameters are NULL. However, we must generate
* a name with less than NAMEDATALEN characters! So, we truncate one or
* both names if necessary to make a short-enough string. The type part
* is never truncated (so it had better be reasonably short).
*
* To reduce the probability of collisions, we might someday add more
* smarts to this routine, like including some "hash" characters computed
* from the truncated characters. Currently it seems best to keep it simple,
* so that the generated names are easily predictable by a person.
*/
char *
makeObjectName(char *name1, char *name2, char *typename)
{
char *name;
int overhead = 0; /* chars needed for type and underscores */
int availchars; /* chars available for name(s) */
int name1chars; /* chars allocated to name1 */
int name2chars; /* chars allocated to name2 */
int ndx;
name1chars = strlen(name1);
if (name2)
{
name2chars = strlen(name2);
overhead++; /* allow for separating underscore */
}
else
name2chars = 0;
if (typename)
overhead += strlen(typename) + 1;
availchars = NAMEDATALEN - 1 - overhead;
/*
* If we must truncate, preferentially truncate the longer name. This
* logic could be expressed without a loop, but it's simple and
* obvious as a loop.
*/
while (name1chars + name2chars > availchars)
{
if (name1chars > name2chars)
name1chars--;
else
name2chars--;
}
if (name1)
name1chars = pg_mbcliplen(name1, name1chars, name1chars);
if (name2)
name2chars = pg_mbcliplen(name2, name2chars, name2chars);
/* Now construct the string using the chosen lengths */
name = palloc(name1chars + name2chars + overhead + 1);
strncpy(name, name1, name1chars);
ndx = name1chars;
if (name2)
{
name[ndx++] = '_';
strncpy(name + ndx, name2, name2chars);
ndx += name2chars;
}
if (typename)
{
name[ndx++] = '_';
strcpy(name + ndx, typename);
}
else
name[ndx] = '\0';
1998-09-01 05:29:17 +02:00
return name;
1998-01-20 06:05:08 +01:00
}
1998-01-20 06:05:08 +01:00
static char *
CreateIndexName(char *table_name, char *column_name,
char *label, List *indices)
{
int pass = 0;
char *iname = NULL;
List *ilist;
char typename[NAMEDATALEN];
/*
* The type name for makeObjectName is label, or labelN if that's
* necessary to prevent collisions among multiple indexes for the same
* table. Note there is no check for collisions with already-existing
2001-03-22 05:01:46 +01:00
* indexes, only among the indexes we're about to create now; this
* ought to be improved someday.
*/
strncpy(typename, label, sizeof(typename));
for (;;)
{
iname = makeObjectName(table_name, column_name, typename);
foreach(ilist, indices)
{
IndexStmt *index = lfirst(ilist);
if (index->idxname != NULL &&
strcmp(iname, index->idxname) == 0)
break;
}
/* ran through entire list? then no name conflict found so done */
if (ilist == NIL)
break;
/* found a conflict, so try a new name component */
pfree(iname);
snprintf(typename, sizeof(typename), "%s%d", label, ++pass);
}
1998-09-01 05:29:17 +02:00
return iname;
1998-01-20 06:05:08 +01:00
}
/*
* transformCreateStmt -
* transforms the "create table" statement
* SQL92 allows constraints to be scattered all over, so thumb through
* the columns and collect all constraints into one place.
* If there are any implied indices (e.g. UNIQUE or PRIMARY KEY)
* then expand those into multiple IndexStmt blocks.
* - thomas 1997-12-02
*/
static Query *
transformCreateStmt(ParseState *pstate, CreateStmt *stmt,
List **extras_before, List **extras_after)
{
CreateStmtContext cxt;
Query *q;
List *elements;
cxt.stmtType = "CREATE TABLE";
cxt.relation = stmt->relation;
cxt.inhRelations = stmt->inhRelations;
cxt.hasoids = stmt->hasoids;
cxt.relOid = InvalidOid;
cxt.columns = NIL;
cxt.ckconstraints = NIL;
cxt.fkconstraints = NIL;
cxt.ixconstraints = NIL;
cxt.blist = NIL;
cxt.alist = NIL;
cxt.pkey = NULL;
/*
* Run through each primary element in the table creation clause.
* Separate column defs from constraints, and do preliminary analysis.
*/
foreach(elements, stmt->tableElts)
{
Node *element = lfirst(elements);
switch (nodeTag(element))
{
case T_ColumnDef:
transformColumnDefinition(pstate, &cxt,
(ColumnDef *) element);
break;
case T_Constraint:
transformTableConstraint(pstate, &cxt,
(Constraint *) element);
break;
case T_FkConstraint:
/* No pre-transformation needed */
cxt.fkconstraints = lappend(cxt.fkconstraints, element);
break;
default:
elog(ERROR, "parser: unrecognized node (internal error)");
}
}
Assert(stmt->constraints == NIL);
/*
* Postprocess constraints that give rise to index definitions.
*/
transformIndexConstraints(pstate, &cxt);
/*
* Postprocess foreign-key constraints.
*/
transformFKConstraints(pstate, &cxt);
/*
* Output results.
*/
q = makeNode(Query);
q->commandType = CMD_UTILITY;
q->utilityStmt = (Node *) stmt;
stmt->tableElts = cxt.columns;
stmt->constraints = cxt.ckconstraints;
2002-09-04 22:31:48 +02:00
*extras_before = nconc(*extras_before, cxt.blist);
*extras_after = nconc(cxt.alist, *extras_after);
return q;
}
static void
transformColumnDefinition(ParseState *pstate, CreateStmtContext *cxt,
ColumnDef *column)
{
bool is_serial;
bool saw_nullable;
Constraint *constraint;
List *clist;
cxt->columns = lappend(cxt->columns, column);
/* Check for SERIAL pseudo-types */
is_serial = false;
if (length(column->typename->names) == 1)
{
char *typname = strVal(lfirst(column->typename->names));
if (strcmp(typname, "serial") == 0 ||
strcmp(typname, "serial4") == 0)
{
is_serial = true;
column->typename->names = NIL;
column->typename->typeid = INT4OID;
}
else if (strcmp(typname, "bigserial") == 0 ||
strcmp(typname, "serial8") == 0)
{
is_serial = true;
column->typename->names = NIL;
column->typename->typeid = INT8OID;
}
}
/* Do necessary work on the column type declaration */
transformColumnType(pstate, column);
/* Special actions for SERIAL pseudo-types */
if (is_serial)
{
char *sname;
char *snamespace;
char *qstring;
A_Const *snamenode;
FuncCall *funccallnode;
CreateSeqStmt *seqstmt;
/*
* Determine name and namespace to use for the sequence.
*/
sname = makeObjectName(cxt->relation->relname, column->colname, "seq");
snamespace = get_namespace_name(RangeVarGetCreationNamespace(cxt->relation));
elog(NOTICE, "%s will create implicit sequence '%s' for SERIAL column '%s.%s'",
2002-09-04 22:31:48 +02:00
cxt->stmtType, sname, cxt->relation->relname, column->colname);
/*
* Build a CREATE SEQUENCE command to create the sequence object,
* and add it to the list of things to be done before this
* CREATE/ALTER TABLE.
*/
seqstmt = makeNode(CreateSeqStmt);
seqstmt->sequence = makeRangeVar(snamespace, sname);
seqstmt->options = NIL;
cxt->blist = lappend(cxt->blist, seqstmt);
/*
* Mark the ColumnDef so that during execution, an appropriate
* dependency will be added from the sequence to the column.
*/
column->support = makeRangeVar(snamespace, sname);
/*
* Create appropriate constraints for SERIAL. We do this in full,
* rather than shortcutting, so that we will detect any
* conflicting constraints the user wrote (like a different
* DEFAULT).
*
* Create an expression tree representing the function call
* nextval('"sequencename"')
*/
qstring = quote_qualified_identifier(snamespace, sname);
snamenode = makeNode(A_Const);
snamenode->val.type = T_String;
snamenode->val.val.str = qstring;
funccallnode = makeNode(FuncCall);
funccallnode->funcname = SystemFuncName("nextval");
funccallnode->args = makeList1(snamenode);
funccallnode->agg_star = false;
funccallnode->agg_distinct = false;
constraint = makeNode(Constraint);
constraint->contype = CONSTR_DEFAULT;
constraint->name = sname;
constraint->raw_expr = (Node *) funccallnode;
constraint->cooked_expr = NULL;
constraint->keys = NIL;
column->constraints = lappend(column->constraints, constraint);
constraint = makeNode(Constraint);
constraint->contype = CONSTR_NOTNULL;
column->constraints = lappend(column->constraints, constraint);
}
/* Process column constraints, if any... */
transformConstraintAttrs(column->constraints);
saw_nullable = false;
foreach(clist, column->constraints)
{
constraint = lfirst(clist);
/*
* If this column constraint is a FOREIGN KEY constraint, then we
* fill in the current attribute's name and throw it into the list
* of FK constraints to be processed later.
*/
if (IsA(constraint, FkConstraint))
{
FkConstraint *fkconstraint = (FkConstraint *) constraint;
fkconstraint->fk_attrs = makeList1(makeString(column->colname));
cxt->fkconstraints = lappend(cxt->fkconstraints, fkconstraint);
continue;
}
Assert(IsA(constraint, Constraint));
switch (constraint->contype)
{
case CONSTR_NULL:
if (saw_nullable && column->is_not_null)
elog(ERROR, "%s/(NOT) NULL conflicting declaration for '%s.%s'",
cxt->stmtType, cxt->relation->relname, column->colname);
column->is_not_null = FALSE;
saw_nullable = true;
break;
case CONSTR_NOTNULL:
if (saw_nullable && !column->is_not_null)
elog(ERROR, "%s/(NOT) NULL conflicting declaration for '%s.%s'",
cxt->stmtType, cxt->relation->relname, column->colname);
column->is_not_null = TRUE;
saw_nullable = true;
break;
case CONSTR_DEFAULT:
if (column->raw_default != NULL)
elog(ERROR, "%s/DEFAULT multiple values specified for '%s.%s'",
cxt->stmtType, cxt->relation->relname, column->colname);
column->raw_default = constraint->raw_expr;
Assert(constraint->cooked_expr == NULL);
break;
case CONSTR_PRIMARY:
if (constraint->name == NULL)
constraint->name = makeObjectName(cxt->relation->relname,
NULL,
"pkey");
if (constraint->keys == NIL)
constraint->keys = makeList1(makeString(column->colname));
cxt->ixconstraints = lappend(cxt->ixconstraints, constraint);
break;
case CONSTR_UNIQUE:
if (constraint->name == NULL)
constraint->name = makeObjectName(cxt->relation->relname,
column->colname,
"key");
if (constraint->keys == NIL)
constraint->keys = makeList1(makeString(column->colname));
cxt->ixconstraints = lappend(cxt->ixconstraints, constraint);
break;
case CONSTR_CHECK:
if (constraint->name == NULL)
constraint->name = makeObjectName(cxt->relation->relname,
column->colname,
NULL);
cxt->ckconstraints = lappend(cxt->ckconstraints, constraint);
break;
case CONSTR_ATTR_DEFERRABLE:
case CONSTR_ATTR_NOT_DEFERRABLE:
case CONSTR_ATTR_DEFERRED:
case CONSTR_ATTR_IMMEDIATE:
/* transformConstraintAttrs took care of these */
break;
default:
elog(ERROR, "parser: unrecognized constraint (internal error)");
break;
}
}
}
static void
transformTableConstraint(ParseState *pstate, CreateStmtContext *cxt,
Constraint *constraint)
{
switch (constraint->contype)
{
case CONSTR_PRIMARY:
if (constraint->name == NULL)
constraint->name = makeObjectName(cxt->relation->relname,
NULL,
"pkey");
cxt->ixconstraints = lappend(cxt->ixconstraints, constraint);
break;
case CONSTR_UNIQUE:
cxt->ixconstraints = lappend(cxt->ixconstraints, constraint);
break;
case CONSTR_CHECK:
cxt->ckconstraints = lappend(cxt->ckconstraints, constraint);
break;
case CONSTR_NULL:
case CONSTR_NOTNULL:
case CONSTR_DEFAULT:
case CONSTR_ATTR_DEFERRABLE:
case CONSTR_ATTR_NOT_DEFERRABLE:
case CONSTR_ATTR_DEFERRED:
case CONSTR_ATTR_IMMEDIATE:
elog(ERROR, "parser: illegal context for constraint (internal error)");
break;
default:
elog(ERROR, "parser: unrecognized constraint (internal error)");
break;
}
}
static void
transformIndexConstraints(ParseState *pstate, CreateStmtContext *cxt)
{
List *listptr;
List *keys;
IndexStmt *index;
IndexElem *iparam;
ColumnDef *column;
List *columns;
List *indexlist = NIL;
/*
* Run through the constraints that need to generate an index. For
* PRIMARY KEY, mark each column as NOT NULL and create an index. For
* UNIQUE, create an index as for PRIMARY KEY, but do not insist on
* NOT NULL.
*/
foreach(listptr, cxt->ixconstraints)
{
Constraint *constraint = lfirst(listptr);
Assert(IsA(constraint, Constraint));
Assert((constraint->contype == CONSTR_PRIMARY)
1999-05-25 18:15:34 +02:00
|| (constraint->contype == CONSTR_UNIQUE));
index = makeNode(IndexStmt);
index->unique = true;
index->primary = (constraint->contype == CONSTR_PRIMARY);
if (index->primary)
{
/* In ALTER TABLE case, a primary index might already exist */
if (cxt->pkey != NULL ||
(OidIsValid(cxt->relOid) &&
relationHasPrimaryKey(cxt->relOid)))
elog(ERROR, "%s / PRIMARY KEY multiple primary keys"
" for table '%s' are not allowed",
cxt->stmtType, cxt->relation->relname);
cxt->pkey = index;
}
index->isconstraint = true;
if (constraint->name != NULL)
index->idxname = pstrdup(constraint->name);
else if (constraint->contype == CONSTR_PRIMARY)
index->idxname = makeObjectName(cxt->relation->relname, NULL, "pkey");
else
2001-03-22 05:01:46 +01:00
index->idxname = NULL; /* will set it later */
index->relation = cxt->relation;
index->accessMethod = DEFAULT_INDEX_TYPE;
index->indexParams = NIL;
index->whereClause = NULL;
/*
* Make sure referenced keys exist. If we are making a PRIMARY
* KEY index, also make sure they are NOT NULL, if possible.
* (Although we could leave it to DefineIndex to mark the columns NOT
* NULL, it's more efficient to get it right the first time.)
*/
2001-03-22 05:01:46 +01:00
foreach(keys, constraint->keys)
{
char *key = strVal(lfirst(keys));
2001-03-22 05:01:46 +01:00
bool found = false;
column = NULL;
foreach(columns, cxt->columns)
{
column = lfirst(columns);
Assert(IsA(column, ColumnDef));
if (strcmp(column->colname, key) == 0)
{
found = true;
break;
}
}
if (found)
{
/* found column in the new table; force it to be NOT NULL */
if (constraint->contype == CONSTR_PRIMARY)
column->is_not_null = TRUE;
}
else if (SystemAttributeByName(key, cxt->hasoids) != NULL)
{
/*
* column will be a system column in the new table, so
* accept it. System columns can't ever be null, so no
* need to worry about PRIMARY/NOT NULL constraint.
*/
found = true;
}
else if (cxt->inhRelations)
{
/* try inherited tables */
2001-03-22 05:01:46 +01:00
List *inher;
foreach(inher, cxt->inhRelations)
{
RangeVar *inh = lfirst(inher);
2001-03-22 05:01:46 +01:00
Relation rel;
int count;
Assert(IsA(inh, RangeVar));
rel = heap_openrv(inh, AccessShareLock);
if (rel->rd_rel->relkind != RELKIND_RELATION)
elog(ERROR, "inherited table \"%s\" is not a relation",
inh->relname);
for (count = 0; count < rel->rd_att->natts; count++)
{
Form_pg_attribute inhattr = rel->rd_att->attrs[count];
2001-03-22 05:01:46 +01:00
char *inhname = NameStr(inhattr->attname);
if (inhattr->attisdropped)
continue;
if (strcmp(key, inhname) == 0)
{
found = true;
/*
* We currently have no easy way to force an
* inherited column to be NOT NULL at creation, if
* its parent wasn't so already. We leave it to
* DefineIndex to fix things up in this case.
*/
break;
}
}
heap_close(rel, NoLock);
if (found)
break;
}
}
else if (OidIsValid(cxt->relOid))
{
/* ALTER TABLE case: does column already exist? */
HeapTuple atttuple;
atttuple = SearchSysCacheAttName(cxt->relOid, key);
if (HeapTupleIsValid(atttuple))
{
found = true;
/*
* If it's not already NOT NULL, leave it to DefineIndex
* to fix later.
*/
ReleaseSysCache(atttuple);
}
}
if (!found)
elog(ERROR, "%s: column \"%s\" named in key does not exist",
cxt->stmtType, key);
/* Check for PRIMARY KEY(foo, foo) */
foreach(columns, index->indexParams)
{
iparam = (IndexElem *) lfirst(columns);
if (iparam->name && strcmp(key, iparam->name) == 0)
elog(ERROR, "%s: column \"%s\" appears twice in %s constraint",
cxt->stmtType, key,
index->primary ? "PRIMARY KEY" : "UNIQUE");
}
/* OK, add it to the index definition */
iparam = makeNode(IndexElem);
iparam->name = pstrdup(key);
iparam->funcname = NIL;
iparam->args = NIL;
iparam->opclass = NIL;
index->indexParams = lappend(index->indexParams, iparam);
}
indexlist = lappend(indexlist, index);
}
/*
* Scan the index list and remove any redundant index specifications.
2002-09-04 22:31:48 +02:00
* This can happen if, for instance, the user writes UNIQUE PRIMARY
* KEY. A strict reading of SQL92 would suggest raising an error
* instead, but that strikes me as too anal-retentive. - tgl
* 2001-02-14
*
* XXX in ALTER TABLE case, it'd be nice to look for duplicate
* pre-existing indexes, too.
*/
cxt->alist = NIL;
if (cxt->pkey != NULL)
{
/* Make sure we keep the PKEY index in preference to others... */
cxt->alist = makeList1(cxt->pkey);
}
while (indexlist != NIL)
{
index = lfirst(indexlist);
/* if it's pkey, it's already in cxt->alist */
if (index != cxt->pkey)
{
bool keep = true;
List *priorlist;
foreach(priorlist, cxt->alist)
{
IndexStmt *priorindex = lfirst(priorlist);
1999-05-25 18:15:34 +02:00
if (equal(index->indexParams, priorindex->indexParams))
{
/*
* If the prior index is as yet unnamed, and this one
2001-03-22 05:01:46 +01:00
* is named, then transfer the name to the prior
* index. This ensures that if we have named and
* unnamed constraints, we'll use (at least one of)
* the names for the index.
*/
if (priorindex->idxname == NULL)
priorindex->idxname = index->idxname;
keep = false;
break;
}
}
if (keep)
cxt->alist = lappend(cxt->alist, index);
}
indexlist = lnext(indexlist);
}
/*
* Finally, select unique names for all not-previously-named indices,
* and display NOTICE messages.
*
* XXX in ALTER TABLE case, we fail to consider name collisions against
* pre-existing indexes.
*/
foreach(indexlist, cxt->alist)
{
index = lfirst(indexlist);
if (index->idxname == NULL && index->indexParams != NIL)
{
iparam = lfirst(index->indexParams);
index->idxname = CreateIndexName(cxt->relation->relname,
iparam->name ? iparam->name :
2002-09-04 22:31:48 +02:00
strVal(llast(iparam->funcname)),
"key", cxt->alist);
}
if (index->idxname == NULL) /* should not happen */
elog(ERROR, "%s: failed to make implicit index name",
cxt->stmtType);
elog(NOTICE, "%s / %s%s will create implicit index '%s' for table '%s'",
cxt->stmtType,
(strcmp(cxt->stmtType, "ALTER TABLE") == 0) ? "ADD " : "",
(index->primary ? "PRIMARY KEY" : "UNIQUE"),
index->idxname, cxt->relation->relname);
}
}
static void
transformFKConstraints(ParseState *pstate, CreateStmtContext *cxt)
{
if (cxt->fkconstraints == NIL)
return;
elog(NOTICE, "%s will create implicit trigger(s) for FOREIGN KEY check(s)",
cxt->stmtType);
/*
* For ALTER TABLE ADD CONSTRAINT, nothing to do. For CREATE TABLE,
* gin up an ALTER TABLE ADD CONSTRAINT command to execute after
* the basic CREATE TABLE is complete.
*
* Note: the ADD CONSTRAINT command must also execute after any index
* creation commands. Thus, this should run after
* transformIndexConstraints, so that the CREATE INDEX commands are
* already in cxt->alist.
*/
if (strcmp(cxt->stmtType, "CREATE TABLE") == 0)
{
AlterTableStmt *alterstmt = makeNode(AlterTableStmt);
List *fkclist;
alterstmt->subtype = 'c'; /* preprocessed add constraint */
alterstmt->relation = cxt->relation;
alterstmt->name = NULL;
alterstmt->def = (Node *) cxt->fkconstraints;
/* Don't need to scan the table contents in this case */
foreach(fkclist, cxt->fkconstraints)
{
FkConstraint *fkconstraint = (FkConstraint *) lfirst(fkclist);
2001-03-22 05:01:46 +01:00
fkconstraint->skip_validation = true;
}
cxt->alist = lappend(cxt->alist, (Node *) alterstmt);
}
}
/*
* transformIndexStmt -
* transforms the qualification of the index statement
*/
static Query *
transformIndexStmt(ParseState *pstate, IndexStmt *stmt)
{
1998-02-06 17:46:29 +01:00
Query *qry;
RangeTblEntry *rte;
1998-02-06 17:46:29 +01:00
qry = makeNode(Query);
qry->commandType = CMD_UTILITY;
/* take care of the where clause */
if (stmt->whereClause)
{
/*
* Put the parent table into the rtable so that the WHERE clause
* can refer to its fields without qualification. Note that this
* only works if the parent table already exists --- so we can't
* easily support predicates on indexes created implicitly by
* CREATE TABLE. Fortunately, that's not necessary.
*/
rte = addRangeTableEntry(pstate, stmt->relation, NULL, false, true);
/* no to join list, yes to namespace */
addRTEtoQuery(pstate, rte, false, true);
stmt->whereClause = transformWhereClause(pstate, stmt->whereClause);
}
1998-02-06 17:46:29 +01:00
qry->hasSubLinks = pstate->p_hasSubLinks;
stmt->rangetable = pstate->p_rtable;
1998-02-06 17:46:29 +01:00
qry->utilityStmt = (Node *) stmt;
1998-02-06 17:46:29 +01:00
return qry;
}
/*
* transformRuleStmt -
* transform a Create Rule Statement. The actions is a list of parse
* trees which is transformed into a list of query trees.
*/
static Query *
transformRuleStmt(ParseState *pstate, RuleStmt *stmt,
2002-09-04 22:31:48 +02:00
List **extras_before, List **extras_after)
{
1998-02-06 17:46:29 +01:00
Query *qry;
RangeTblEntry *oldrte;
RangeTblEntry *newrte;
1998-02-06 17:46:29 +01:00
qry = makeNode(Query);
qry->commandType = CMD_UTILITY;
qry->utilityStmt = (Node *) stmt;
/*
* To avoid deadlock, make sure the first thing we do is grab
2001-03-22 05:01:46 +01:00
* AccessExclusiveLock on the target relation. This will be needed by
* DefineQueryRewrite(), and we don't want to grab a lesser lock
* beforehand. We don't need to hold a refcount on the relcache
* entry, however.
*/
heap_close(heap_openrv(stmt->relation, AccessExclusiveLock),
NoLock);
/*
2001-03-22 05:01:46 +01:00
* NOTE: 'OLD' must always have a varno equal to 1 and 'NEW' equal to
* 2. Set up their RTEs in the main pstate for use in parsing the
* rule qualification.
*/
Assert(pstate->p_rtable == NIL);
oldrte = addRangeTableEntry(pstate, stmt->relation,
makeAlias("*OLD*", NIL),
false, true);
newrte = addRangeTableEntry(pstate, stmt->relation,
makeAlias("*NEW*", NIL),
false, true);
/* Must override addRangeTableEntry's default access-check flags */
oldrte->checkForRead = false;
newrte->checkForRead = false;
2001-03-22 05:01:46 +01:00
/*
* They must be in the namespace too for lookup purposes, but only add
* the one(s) that are relevant for the current kind of rule. In an
* UPDATE rule, quals must refer to OLD.field or NEW.field to be
2001-03-22 05:01:46 +01:00
* unambiguous, but there's no need to be so picky for INSERT &
* DELETE. (Note we marked the RTEs "inFromCl = true" above to allow
* unqualified references to their fields.) We do not add them to the
* joinlist.
*/
switch (stmt->event)
{
case CMD_SELECT:
addRTEtoQuery(pstate, oldrte, false, true);
break;
case CMD_UPDATE:
addRTEtoQuery(pstate, oldrte, false, true);
addRTEtoQuery(pstate, newrte, false, true);
break;
case CMD_INSERT:
addRTEtoQuery(pstate, newrte, false, true);
break;
case CMD_DELETE:
addRTEtoQuery(pstate, oldrte, false, true);
break;
default:
elog(ERROR, "transformRuleStmt: unexpected event type %d",
(int) stmt->event);
break;
}
/* take care of the where clause */
stmt->whereClause = transformWhereClause(pstate, stmt->whereClause);
2001-03-22 05:01:46 +01:00
if (length(pstate->p_rtable) != 2) /* naughty, naughty... */
elog(ERROR, "Rule WHERE condition may not contain references to other relations");
/* aggregates not allowed (but subselects are okay) */
if (contain_agg_clause(stmt->whereClause))
elog(ERROR, "Rule WHERE condition may not contain aggregate functions");
/* save info about sublinks in where clause */
qry->hasSubLinks = pstate->p_hasSubLinks;
/*
* 'instead nothing' rules with a qualification need a query
* rangetable so the rewrite handler can add the negated rule
* qualification to the original query. We create a query with the new
* command type CMD_NOTHING here that is treated specially by the
* rewrite system.
*/
if (stmt->actions == NIL)
{
Query *nothing_qry = makeNode(Query);
nothing_qry->commandType = CMD_NOTHING;
nothing_qry->rtable = pstate->p_rtable;
2001-03-22 05:01:46 +01:00
nothing_qry->jointree = makeFromExpr(NIL, NULL); /* no join wanted */
stmt->actions = makeList1(nothing_qry);
}
else
{
List *oldactions;
List *newactions = NIL;
/*
* transform each statement, like parse_analyze()
*/
foreach(oldactions, stmt->actions)
{
Node *action = (Node *) lfirst(oldactions);
ParseState *sub_pstate = make_parsestate(pstate->parentParseState);
Query *sub_qry,
*top_subqry;
bool has_old,
has_new;
/*
2001-03-22 05:01:46 +01:00
* Set up OLD/NEW in the rtable for this statement. The
* entries are marked not inFromCl because we don't want them
* to be referred to by unqualified field names nor "*" in the
* rule actions. We must add them to the namespace, however,
* or they won't be accessible at all. We decide later
* whether to put them in the joinlist.
*/
oldrte = addRangeTableEntry(sub_pstate, stmt->relation,
makeAlias("*OLD*", NIL),
false, false);
newrte = addRangeTableEntry(sub_pstate, stmt->relation,
makeAlias("*NEW*", NIL),
false, false);
oldrte->checkForRead = false;
newrte->checkForRead = false;
addRTEtoQuery(sub_pstate, oldrte, false, true);
addRTEtoQuery(sub_pstate, newrte, false, true);
/* Transform the rule action statement */
top_subqry = transformStmt(sub_pstate, action,
2002-09-04 22:31:48 +02:00
extras_before, extras_after);
/*
* We cannot support utility-statement actions (eg NOTIFY)
* with nonempty rule WHERE conditions, because there's no way
* to make the utility action execute conditionally.
*/
if (top_subqry->commandType == CMD_UTILITY &&
stmt->whereClause != NULL)
elog(ERROR, "Rules with WHERE conditions may only have SELECT, INSERT, UPDATE, or DELETE actions");
/*
* If the action is INSERT...SELECT, OLD/NEW have been pushed
* down into the SELECT, and that's what we need to look at.
2001-03-22 05:01:46 +01:00
* (Ugly kluge ... try to fix this when we redesign
* querytrees.)
*/
sub_qry = getInsertSelectQuery(top_subqry, NULL);
/*
* Validate action's use of OLD/NEW, qual too
*/
has_old =
rangeTableEntry_used((Node *) sub_qry, PRS2_OLD_VARNO, 0) ||
rangeTableEntry_used(stmt->whereClause, PRS2_OLD_VARNO, 0);
has_new =
rangeTableEntry_used((Node *) sub_qry, PRS2_NEW_VARNO, 0) ||
rangeTableEntry_used(stmt->whereClause, PRS2_NEW_VARNO, 0);
switch (stmt->event)
{
case CMD_SELECT:
if (has_old)
elog(ERROR, "ON SELECT rule may not use OLD");
if (has_new)
elog(ERROR, "ON SELECT rule may not use NEW");
break;
case CMD_UPDATE:
/* both are OK */
break;
case CMD_INSERT:
if (has_old)
elog(ERROR, "ON INSERT rule may not use OLD");
break;
case CMD_DELETE:
if (has_new)
elog(ERROR, "ON DELETE rule may not use NEW");
break;
default:
elog(ERROR, "transformRuleStmt: unexpected event type %d",
(int) stmt->event);
break;
}
/*
2001-03-22 05:01:46 +01:00
* For efficiency's sake, add OLD to the rule action's
* jointree only if it was actually referenced in the
* statement or qual.
*
* For INSERT, NEW is not really a relation (only a reference to
* the to-be-inserted tuple) and should never be added to the
* jointree.
*
* For UPDATE, we treat NEW as being another kind of reference to
2001-03-22 05:01:46 +01:00
* OLD, because it represents references to *transformed*
* tuples of the existing relation. It would be wrong to
* enter NEW separately in the jointree, since that would
* cause a double join of the updated relation. It's also
* wrong to fail to make a jointree entry if only NEW and not
* OLD is mentioned.
*/
if (has_old || (has_new && stmt->event == CMD_UPDATE))
{
/* hack so we can use addRTEtoQuery() */
sub_pstate->p_rtable = sub_qry->rtable;
sub_pstate->p_joinlist = sub_qry->jointree->fromlist;
addRTEtoQuery(sub_pstate, oldrte, true, false);
sub_qry->jointree->fromlist = sub_pstate->p_joinlist;
}
newactions = lappend(newactions, top_subqry);
release_pstate_resources(sub_pstate);
pfree(sub_pstate);
}
stmt->actions = newactions;
}
1998-02-06 17:46:29 +01:00
return qry;
}
/*
* transformSelectStmt -
* transforms a Select Statement
*
* Note: this is also used for DECLARE CURSOR statements.
*/
static Query *
1998-01-09 21:06:08 +01:00
transformSelectStmt(ParseState *pstate, SelectStmt *stmt)
{
Query *qry = makeNode(Query);
Node *qual;
qry->commandType = CMD_SELECT;
if (stmt->portalname)
{
/* DECLARE CURSOR */
if (stmt->into)
elog(ERROR, "DECLARE CURSOR must not specify INTO");
if (stmt->forUpdate)
elog(ERROR, "DECLARE/UPDATE is not supported"
"\n\tCursors must be READ ONLY");
qry->into = makeNode(RangeVar);
qry->into->relname = stmt->portalname;
qry->isPortal = TRUE;
2001-03-22 05:01:46 +01:00
qry->isBinary = stmt->binary; /* internal portal */
}
else
{
/* SELECT */
qry->into = stmt->into;
qry->isPortal = FALSE;
qry->isBinary = FALSE;
}
/* make FOR UPDATE clause available to addRangeTableEntry */
pstate->p_forUpdate = stmt->forUpdate;
/* process the FROM clause */
transformFromClause(pstate, stmt->fromClause);
/* transform targetlist */
qry->targetList = transformTargetList(pstate, stmt->targetList);
if (stmt->intoColNames)
applyColumnNames(qry->targetList, stmt->intoColNames);
/* transform WHERE */
qual = transformWhereClause(pstate, stmt->whereClause);
/*
* Initial processing of HAVING clause is just like WHERE clause.
* Additional work will be done in optimizer/plan/planner.c.
*/
qry->havingQual = transformWhereClause(pstate, stmt->havingClause);
qry->groupClause = transformGroupClause(pstate,
stmt->groupClause,
qry->targetList);
qry->sortClause = transformSortClause(pstate,
stmt->sortClause,
qry->targetList);
qry->distinctClause = transformDistinctClause(pstate,
stmt->distinctClause,
qry->targetList,
&qry->sortClause);
qry->limitOffset = stmt->limitOffset;
qry->limitCount = stmt->limitCount;
qry->rtable = pstate->p_rtable;
qry->jointree = makeFromExpr(pstate->p_joinlist, qual);
qry->hasSubLinks = pstate->p_hasSubLinks;
1998-01-20 06:05:08 +01:00
qry->hasAggs = pstate->p_hasAggs;
if (pstate->p_hasAggs || qry->groupClause || qry->havingQual)
parseCheckAggregates(pstate, qry);
if (stmt->forUpdate != NIL)
transformForUpdate(qry, stmt->forUpdate);
return qry;
}
/*
* transformSetOperationsStmt -
* transforms a set-operations tree
*
* A set-operation tree is just a SELECT, but with UNION/INTERSECT/EXCEPT
* structure to it. We must transform each leaf SELECT and build up a top-
* level Query that contains the leaf SELECTs as subqueries in its rangetable.
* The tree of set operations is converted into the setOperations field of
* the top-level Query.
*/
static Query *
transformSetOperationStmt(ParseState *pstate, SelectStmt *stmt)
{
Query *qry = makeNode(Query);
SelectStmt *leftmostSelect;
int leftmostRTI;
Query *leftmostQuery;
SetOperationStmt *sostmt;
RangeVar *into;
List *intoColNames;
char *portalname;
bool binary;
List *sortClause;
Node *limitOffset;
Node *limitCount;
List *forUpdate;
Node *node;
List *lefttl,
*dtlist,
*targetvars,
*targetnames,
*sv_namespace,
*sv_rtable;
RangeTblEntry *jrte;
RangeTblRef *jrtr;
int tllen;
qry->commandType = CMD_SELECT;
/*
2001-03-22 05:01:46 +01:00
* Find leftmost leaf SelectStmt; extract the one-time-only items from
* it and from the top-level node.
*/
leftmostSelect = stmt->larg;
while (leftmostSelect && leftmostSelect->op != SETOP_NONE)
leftmostSelect = leftmostSelect->larg;
Assert(leftmostSelect && IsA(leftmostSelect, SelectStmt) &&
leftmostSelect->larg == NULL);
into = leftmostSelect->into;
intoColNames = leftmostSelect->intoColNames;
portalname = stmt->portalname;
binary = stmt->binary;
/* clear them to prevent complaints in transformSetOperationTree() */
leftmostSelect->into = NULL;
leftmostSelect->intoColNames = NIL;
stmt->portalname = NULL;
stmt->binary = false;
/*
* These are not one-time, exactly, but we want to process them here
2001-03-22 05:01:46 +01:00
* and not let transformSetOperationTree() see them --- else it'll
* just recurse right back here!
*/
sortClause = stmt->sortClause;
limitOffset = stmt->limitOffset;
limitCount = stmt->limitCount;
forUpdate = stmt->forUpdate;
stmt->sortClause = NIL;
stmt->limitOffset = NULL;
stmt->limitCount = NULL;
stmt->forUpdate = NIL;
/* We don't support forUpdate with set ops at the moment. */
if (forUpdate)
elog(ERROR, "SELECT FOR UPDATE is not allowed with UNION/INTERSECT/EXCEPT");
Hi! INTERSECT and EXCEPT is available for postgresql-v6.4! The patch against v6.4 is included at the end of the current text (in uuencoded form!) I also included the text of my Master's Thesis. (a postscript version). I hope that you find something of it useful and would be happy if parts of it find their way into the PostgreSQL documentation project (If so, tell me, then I send the sources of the document!) The contents of the document are: -) The first chapter might be of less interest as it gives only an overview on SQL. -) The second chapter gives a description on much of PostgreSQL's features (like user defined types etc. and how to use these features) -) The third chapter starts with an overview of PostgreSQL's internal structure with focus on the stages a query has to pass (i.e. parser, planner/optimizer, executor). Then a detailed description of the implementation of the Having clause and the Intersect/Except logic is given. Originally I worked on v6.3.2 but never found time enough to prepare and post a patch. Now I applied the changes to v6.4 to get Intersect and Except working with the new version. Chapter 3 of my documentation deals with the changes against v6.3.2, so keep that in mind when comparing the parts of the code printed there with the patched sources of v6.4. Here are some remarks on the patch. There are some things that have still to be done but at the moment I don't have time to do them myself. (I'm doing my military service at the moment) Sorry for that :-( -) I used a rewrite technique for the implementation of the Except/Intersect logic which rewrites the query to a semantically equivalent query before it is handed to the rewrite system (for views, rules etc.), planner, executor etc. -) In v6.3.2 the types of the attributes of two select statements connected by the UNION keyword had to match 100%. In v6.4 the types only need to be familiar (i.e. int and float can be mixed). Since this feature did not exist when I worked on Intersect/Except it does not work correctly for Except/Intersect queries WHEN USED IN COMBINATION WITH UNIONS! (i.e. sometimes the wrong type is used for the resulting table. This is because until now the types of the attributes of the first select statement have been used for the resulting table. When Intersects and/or Excepts are used in combination with Unions it might happen, that the first select statement of the original query appears at another position in the query which will be executed. The reason for this is the technique used for the implementation of Except/Intersect which does a query rewrite!) NOTE: It is NOT broken for pure UNION queries and pure INTERSECT/EXCEPT queries!!! -) I had to add the field intersect_clause to some data structures but did not find time to implement printfuncs for the new field. This does NOT break the debug modes but when an Except/Intersect is used the query debug output will be the already rewritten query. -) Massive changes to the grammar rules for SELECT and INSERT statements have been necessary (see comments in gram.y and documentation for deatails) in order to be able to use mixed queries like (SELECT ... UNION (SELECT ... EXCEPT SELECT)) INTERSECT SELECT...; -) When using UNION/EXCEPT/INTERSECT you will get: NOTICE: equal: "Don't know if nodes of type xxx are equal". I did not have time to add comparsion support for all the needed nodes, but the default behaviour of the function equal met my requirements. I did not dare to supress this message! That's the reason why the regression test for union will fail: These messages are also included in the union.out file! -) Somebody of you changed the union_planner() function for v6.4 (I copied the targetlist to new_tlist and that was removed and replaced by a cleanup of the original targetlist). These chnages violated some having queries executed against views so I changed it back again. I did not have time to examine the differences between the two versions but now it works :-) If you want to find out, try the file queries/view_having.sql on both versions and compare the results . Two queries won't produce a correct result with your version. regards Stefan
1999-01-18 01:10:17 +01:00
1999-05-25 18:15:34 +02:00
/*
* Recursively transform the components of the tree.
1999-05-25 18:15:34 +02:00
*/
sostmt = (SetOperationStmt *) transformSetOperationTree(pstate, stmt);
Assert(sostmt && IsA(sostmt, SetOperationStmt));
qry->setOperations = (Node *) sostmt;
/*
* Re-find leftmost SELECT (now it's a sub-query in rangetable)
*/
node = sostmt->larg;
while (node && IsA(node, SetOperationStmt))
node = ((SetOperationStmt *) node)->larg;
Assert(node && IsA(node, RangeTblRef));
leftmostRTI = ((RangeTblRef *) node)->rtindex;
leftmostQuery = rt_fetch(leftmostRTI, pstate->p_rtable)->subquery;
Assert(leftmostQuery != NULL);
2001-03-22 05:01:46 +01:00
/*
* Generate dummy targetlist for outer query using column names of
2001-03-22 05:01:46 +01:00
* leftmost select and common datatypes of topmost set operation. Also
* make lists of the dummy vars and their names for use in parsing
* ORDER BY.
*/
qry->targetList = NIL;
targetvars = NIL;
targetnames = NIL;
lefttl = leftmostQuery->targetList;
foreach(dtlist, sostmt->colTypes)
{
Oid colType = lfirsto(dtlist);
2001-03-22 05:01:46 +01:00
Resdom *leftResdom = ((TargetEntry *) lfirst(lefttl))->resdom;
char *colName = pstrdup(leftResdom->resname);
Resdom *resdom;
Expr *expr;
resdom = makeResdom((AttrNumber) pstate->p_last_resno++,
colType,
-1,
colName,
false);
expr = (Expr *) makeVar(1,
leftResdom->resno,
colType,
-1,
0);
qry->targetList = lappend(qry->targetList,
makeTargetEntry(resdom, expr));
targetvars = lappend(targetvars, expr);
targetnames = lappend(targetnames, makeString(colName));
lefttl = lnext(lefttl);
}
2001-03-22 05:01:46 +01:00
/*
* Insert one-time items into top-level query
*
* This needs to agree with transformSelectStmt!
*/
if (portalname)
{
/* DECLARE CURSOR */
if (into)
elog(ERROR, "DECLARE CURSOR must not specify INTO");
if (forUpdate)
elog(ERROR, "DECLARE/UPDATE is not supported"
"\n\tCursors must be READ ONLY");
qry->into = makeNode(RangeVar);
qry->into->relname = portalname;
qry->isPortal = TRUE;
qry->isBinary = binary; /* internal portal */
}
else
{
/* SELECT */
qry->into = into;
qry->isPortal = FALSE;
qry->isBinary = FALSE;
}
/*
2002-09-04 22:31:48 +02:00
* Any column names from CREATE TABLE AS need to be attached to both
* the top level and the leftmost subquery. We do not do this earlier
* because we do *not* want the targetnames list to be affected.
*/
if (intoColNames)
{
applyColumnNames(qry->targetList, intoColNames);
applyColumnNames(leftmostQuery->targetList, intoColNames);
}
/*
2001-03-22 05:01:46 +01:00
* As a first step towards supporting sort clauses that are
* expressions using the output columns, generate a namespace entry
* that makes the output columns visible. A Join RTE node is handy
2001-03-22 05:01:46 +01:00
* for this, since we can easily control the Vars generated upon
* matches.
*
* Note: we don't yet do anything useful with such cases, but at least
* "ORDER BY upper(foo)" will draw the right error message rather than
* "foo not found".
*/
jrte = addRangeTableEntryForJoin(NULL,
targetnames,
JOIN_INNER,
targetvars,
NULL,
true);
jrtr = makeNode(RangeTblRef);
jrtr->rtindex = 1;
sv_rtable = pstate->p_rtable;
pstate->p_rtable = makeList1(jrte);
sv_namespace = pstate->p_namespace;
pstate->p_namespace = makeList1(jrtr);
/*
* For now, we don't support resjunk sort clauses on the output of a
* setOperation tree --- you can only use the SQL92-spec options of
* selecting an output column by name or number. Enforce by checking
* that transformSortClause doesn't add any items to tlist.
*/
tllen = length(qry->targetList);
qry->sortClause = transformSortClause(pstate,
sortClause,
qry->targetList);
pstate->p_namespace = sv_namespace;
pstate->p_rtable = sv_rtable;
if (tllen != length(qry->targetList))
elog(ERROR, "ORDER BY on a UNION/INTERSECT/EXCEPT result must be on one of the result columns");
qry->limitOffset = limitOffset;
qry->limitCount = limitCount;
qry->rtable = pstate->p_rtable;
qry->jointree = makeFromExpr(pstate->p_joinlist, NULL);
qry->hasSubLinks = pstate->p_hasSubLinks;
qry->hasAggs = pstate->p_hasAggs;
if (pstate->p_hasAggs || qry->groupClause || qry->havingQual)
parseCheckAggregates(pstate, qry);
if (forUpdate != NIL)
transformForUpdate(qry, forUpdate);
1999-01-21 17:08:55 +01:00
return qry;
}
/*
* transformSetOperationTree
* Recursively transform leaves and internal nodes of a set-op tree
*/
static Node *
transformSetOperationTree(ParseState *pstate, SelectStmt *stmt)
{
2001-03-22 05:01:46 +01:00
bool isLeaf;
Assert(stmt && IsA(stmt, SelectStmt));
/*
* Validity-check both leaf and internal SELECTs for disallowed ops.
*/
if (stmt->into)
elog(ERROR, "INTO is only allowed on first SELECT of UNION/INTERSECT/EXCEPT");
if (stmt->portalname) /* should not happen */
elog(ERROR, "Portal may not appear in UNION/INTERSECT/EXCEPT");
/* We don't support forUpdate with set ops at the moment. */
if (stmt->forUpdate)
elog(ERROR, "SELECT FOR UPDATE is not allowed with UNION/INTERSECT/EXCEPT");
/*
* If an internal node of a set-op tree has ORDER BY, UPDATE, or LIMIT
* clauses attached, we need to treat it like a leaf node to generate
* an independent sub-Query tree. Otherwise, it can be represented by
* a SetOperationStmt node underneath the parent Query.
*/
if (stmt->op == SETOP_NONE)
{
Assert(stmt->larg == NULL && stmt->rarg == NULL);
isLeaf = true;
}
else
{
Assert(stmt->larg != NULL && stmt->rarg != NULL);
if (stmt->sortClause || stmt->limitOffset || stmt->limitCount ||
stmt->forUpdate)
isLeaf = true;
else
isLeaf = false;
}
if (isLeaf)
{
/* Process leaf SELECT */
2001-03-22 05:01:46 +01:00
List *selectList;
Query *selectQuery;
char selectName[32];
RangeTblEntry *rte;
RangeTblRef *rtr;
/*
* Transform SelectStmt into a Query.
*
* Note: previously transformed sub-queries don't affect the parsing
2001-03-22 05:01:46 +01:00
* of this sub-query, because they are not in the toplevel
* pstate's namespace list.
*/
selectList = parse_analyze((Node *) stmt, pstate);
Assert(length(selectList) == 1);
selectQuery = (Query *) lfirst(selectList);
2001-03-22 05:01:46 +01:00
/*
* Make the leaf query be a subquery in the top-level rangetable.
*/
snprintf(selectName, sizeof(selectName), "*SELECT* %d", length(pstate->p_rtable) + 1);
rte = addRangeTableEntryForSubquery(pstate,
selectQuery,
makeAlias(selectName, NIL),
false);
2001-03-22 05:01:46 +01:00
/*
2001-03-22 05:01:46 +01:00
* Return a RangeTblRef to replace the SelectStmt in the set-op
* tree.
*/
rtr = makeNode(RangeTblRef);
/* assume new rte is at end */
rtr->rtindex = length(pstate->p_rtable);
Assert(rte == rt_fetch(rtr->rtindex, pstate->p_rtable));
return (Node *) rtr;
}
else
{
/* Process an internal node (set operation node) */
SetOperationStmt *op = makeNode(SetOperationStmt);
2001-03-22 05:01:46 +01:00
List *lcoltypes;
List *rcoltypes;
const char *context;
context = (stmt->op == SETOP_UNION ? "UNION" :
(stmt->op == SETOP_INTERSECT ? "INTERSECT" :
"EXCEPT"));
op->op = stmt->op;
op->all = stmt->all;
/*
* Recursively transform the child nodes.
*/
op->larg = transformSetOperationTree(pstate, stmt->larg);
op->rarg = transformSetOperationTree(pstate, stmt->rarg);
2001-03-22 05:01:46 +01:00
/*
* Verify that the two children have the same number of non-junk
* columns, and determine the types of the merged output columns.
*/
lcoltypes = getSetColTypes(pstate, op->larg);
rcoltypes = getSetColTypes(pstate, op->rarg);
if (length(lcoltypes) != length(rcoltypes))
elog(ERROR, "Each %s query must have the same number of columns",
context);
op->colTypes = NIL;
while (lcoltypes != NIL)
{
Oid lcoltype = lfirsto(lcoltypes);
Oid rcoltype = lfirsto(rcoltypes);
2001-03-22 05:01:46 +01:00
Oid rescoltype;
rescoltype = select_common_type(makeListo2(lcoltype, rcoltype),
context);
op->colTypes = lappendo(op->colTypes, rescoltype);
lcoltypes = lnext(lcoltypes);
rcoltypes = lnext(rcoltypes);
}
return (Node *) op;
}
}
/*
* getSetColTypes
* Get output column types of an (already transformed) set-op node
*/
static List *
getSetColTypes(ParseState *pstate, Node *node)
{
if (IsA(node, RangeTblRef))
{
RangeTblRef *rtr = (RangeTblRef *) node;
RangeTblEntry *rte = rt_fetch(rtr->rtindex, pstate->p_rtable);
2001-03-22 05:01:46 +01:00
Query *selectQuery = rte->subquery;
List *result = NIL;
List *tl;
Assert(selectQuery != NULL);
/* Get types of non-junk columns */
foreach(tl, selectQuery->targetList)
{
TargetEntry *tle = (TargetEntry *) lfirst(tl);
Resdom *resnode = tle->resdom;
if (resnode->resjunk)
continue;
result = lappendo(result, resnode->restype);
}
return result;
}
else if (IsA(node, SetOperationStmt))
{
SetOperationStmt *op = (SetOperationStmt *) node;
/* Result already computed during transformation of node */
Assert(op->colTypes != NIL);
return op->colTypes;
}
else
{
elog(ERROR, "getSetColTypes: unexpected node %d",
(int) nodeTag(node));
return NIL; /* keep compiler quiet */
}
}
/* Attach column names from a ColumnDef list to a TargetEntry list */
static void
applyColumnNames(List *dst, List *src)
{
if (length(src) > length(dst))
elog(ERROR, "CREATE TABLE AS specifies too many column names");
while (src != NIL && dst != NIL)
{
TargetEntry *d = (TargetEntry *) lfirst(dst);
ColumnDef *s = (ColumnDef *) lfirst(src);
Assert(d->resdom && !d->resdom->resjunk);
d->resdom->resname = pstrdup(s->colname);
dst = lnext(dst);
src = lnext(src);
}
}
/*
* transformUpdateStmt -
* transforms an update statement
*/
static Query *
1998-01-09 21:06:08 +01:00
transformUpdateStmt(ParseState *pstate, UpdateStmt *stmt)
{
Query *qry = makeNode(Query);
Node *qual;
List *origTargetList;
List *tl;
qry->commandType = CMD_UPDATE;
pstate->p_is_update = true;
qry->resultRelation = setTargetTable(pstate, stmt->relation,
2002-09-04 22:31:48 +02:00
interpretInhOption(stmt->relation->inhOpt),
true);
/*
* the FROM clause is non-standard SQL syntax. We used to be able to
* do this with REPLACE in POSTQUEL so we keep the feature.
*/
transformFromClause(pstate, stmt->fromClause);
qry->targetList = transformTargetList(pstate, stmt->targetList);
qual = transformWhereClause(pstate, stmt->whereClause);
qry->rtable = pstate->p_rtable;
qry->jointree = makeFromExpr(pstate->p_joinlist, qual);
1998-01-10 05:30:11 +01:00
qry->hasSubLinks = pstate->p_hasSubLinks;
1998-01-20 06:05:08 +01:00
qry->hasAggs = pstate->p_hasAggs;
if (pstate->p_hasAggs)
parseCheckAggregates(pstate, qry);
/*
* Now we are done with SELECT-like processing, and can get on with
* transforming the target list to match the UPDATE target columns.
*/
/* Prepare to assign non-conflicting resnos to resjunk attributes */
if (pstate->p_last_resno <= pstate->p_target_relation->rd_rel->relnatts)
pstate->p_last_resno = pstate->p_target_relation->rd_rel->relnatts + 1;
/* Prepare non-junk columns for assignment to target table */
origTargetList = stmt->targetList;
foreach(tl, qry->targetList)
{
TargetEntry *tle = (TargetEntry *) lfirst(tl);
Resdom *resnode = tle->resdom;
ResTarget *origTarget;
if (resnode->resjunk)
{
/*
* Resjunk nodes need no additional processing, but be sure
* they have names and resnos that do not match any target
* columns; else rewriter or planner might get confused.
*/
resnode->resname = "?resjunk?";
resnode->resno = (AttrNumber) pstate->p_last_resno++;
continue;
}
if (origTargetList == NIL)
elog(ERROR, "UPDATE target count mismatch --- internal error");
origTarget = (ResTarget *) lfirst(origTargetList);
updateTargetListEntry(pstate, tle, origTarget->name,
attnameAttNum(pstate->p_target_relation,
origTarget->name, true),
origTarget->indirection);
origTargetList = lnext(origTargetList);
}
if (origTargetList != NIL)
elog(ERROR, "UPDATE target count mismatch --- internal error");
1998-01-10 05:30:11 +01:00
return qry;
}
Hi! INTERSECT and EXCEPT is available for postgresql-v6.4! The patch against v6.4 is included at the end of the current text (in uuencoded form!) I also included the text of my Master's Thesis. (a postscript version). I hope that you find something of it useful and would be happy if parts of it find their way into the PostgreSQL documentation project (If so, tell me, then I send the sources of the document!) The contents of the document are: -) The first chapter might be of less interest as it gives only an overview on SQL. -) The second chapter gives a description on much of PostgreSQL's features (like user defined types etc. and how to use these features) -) The third chapter starts with an overview of PostgreSQL's internal structure with focus on the stages a query has to pass (i.e. parser, planner/optimizer, executor). Then a detailed description of the implementation of the Having clause and the Intersect/Except logic is given. Originally I worked on v6.3.2 but never found time enough to prepare and post a patch. Now I applied the changes to v6.4 to get Intersect and Except working with the new version. Chapter 3 of my documentation deals with the changes against v6.3.2, so keep that in mind when comparing the parts of the code printed there with the patched sources of v6.4. Here are some remarks on the patch. There are some things that have still to be done but at the moment I don't have time to do them myself. (I'm doing my military service at the moment) Sorry for that :-( -) I used a rewrite technique for the implementation of the Except/Intersect logic which rewrites the query to a semantically equivalent query before it is handed to the rewrite system (for views, rules etc.), planner, executor etc. -) In v6.3.2 the types of the attributes of two select statements connected by the UNION keyword had to match 100%. In v6.4 the types only need to be familiar (i.e. int and float can be mixed). Since this feature did not exist when I worked on Intersect/Except it does not work correctly for Except/Intersect queries WHEN USED IN COMBINATION WITH UNIONS! (i.e. sometimes the wrong type is used for the resulting table. This is because until now the types of the attributes of the first select statement have been used for the resulting table. When Intersects and/or Excepts are used in combination with Unions it might happen, that the first select statement of the original query appears at another position in the query which will be executed. The reason for this is the technique used for the implementation of Except/Intersect which does a query rewrite!) NOTE: It is NOT broken for pure UNION queries and pure INTERSECT/EXCEPT queries!!! -) I had to add the field intersect_clause to some data structures but did not find time to implement printfuncs for the new field. This does NOT break the debug modes but when an Except/Intersect is used the query debug output will be the already rewritten query. -) Massive changes to the grammar rules for SELECT and INSERT statements have been necessary (see comments in gram.y and documentation for deatails) in order to be able to use mixed queries like (SELECT ... UNION (SELECT ... EXCEPT SELECT)) INTERSECT SELECT...; -) When using UNION/EXCEPT/INTERSECT you will get: NOTICE: equal: "Don't know if nodes of type xxx are equal". I did not have time to add comparsion support for all the needed nodes, but the default behaviour of the function equal met my requirements. I did not dare to supress this message! That's the reason why the regression test for union will fail: These messages are also included in the union.out file! -) Somebody of you changed the union_planner() function for v6.4 (I copied the targetlist to new_tlist and that was removed and replaced by a cleanup of the original targetlist). These chnages violated some having queries executed against views so I changed it back again. I did not have time to examine the differences between the two versions but now it works :-) If you want to find out, try the file queries/view_having.sql on both versions and compare the results . Two queries won't produce a correct result with your version. regards Stefan
1999-01-18 01:10:17 +01:00
/*
* tranformAlterTableStmt -
* transform an Alter Table Statement
*/
static Query *
transformAlterTableStmt(ParseState *pstate, AlterTableStmt *stmt,
List **extras_before, List **extras_after)
{
Relation rel;
CreateStmtContext cxt;
Query *qry;
/*
* The only subtypes that currently require parse transformation
* handling are 'A'dd column and Add 'C'onstraint. These largely
* re-use code from CREATE TABLE.
*
* If we need to do any parse transformation, get exclusive lock on
* the relation to make sure it won't change before we execute the
* command.
*/
switch (stmt->subtype)
{
case 'A':
rel = heap_openrv(stmt->relation, AccessExclusiveLock);
cxt.stmtType = "ALTER TABLE";
cxt.relation = stmt->relation;
cxt.inhRelations = NIL;
cxt.relOid = RelationGetRelid(rel);
cxt.hasoids = SearchSysCacheExists(ATTNUM,
ObjectIdGetDatum(cxt.relOid),
Int16GetDatum(ObjectIdAttributeNumber),
0, 0);
cxt.columns = NIL;
cxt.ckconstraints = NIL;
cxt.fkconstraints = NIL;
cxt.ixconstraints = NIL;
cxt.blist = NIL;
cxt.alist = NIL;
cxt.pkey = NULL;
Assert(IsA(stmt->def, ColumnDef));
transformColumnDefinition(pstate, &cxt,
(ColumnDef *) stmt->def);
transformIndexConstraints(pstate, &cxt);
transformFKConstraints(pstate, &cxt);
((ColumnDef *) stmt->def)->constraints = cxt.ckconstraints;
*extras_before = nconc(*extras_before, cxt.blist);
*extras_after = nconc(cxt.alist, *extras_after);
heap_close(rel, NoLock); /* close rel, keep lock */
break;
case 'C':
rel = heap_openrv(stmt->relation, AccessExclusiveLock);
cxt.stmtType = "ALTER TABLE";
cxt.relation = stmt->relation;
cxt.inhRelations = NIL;
cxt.relOid = RelationGetRelid(rel);
cxt.hasoids = SearchSysCacheExists(ATTNUM,
ObjectIdGetDatum(cxt.relOid),
Int16GetDatum(ObjectIdAttributeNumber),
0, 0);
cxt.columns = NIL;
cxt.ckconstraints = NIL;
cxt.fkconstraints = NIL;
cxt.ixconstraints = NIL;
cxt.blist = NIL;
cxt.alist = NIL;
cxt.pkey = NULL;
if (IsA(stmt->def, Constraint))
transformTableConstraint(pstate, &cxt,
(Constraint *) stmt->def);
else if (IsA(stmt->def, FkConstraint))
cxt.fkconstraints = lappend(cxt.fkconstraints, stmt->def);
else
elog(ERROR, "Unexpected node type in ALTER TABLE ADD CONSTRAINT");
transformIndexConstraints(pstate, &cxt);
transformFKConstraints(pstate, &cxt);
Assert(cxt.columns == NIL);
stmt->def = (Node *) nconc(cxt.ckconstraints, cxt.fkconstraints);
*extras_before = nconc(*extras_before, cxt.blist);
*extras_after = nconc(cxt.alist, *extras_after);
heap_close(rel, NoLock); /* close rel, keep lock */
break;
case 'c':
2002-09-04 22:31:48 +02:00
/*
* Already-transformed ADD CONSTRAINT, so just make it look
* like the standard case.
*/
stmt->subtype = 'C';
break;
default:
break;
}
qry = makeNode(Query);
qry->commandType = CMD_UTILITY;
qry->utilityStmt = (Node *) stmt;
return qry;
}
static Query *
transformPrepareStmt(ParseState *pstate, PrepareStmt *stmt)
{
2002-09-04 22:31:48 +02:00
Query *result = makeNode(Query);
List *extras_before = NIL,
*extras_after = NIL;
List *argtype_oids = NIL; /* argtype OIDs in a list */
Oid *argtoids = NULL; /* as an array for parser_param_set */
int nargs;
result->commandType = CMD_UTILITY;
result->utilityStmt = (Node *) stmt;
/* Transform list of TypeNames to list (and array) of type OIDs */
nargs = length(stmt->argtypes);
if (nargs)
{
2002-09-04 22:31:48 +02:00
List *l;
int i = 0;
argtoids = (Oid *) palloc(nargs * sizeof(Oid));
2002-09-04 22:31:48 +02:00
foreach(l, stmt->argtypes)
{
2002-09-04 22:31:48 +02:00
TypeName *tn = lfirst(l);
Oid toid = typenameTypeId(tn);
argtype_oids = lappendo(argtype_oids, toid);
argtoids[i++] = toid;
}
}
stmt->argtype_oids = argtype_oids;
/*
2002-09-04 22:31:48 +02:00
* We need to adjust the parameters expected by the rest of the
* system, so that $1, ... $n are parsed properly.
*
2002-09-04 22:31:48 +02:00
* This is somewhat of a hack; however, the main parser interface only
* allows parameters to be specified when working with a raw query
* string, which is not helpful here.
*/
parser_param_set(argtoids, nargs);
stmt->query = transformStmt(pstate, (Node *) stmt->query,
&extras_before, &extras_after);
/* Shouldn't get any extras, since grammar only allows OptimizableStmt */
if (extras_before || extras_after)
elog(ERROR, "transformPrepareStmt: internal error");
/* Remove links to our local parameters */
parser_param_set(NULL, 0);
return result;
}
static Query *
transformExecuteStmt(ParseState *pstate, ExecuteStmt *stmt)
{
2002-09-04 22:31:48 +02:00
Query *result = makeNode(Query);
List *paramtypes;
result->commandType = CMD_UTILITY;
result->utilityStmt = (Node *) stmt;
paramtypes = FetchQueryParams(stmt->name);
if (stmt->params || paramtypes)
{
2002-09-04 22:31:48 +02:00
int nparams = length(stmt->params);
int nexpected = length(paramtypes);
List *l;
int i = 1;
if (nparams != nexpected)
elog(ERROR, "Wrong number of parameters, expected %d but got %d",
nexpected, nparams);
2002-09-04 22:31:48 +02:00
foreach(l, stmt->params)
{
2002-09-04 22:31:48 +02:00
Node *expr = lfirst(l);
Oid expected_type_id,
given_type_id;
expr = transformExpr(pstate, expr);
/* Cannot contain subselects or aggregates */
if (contain_subplans(expr))
elog(ERROR, "Cannot use subselects in EXECUTE parameters");
if (contain_agg_clause(expr))
elog(ERROR, "Cannot use aggregates in EXECUTE parameters");
given_type_id = exprType(expr);
expected_type_id = lfirsto(paramtypes);
expr = coerce_to_target_type(expr, given_type_id,
expected_type_id, -1,
COERCION_ASSIGNMENT,
COERCE_IMPLICIT_CAST);
if (expr == NULL)
elog(ERROR, "Parameter $%d of type %s cannot be coerced into the expected type %s"
"\n\tYou will need to rewrite or cast the expression",
i,
format_type_be(given_type_id),
format_type_be(expected_type_id));
lfirst(l) = expr;
paramtypes = lnext(paramtypes);
i++;
}
}
return result;
}
/* exported so planner can check again after rewriting, query pullup, etc */
1999-01-25 13:01:19 +01:00
void
CheckSelectForUpdate(Query *qry)
{
if (qry->setOperations)
elog(ERROR, "SELECT FOR UPDATE is not allowed with UNION/INTERSECT/EXCEPT");
if (qry->distinctClause != NIL)
1999-01-25 13:01:19 +01:00
elog(ERROR, "SELECT FOR UPDATE is not allowed with DISTINCT clause");
if (qry->groupClause != NIL)
1999-01-25 13:01:19 +01:00
elog(ERROR, "SELECT FOR UPDATE is not allowed with GROUP BY clause");
if (qry->hasAggs)
elog(ERROR, "SELECT FOR UPDATE is not allowed with AGGREGATE");
}
1999-01-21 17:08:55 +01:00
static void
transformForUpdate(Query *qry, List *forUpdate)
{
List *rowMarks = qry->rowMarks;
1999-01-21 17:08:55 +01:00
List *l;
List *rt;
1999-01-21 17:08:55 +01:00
Index i;
1999-01-25 13:01:19 +01:00
CheckSelectForUpdate(qry);
if (lfirst(forUpdate) == NULL)
1999-01-21 17:08:55 +01:00
{
/* all tables used in query */
i = 0;
foreach(rt, qry->rtable)
1999-01-21 17:08:55 +01:00
{
RangeTblEntry *rte = (RangeTblEntry *) lfirst(rt);
++i;
if (rte->rtekind == RTE_SUBQUERY)
{
/* FOR UPDATE of subquery is propagated to subquery's rels */
transformForUpdate(rte->subquery, makeList1(NULL));
}
else
{
2001-03-22 05:01:46 +01:00
if (!intMember(i, rowMarks)) /* avoid duplicates */
rowMarks = lappendi(rowMarks, i);
rte->checkForWrite = true;
}
1999-01-21 17:08:55 +01:00
}
}
else
1999-01-21 17:08:55 +01:00
{
/* just the named tables */
foreach(l, forUpdate)
1999-01-21 17:08:55 +01:00
{
char *relname = strVal(lfirst(l));
i = 0;
foreach(rt, qry->rtable)
1999-01-21 17:08:55 +01:00
{
RangeTblEntry *rte = (RangeTblEntry *) lfirst(rt);
2000-01-16 09:21:59 +01:00
++i;
if (strcmp(rte->eref->aliasname, relname) == 0)
1999-01-21 17:08:55 +01:00
{
if (rte->rtekind == RTE_SUBQUERY)
{
/* propagate to subquery */
transformForUpdate(rte->subquery, makeList1(NULL));
}
else
{
2001-03-22 05:01:46 +01:00
if (!intMember(i, rowMarks)) /* avoid duplicates */
rowMarks = lappendi(rowMarks, i);
rte->checkForWrite = true;
}
break;
1999-01-21 17:08:55 +01:00
}
}
if (rt == NIL)
elog(ERROR, "FOR UPDATE: relation \"%s\" not found in FROM clause",
relname);
1999-01-21 17:08:55 +01:00
}
}
qry->rowMarks = rowMarks;
1999-01-21 17:08:55 +01:00
}
/*
* relationHasPrimaryKey -
*
* See whether an existing relation has a primary key.
*/
static bool
relationHasPrimaryKey(Oid relationOid)
{
bool result = false;
Relation rel;
List *indexoidlist,
*indexoidscan;
rel = heap_open(relationOid, AccessShareLock);
/*
* Get the list of index OIDs for the table from the relcache, and
* look up each one in the pg_index syscache until we find one marked
* primary key (hopefully there isn't more than one such).
*/
indexoidlist = RelationGetIndexList(rel);
foreach(indexoidscan, indexoidlist)
{
Oid indexoid = lfirsto(indexoidscan);
HeapTuple indexTuple;
indexTuple = SearchSysCache(INDEXRELID,
ObjectIdGetDatum(indexoid),
0, 0, 0);
if (!HeapTupleIsValid(indexTuple))
elog(ERROR, "relationHasPrimaryKey: index %u not found",
indexoid);
result = ((Form_pg_index) GETSTRUCT(indexTuple))->indisprimary;
ReleaseSysCache(indexTuple);
if (result)
break;
}
freeList(indexoidlist);
heap_close(rel, AccessShareLock);
return result;
}
/*
* Preprocess a list of column constraint clauses
* to attach constraint attributes to their primary constraint nodes
* and detect inconsistent/misplaced constraint attributes.
*
* NOTE: currently, attributes are only supported for FOREIGN KEY primary
* constraints, but someday they ought to be supported for other constraints.
*/
static void
transformConstraintAttrs(List *constraintList)
{
Node *lastprimarynode = NULL;
bool saw_deferrability = false;
bool saw_initially = false;
List *clist;
foreach(clist, constraintList)
{
Node *node = lfirst(clist);
if (!IsA(node, Constraint))
{
lastprimarynode = node;
/* reset flags for new primary node */
saw_deferrability = false;
saw_initially = false;
}
else
{
Constraint *con = (Constraint *) node;
switch (con->contype)
{
case CONSTR_ATTR_DEFERRABLE:
if (lastprimarynode == NULL ||
!IsA(lastprimarynode, FkConstraint))
elog(ERROR, "Misplaced DEFERRABLE clause");
if (saw_deferrability)
elog(ERROR, "Multiple DEFERRABLE/NOT DEFERRABLE clauses not allowed");
saw_deferrability = true;
((FkConstraint *) lastprimarynode)->deferrable = true;
break;
case CONSTR_ATTR_NOT_DEFERRABLE:
if (lastprimarynode == NULL ||
!IsA(lastprimarynode, FkConstraint))
elog(ERROR, "Misplaced NOT DEFERRABLE clause");
if (saw_deferrability)
elog(ERROR, "Multiple DEFERRABLE/NOT DEFERRABLE clauses not allowed");
saw_deferrability = true;
((FkConstraint *) lastprimarynode)->deferrable = false;
if (saw_initially &&
((FkConstraint *) lastprimarynode)->initdeferred)
elog(ERROR, "INITIALLY DEFERRED constraint must be DEFERRABLE");
break;
case CONSTR_ATTR_DEFERRED:
if (lastprimarynode == NULL ||
!IsA(lastprimarynode, FkConstraint))
elog(ERROR, "Misplaced INITIALLY DEFERRED clause");
if (saw_initially)
elog(ERROR, "Multiple INITIALLY IMMEDIATE/DEFERRED clauses not allowed");
saw_initially = true;
((FkConstraint *) lastprimarynode)->initdeferred = true;
/*
* If only INITIALLY DEFERRED appears, assume
* DEFERRABLE
*/
if (!saw_deferrability)
((FkConstraint *) lastprimarynode)->deferrable = true;
else if (!((FkConstraint *) lastprimarynode)->deferrable)
elog(ERROR, "INITIALLY DEFERRED constraint must be DEFERRABLE");
break;
case CONSTR_ATTR_IMMEDIATE:
if (lastprimarynode == NULL ||
!IsA(lastprimarynode, FkConstraint))
elog(ERROR, "Misplaced INITIALLY IMMEDIATE clause");
if (saw_initially)
elog(ERROR, "Multiple INITIALLY IMMEDIATE/DEFERRED clauses not allowed");
saw_initially = true;
((FkConstraint *) lastprimarynode)->initdeferred = false;
break;
default:
/* Otherwise it's not an attribute */
lastprimarynode = node;
/* reset flags for new primary node */
saw_deferrability = false;
saw_initially = false;
break;
}
}
}
}
/* Build a FromExpr node */
static FromExpr *
makeFromExpr(List *fromlist, Node *quals)
{
2001-03-22 05:01:46 +01:00
FromExpr *f = makeNode(FromExpr);
f->fromlist = fromlist;
f->quals = quals;
return f;
}
/*
* Special handling of type definition for a column
*/
static void
transformColumnType(ParseState *pstate, ColumnDef *column)
{
TypeName *typename = column->typename;
Type ctype = typenameType(typename);
/*
2001-03-22 05:01:46 +01:00
* Is this the name of a complex type? If so, implement it as a set.
*
* XXX this is a hangover from ancient Berkeley code that probably
* doesn't work anymore anyway.
*/
2001-03-22 05:01:46 +01:00
if (typeTypeRelid(ctype) != InvalidOid)
{
/*
* (Eventually add in here that the set can only contain one
* element.)
*/
typename->setof = true;
}
ReleaseSysCache(ctype);
}
/*
* analyzeCreateSchemaStmt -
* analyzes the "create schema" statement
*
* Split the schema element list into individual commands and place
* them in the result list in an order such that there are no
* forward references (e.g. GRANT to a table created later in the list).
*
* SQL92 also allows constraints to make forward references, so thumb through
* the table columns and move forward references to a posterior alter-table
* command.
*
* The result is a list of parse nodes that still need to be analyzed ---
* but we can't analyze the later commands until we've executed the earlier
* ones, because of possible inter-object references.
*
* Note: Called from commands/command.c
*/
List *
analyzeCreateSchemaStmt(CreateSchemaStmt *stmt)
{
CreateSchemaStmtContext cxt;
List *result;
List *elements;
cxt.stmtType = "CREATE SCHEMA";
cxt.schemaname = stmt->schemaname;
cxt.authid = stmt->authid;
cxt.tables = NIL;
cxt.views = NIL;
cxt.grants = NIL;
cxt.fwconstraints = NIL;
cxt.alters = NIL;
cxt.blist = NIL;
cxt.alist = NIL;
/*
* Run through each schema element in the schema element list.
* Separate statements by type, and do preliminary analysis.
*/
foreach(elements, stmt->schemaElts)
{
Node *element = lfirst(elements);
switch (nodeTag(element))
{
case T_CreateStmt:
{
CreateStmt *elp = (CreateStmt *) element;
if (elp->relation->schemaname == NULL)
elp->relation->schemaname = cxt.schemaname;
2002-04-12 21:11:49 +02:00
else if (strcmp(cxt.schemaname, elp->relation->schemaname) != 0)
elog(ERROR, "New table specifies a schema (%s)"
2002-09-04 22:31:48 +02:00
" different from the one being created (%s)",
elp->relation->schemaname, cxt.schemaname);
/*
* XXX todo: deal with constraints
*/
cxt.tables = lappend(cxt.tables, element);
}
break;
case T_ViewStmt:
{
2002-09-04 22:31:48 +02:00
ViewStmt *elp = (ViewStmt *) element;
if (elp->view->schemaname == NULL)
elp->view->schemaname = cxt.schemaname;
2002-04-12 21:11:49 +02:00
else if (strcmp(cxt.schemaname, elp->view->schemaname) != 0)
elog(ERROR, "New view specifies a schema (%s)"
2002-09-04 22:31:48 +02:00
" different from the one being created (%s)",
elp->view->schemaname, cxt.schemaname);
/*
* XXX todo: deal with references between views
*/
cxt.views = lappend(cxt.views, element);
}
break;
case T_GrantStmt:
cxt.grants = lappend(cxt.grants, element);
break;
default:
elog(ERROR, "parser: unsupported schema node (internal error)");
}
}
result = NIL;
result = nconc(result, cxt.tables);
result = nconc(result, cxt.views);
result = nconc(result, cxt.grants);
return result;
}