Avoid recursion when processing simple lists of AND'ed or OR'ed clauses.

tglsfdc · tglsfdc · commit 2146f13408cd · 2014-06-16T15:55:30.000-04:00
Since most of the system thinks AND and OR are N-argument expressions
anyway, let's have the grammar generate a representation of that form when
dealing with input like "x AND y AND z AND ...", rather than generating
a deeply-nested binary tree that just has to be flattened later by the
planner.  This avoids stack overflow in parse analysis when dealing with
queries having more than a few thousand such clauses; and in any case it
removes some rather unsightly inconsistencies, since some parts of parse
analysis were generating N-argument ANDs/ORs already.

It's still possible to get a stack overflow with weirdly parenthesized
input, such as "x AND (y AND (z AND ( ... )))", but such cases are not
mainstream usage.  The maximum depth of parenthesization is already
limited by Bison's stack in such cases, anyway, so that the limit is
probably fairly platform-independent.

Patch originally by Gurjeet Singh, heavily revised by me
diff --git a/contrib/postgres_fdw/deparse.c b/contrib/postgres_fdw/deparse.c
@@ -1716,9 +1716,6 @@ deparseRelabelType(RelabelType *node, deparse_expr_cxt *context)
 
 /*
  * Deparse a BoolExpr node.
- *
- * Note: by the time we get here, AND and OR expressions have been flattened
- * into N-argument form, so we'd better be prepared to deal with that.
  */
 static void
 deparseBoolExpr(BoolExpr *node, deparse_expr_cxt *context)
diff --git a/src/backend/nodes/nodeFuncs.c b/src/backend/nodes/nodeFuncs.c
@@ -3047,6 +3047,14 @@ raw_expression_tree_walker(Node *node,
 				/* operator name is deemed uninteresting */
 			}
 			break;
+		case T_BoolExpr:
+			{
+				BoolExpr   *expr = (BoolExpr *) node;
+
+				if (walker(expr->args, context))
+					return true;
+			}
+			break;
 		case T_ColumnRef:
 			/* we assume the fields contain nothing interesting */
 			break;
diff --git a/src/backend/nodes/outfuncs.c b/src/backend/nodes/outfuncs.c
@@ -2437,15 +2437,6 @@ _outAExpr(StringInfo str, const A_Expr *node)
 			appendStringInfoChar(str, ' ');
 			WRITE_NODE_FIELD(name);
 			break;
-		case AEXPR_AND:
-			appendStringInfoString(str, " AND");
-			break;
-		case AEXPR_OR:
-			appendStringInfoString(str, " OR");
-			break;
-		case AEXPR_NOT:
-			appendStringInfoString(str, " NOT");
-			break;
 		case AEXPR_OP_ANY:
 			appendStringInfoChar(str, ' ');
 			WRITE_NODE_FIELD(name);
diff --git a/src/backend/optimizer/prep/prepjointree.c b/src/backend/optimizer/prep/prepjointree.c
@@ -133,9 +133,9 @@ static Node *find_jointree_node_for_rel(Node *jtnode, int relid);
  * transformations if any are found.
  *
  * This routine has to run before preprocess_expression(), so the quals
- * clauses are not yet reduced to implicit-AND format.  That means we need
- * to recursively search through explicit AND clauses, which are
- * probably only binary ANDs.  We stop as soon as we hit a non-AND item.
+ * clauses are not yet reduced to implicit-AND format, and are not guaranteed
+ * to be AND/OR-flat either.  That means we need to recursively search through
+ * explicit AND clauses.  We stop as soon as we hit a non-AND item.
  */
 void
 pull_up_sublinks(PlannerInfo *root)
diff --git a/src/backend/optimizer/prep/prepqual.c b/src/backend/optimizer/prep/prepqual.c
@@ -4,13 +4,12 @@
  *	  Routines for preprocessing qualification expressions
  *
  *
- * The parser regards AND and OR as purely binary operators, so a qual like
- *		(A = 1) OR (A = 2) OR (A = 3) ...
- * will produce a nested parsetree
- *		(OR (A = 1) (OR (A = 2) (OR (A = 3) ...)))
- * In reality, the optimizer and executor regard AND and OR as N-argument
- * operators, so this tree can be flattened to
- *		(OR (A = 1) (A = 2) (A = 3) ...)
+ * While the parser will produce flattened (N-argument) AND/OR trees from
+ * simple sequences of AND'ed or OR'ed clauses, there might be an AND clause
+ * directly underneath another AND, or OR underneath OR, if the input was
+ * oddly parenthesized.  Also, rule expansion and subquery flattening could
+ * produce such parsetrees.  The planner wants to flatten all such cases
+ * to ensure consistent optimization behavior.
  *
  * Formerly, this module was responsible for doing the initial flattening,
  * but now we leave it to eval_const_expressions to do that since it has to
diff --git a/src/backend/optimizer/util/clauses.c b/src/backend/optimizer/util/clauses.c
@@ -3447,12 +3447,15 @@ simplify_or_arguments(List *args,
 	List	   *unprocessed_args;
 
 	/*
-	 * Since the parser considers OR to be a binary operator, long OR lists
-	 * become deeply nested expressions.  We must flatten these into long
-	 * argument lists of a single OR operator.  To avoid blowing out the stack
-	 * with recursion of eval_const_expressions, we resort to some tenseness
-	 * here: we keep a list of not-yet-processed inputs, and handle flattening
-	 * of nested ORs by prepending to the to-do list instead of recursing.
+	 * We want to ensure that any OR immediately beneath another OR gets
+	 * flattened into a single OR-list, so as to simplify later reasoning.
+	 *
+	 * To avoid stack overflow from recursion of eval_const_expressions, we
+	 * resort to some tenseness here: we keep a list of not-yet-processed
+	 * inputs, and handle flattening of nested ORs by prepending to the to-do
+	 * list instead of recursing.  Now that the parser generates N-argument
+	 * ORs from simple lists, this complexity is probably less necessary than
+	 * it once was, but we might as well keep the logic.
 	 */
 	unprocessed_args = list_copy(args);
 	while (unprocessed_args)
diff --git a/src/backend/parser/gram.y b/src/backend/parser/gram.y
@@ -151,6 +151,9 @@ static void insertSelectOptions(SelectStmt *stmt,
 static Node *makeSetOp(SetOperation op, bool all, Node *larg, Node *rarg);
 static Node *doNegate(Node *n, int location);
 static void doNegateFloat(Value *v);
+static Node *makeAndExpr(Node *lexpr, Node *rexpr, int location);
+static Node *makeOrExpr(Node *lexpr, Node *rexpr, int location);
+static Node *makeNotExpr(Node *expr, int location);
 static Node *makeAArrayExpr(List *elements, int location);
 static Node *makeXmlExpr(XmlExprOp op, char *name, List *named_args,
 						 List *args, int location);
@@ -10849,11 +10852,11 @@ a_expr:		c_expr									{ $$ = $1; }
 				{ $$ = (Node *) makeA_Expr(AEXPR_OP, $2, $1, NULL, @2); }
 
 			| a_expr AND a_expr
-				{ $$ = (Node *) makeA_Expr(AEXPR_AND, NIL, $1, $3, @2); }
+				{ $$ = makeAndExpr($1, $3, @2); }
 			| a_expr OR a_expr
-				{ $$ = (Node *) makeA_Expr(AEXPR_OR, NIL, $1, $3, @2); }
+				{ $$ = makeOrExpr($1, $3, @2); }
 			| NOT a_expr
-				{ $$ = (Node *) makeA_Expr(AEXPR_NOT, NIL, NULL, $2, @1); }
+				{ $$ = makeNotExpr($2, @1); }
 
 			| a_expr LIKE a_expr
 				{ $$ = (Node *) makeSimpleA_Expr(AEXPR_OP, "~~", $1, $3, @2); }
@@ -11022,11 +11025,9 @@ a_expr:		c_expr									{ $$ = $1; }
 				}
 			| a_expr IS NOT DISTINCT FROM a_expr		%prec IS
 				{
-					$$ = (Node *) makeA_Expr(AEXPR_NOT, NIL, NULL,
-									(Node *) makeSimpleA_Expr(AEXPR_DISTINCT,
-															  "=", $1, $6, @2),
-											 @2);
-
+					$$ = makeNotExpr((Node *) makeSimpleA_Expr(AEXPR_DISTINCT,
+															   "=", $1, $6, @2),
+									 @2);
 				}
 			| a_expr IS OF '(' type_list ')'			%prec IS
 				{
@@ -11044,43 +11045,43 @@ a_expr:		c_expr									{ $$ = $1; }
 			 */
 			| a_expr BETWEEN opt_asymmetric b_expr AND b_expr		%prec BETWEEN
 				{
-					$$ = (Node *) makeA_Expr(AEXPR_AND, NIL,
+					$$ = makeAndExpr(
 						(Node *) makeSimpleA_Expr(AEXPR_OP, ">=", $1, $4, @2),
 						(Node *) makeSimpleA_Expr(AEXPR_OP, "<=", $1, $6, @2),
-											 @2);
+									 @2);
 				}
 			| a_expr NOT BETWEEN opt_asymmetric b_expr AND b_expr	%prec BETWEEN
 				{
-					$$ = (Node *) makeA_Expr(AEXPR_OR, NIL,
+					$$ = makeOrExpr(
 						(Node *) makeSimpleA_Expr(AEXPR_OP, "<", $1, $5, @2),
 						(Node *) makeSimpleA_Expr(AEXPR_OP, ">", $1, $7, @2),
-											 @2);
+									@2);
 				}
 			| a_expr BETWEEN SYMMETRIC b_expr AND b_expr			%prec BETWEEN
 				{
-					$$ = (Node *) makeA_Expr(AEXPR_OR, NIL,
-						(Node *) makeA_Expr(AEXPR_AND, NIL,
+					$$ = makeOrExpr(
+						  makeAndExpr(
 							(Node *) makeSimpleA_Expr(AEXPR_OP, ">=", $1, $4, @2),
 							(Node *) makeSimpleA_Expr(AEXPR_OP, "<=", $1, $6, @2),
-											@2),
-						(Node *) makeA_Expr(AEXPR_AND, NIL,
+									  @2),
+						  makeAndExpr(
 							(Node *) makeSimpleA_Expr(AEXPR_OP, ">=", $1, $6, @2),
 							(Node *) makeSimpleA_Expr(AEXPR_OP, "<=", $1, $4, @2),
-											@2),
-											 @2);
+									  @2),
+									@2);
 				}
 			| a_expr NOT BETWEEN SYMMETRIC b_expr AND b_expr		%prec BETWEEN
 				{
-					$$ = (Node *) makeA_Expr(AEXPR_AND, NIL,
-						(Node *) makeA_Expr(AEXPR_OR, NIL,
+					$$ = makeAndExpr(
+						   makeOrExpr(
 							(Node *) makeSimpleA_Expr(AEXPR_OP, "<", $1, $5, @2),
 							(Node *) makeSimpleA_Expr(AEXPR_OP, ">", $1, $7, @2),
-											@2),
-						(Node *) makeA_Expr(AEXPR_OR, NIL,
+									  @2),
+						   makeOrExpr(
 							(Node *) makeSimpleA_Expr(AEXPR_OP, "<", $1, $7, @2),
 							(Node *) makeSimpleA_Expr(AEXPR_OP, ">", $1, $5, @2),
-											@2),
-											 @2);
+									  @2),
+									 @2);
 				}
 			| a_expr IN_P in_expr
 				{
@@ -11114,7 +11115,7 @@ a_expr:		c_expr									{ $$ = $1; }
 						n->operName = list_make1(makeString("="));
 						n->location = @3;
 						/* Stick a NOT on top */
-						$$ = (Node *) makeA_Expr(AEXPR_NOT, NIL, NULL, (Node *) n, @2);
+						$$ = makeNotExpr((Node *) n, @2);
 					}
 					else
 					{
@@ -11162,10 +11163,9 @@ a_expr:		c_expr									{ $$ = $1; }
 				}
 			| a_expr IS NOT DOCUMENT_P				%prec IS
 				{
-					$$ = (Node *) makeA_Expr(AEXPR_NOT, NIL, NULL,
-											 makeXmlExpr(IS_DOCUMENT, NULL, NIL,
-														 list_make1($1), @2),
-											 @2);
+					$$ = makeNotExpr(makeXmlExpr(IS_DOCUMENT, NULL, NIL,
+												 list_make1($1), @2),
+									 @2);
 				}
 		;
 
@@ -11216,8 +11216,9 @@ b_expr:		c_expr
 				}
 			| b_expr IS NOT DISTINCT FROM b_expr	%prec IS
 				{
-					$$ = (Node *) makeA_Expr(AEXPR_NOT, NIL,
-						NULL, (Node *) makeSimpleA_Expr(AEXPR_DISTINCT, "=", $1, $6, @2), @2);
+					$$ = makeNotExpr((Node *) makeSimpleA_Expr(AEXPR_DISTINCT,
+															   "=", $1, $6, @2),
+									 @2);
 				}
 			| b_expr IS OF '(' type_list ')'		%prec IS
 				{
@@ -11234,10 +11235,9 @@ b_expr:		c_expr
 				}
 			| b_expr IS NOT DOCUMENT_P				%prec IS
 				{
-					$$ = (Node *) makeA_Expr(AEXPR_NOT, NIL, NULL,
-											 makeXmlExpr(IS_DOCUMENT, NULL, NIL,
-														 list_make1($1), @2),
-											 @2);
+					$$ = makeNotExpr(makeXmlExpr(IS_DOCUMENT, NULL, NIL,
+												 list_make1($1), @2),
+									 @2);
 				}
 		;
 
@@ -13692,6 +13692,46 @@ doNegateFloat(Value *v)
 		v->val.str = psprintf("-%s", oldval);
 }
 
+static Node *
+makeAndExpr(Node *lexpr, Node *rexpr, int location)
+{
+	/* Flatten "a AND b AND c ..." to a single BoolExpr on sight */
+	if (IsA(lexpr, BoolExpr))
+	{
+		BoolExpr *blexpr = (BoolExpr *) lexpr;
+
+		if (blexpr->boolop == AND_EXPR)
+		{
+			blexpr->args = lappend(blexpr->args, rexpr);
+			return (Node *) blexpr;
+		}
+	}
+	return (Node *) makeBoolExpr(AND_EXPR, list_make2(lexpr, rexpr), location);
+}
+
+static Node *
+makeOrExpr(Node *lexpr, Node *rexpr, int location)
+{
+	/* Flatten "a OR b OR c ..." to a single BoolExpr on sight */
+	if (IsA(lexpr, BoolExpr))
+	{
+		BoolExpr *blexpr = (BoolExpr *) lexpr;
+
+		if (blexpr->boolop == OR_EXPR)
+		{
+			blexpr->args = lappend(blexpr->args, rexpr);
+			return (Node *) blexpr;
+		}
+	}
+	return (Node *) makeBoolExpr(OR_EXPR, list_make2(lexpr, rexpr), location);
+}
+
+static Node *
+makeNotExpr(Node *expr, int location)
+{
+	return (Node *) makeBoolExpr(NOT_EXPR, list_make1(expr), location);
+}
+
 static Node *
 makeAArrayExpr(List *elements, int location)
 {
diff --git a/src/backend/parser/parse_clause.c b/src/backend/parser/parse_clause.c
@@ -332,7 +332,8 @@ transformJoinUsingClause(ParseState *pstate,
 						 RangeTblEntry *leftRTE, RangeTblEntry *rightRTE,
 						 List *leftVars, List *rightVars)
 {
-	Node	   *result = NULL;
+	Node	   *result;
+	List	   *andargs = NIL;
 	ListCell   *lvars,
 			   *rvars;
 
@@ -358,18 +359,16 @@ transformJoinUsingClause(ParseState *pstate,
 							 copyObject(lvar), copyObject(rvar),
 							 -1);
 
-		/* And combine into an AND clause, if multiple join columns */
-		if (result == NULL)
-			result = (Node *) e;
-		else
-		{
-			A_Expr	   *a;
-
-			a = makeA_Expr(AEXPR_AND, NIL, result, (Node *) e, -1);
-			result = (Node *) a;
-		}
+		/* Prepare to combine into an AND clause, if multiple join columns */
+		andargs = lappend(andargs, e);
 	}
 
+	/* Only need an AND if there's more than one join column */
+	if (list_length(andargs) == 1)
+		result = (Node *) linitial(andargs);
+	else
+		result = (Node *) makeBoolExpr(AND_EXPR, andargs, -1);
+
 	/*
 	 * Since the references are already Vars, and are certainly from the input
 	 * relations, we don't have to go through the same pushups that
diff --git a/src/backend/parser/parse_expr.c b/src/backend/parser/parse_expr.c
diff --git a/src/include/nodes/parsenodes.h b/src/include/nodes/parsenodes.h
diff --git a/src/include/nodes/primnodes.h b/src/include/nodes/primnodes.h
diff --git a/src/test/regress/expected/rules.out b/src/test/regress/expected/rules.out
