Add hook in select_common_type for CHAR/NCHAR case (#140)

* Fix SELECT DISTINCT NULL result type

With distinct, Postgres coerces the type of unknown-typed targets so it
may find an equality function. It selects TEXT as the type to coerce to.

The problem in T-SQL is if we are unioning the result of SELECT DISTINCT
NULL with a different type. Some types, BIT for example, do not have
conversion functions between TEXT. Therefore, SELECT CAST(1 as BIT)
UNION SELECT DISTINCT NULL fails unexpectantly in Babelfish.

This change coerces NULL values to INT4 instead of TEXT when using
Babelfish, fixing the result type and the union behavior.

Signed-off-by: Walt Boettge <[email protected]>
Task: BABEL-3348, BABEL-3392, BABEL-4157, BABEL-1874

src/backend/parser/analyze.c

@@ -45,6 +45,7 @@
 #include "parser/parse_relation.h"
 #include "parser/parse_target.h"
 #include "parser/parse_type.h"
+#include "parser/parser.h"
 #include "parser/parsetree.h"
 #include "rewrite/rewriteManip.h"
 #include "utils/backend_status.h"
@@ -85,7 +86,7 @@ set_target_table_alternative_hook_type set_target_table_alternative_hook = NULL;
 pre_transform_setop_tree_hook_type pre_transform_setop_tree_hook = NULL;
 
 /* Hook to reset a query's targetlist after modification in pre_transfrom_sort_clause */
-post_transform_sort_clause_hook_type post_transform_sort_clause_hook = NULL;
+pre_transform_setop_sort_clause_hook_type pre_transform_setop_sort_clause_hook = NULL;
 
 /* Hooks for handling unquoted string argumentss in T-SQL procedure calls */
 call_argument_unquoted_string_hook_type call_argument_unquoted_string_hook = NULL;
@@ -1938,15 +1939,15 @@ transformSetOperationStmt(ParseState *pstate, SelectStmt *stmt)
 	 */
 	tllen = list_length(qry->targetList);
 
+	if (pre_transform_setop_sort_clause_hook)
+		pre_transform_setop_sort_clause_hook(pstate, qry, sortClause, leftmostQuery);
+
 	qry->sortClause = transformSortClause(pstate,
 										  sortClause,
 										  &qry->targetList,
 										  EXPR_KIND_ORDER_BY,
 										  false /* allow SQL92 rules */ );
 
-	if (post_transform_sort_clause_hook)
-		post_transform_sort_clause_hook(qry, leftmostQuery);
-
 	/* restore namespace, remove join RTE from rtable */
 	pstate->p_namespace = sv_namespace;
 	pstate->p_rtable = list_truncate(pstate->p_rtable, sv_rtable_length);
@@ -2282,8 +2283,8 @@ transformSetOperationTree(ParseState *pstate, SelectStmt *stmt,
 			if (lcoltype != UNKNOWNOID)
 				lcolnode = coerce_to_common_type(pstate, lcolnode,
 												 rescoltype, context);
-			else if (IsA(lcolnode, Const) ||
-					 IsA(lcolnode, Param))
+			else if ((IsA(lcolnode, Const) ||
+					 IsA(lcolnode, Param)) && sql_dialect != SQL_DIALECT_TSQL)
 			{
 				lcolnode = coerce_to_common_type(pstate, lcolnode,
 												 rescoltype, context);
@@ -2293,8 +2294,8 @@ transformSetOperationTree(ParseState *pstate, SelectStmt *stmt,
 			if (rcoltype != UNKNOWNOID)
 				rcolnode = coerce_to_common_type(pstate, rcolnode,
 												 rescoltype, context);
-			else if (IsA(rcolnode, Const) ||
-					 IsA(rcolnode, Param))
+			else if ((IsA(lcolnode, Const) ||
+					 IsA(lcolnode, Param)) && sql_dialect != SQL_DIALECT_TSQL)
 			{
 				rcolnode = coerce_to_common_type(pstate, rcolnode,
 												 rescoltype, context);

src/backend/parser/parse_clause.c

@@ -3294,11 +3294,12 @@ addTargetToSortList(ParseState *pstate, TargetEntry *tle,
 	/* if tlist item is an UNKNOWN literal, change it to TEXT */
 	if (restype == UNKNOWNOID)
 	{
-		tle->expr = (Expr *) coerce_type(pstate, (Node *) tle->expr,
-										 restype, TEXTOID, -1,
-										 COERCION_IMPLICIT,
-										 COERCE_IMPLICIT_CAST,
-										 -1);
+		if (sql_dialect != SQL_DIALECT_TSQL || pstate->p_resolve_unknowns)
+			tle->expr = (Expr *) coerce_type(pstate, (Node *) tle->expr,
+											restype, TEXTOID, -1,
+											COERCION_IMPLICIT,
+											COERCE_IMPLICIT_CAST,
+											-1);
 		restype = TEXTOID;
 	}
 
@@ -3435,11 +3436,12 @@ addTargetToGroupList(ParseState *pstate, TargetEntry *tle,
 	/* if tlist item is an UNKNOWN literal, change it to TEXT */
 	if (restype == UNKNOWNOID)
 	{
-		tle->expr = (Expr *) coerce_type(pstate, (Node *) tle->expr,
-										 restype, TEXTOID, -1,
-										 COERCION_IMPLICIT,
-										 COERCE_IMPLICIT_CAST,
-										 -1);
+		if (sql_dialect != SQL_DIALECT_TSQL || pstate->p_resolve_unknowns)
+			tle->expr = (Expr *) coerce_type(pstate, (Node *) tle->expr,
+											restype, TEXTOID, -1,
+											COERCION_IMPLICIT,
+											COERCE_IMPLICIT_CAST,
+											-1);
 		restype = TEXTOID;
 	}
 

src/backend/parser/parse_coerce.c

@@ -38,6 +38,7 @@ determine_datatype_precedence_hook_type determine_datatype_precedence_hook = NUL
 coerce_string_literal_hook_type coerce_string_literal_hook = NULL;
 validate_implicit_conversion_from_string_literal_hook_type validate_implicit_conversion_from_string_literal_hook = NULL;
 select_common_type_hook_type select_common_type_hook = NULL;
+select_common_typmod_hook_type select_common_typmod_hook = NULL;
 
 static Node *coerce_type_typmod(Node *node,
 								Oid targetTypeId, int32 targetTypMod,
@@ -1384,6 +1385,13 @@ select_common_type(ParseState *pstate, List *exprs, const char *context,
 	ListCell   *lc;
 	const char *dump_restore = GetConfigOption("babelfishpg_tsql.dump_restore", true, false);
 
+	if (select_common_type_hook)
+	{
+		Oid result = (*select_common_type_hook)(pstate, exprs, context, which_expr);
+		if (result != InvalidOid)
+			return result;
+	}
+
 	Assert(exprs != NIL);
 	pexpr = (Node *) linitial(exprs);
 	lc = list_second_cell(exprs);
@@ -1719,6 +1727,13 @@ select_common_typmod(ParseState *pstate, List *exprs, Oid common_type)
 	bool		first = true;
 	int32		result = -1;
 
+	if (select_common_typmod_hook)
+	{
+		result = (*select_common_typmod_hook)(pstate, exprs, common_type);
+		if (result != -1)
+			return result;
+	}
+
 	foreach(lc, exprs)
 	{
 		Node	   *expr = (Node *) lfirst(lc);

src/backend/parser/parse_expr.c

@@ -2311,7 +2311,7 @@ transformCoalesceExpr(ParseState *pstate, CoalesceExpr *c)
 	}
 
 	if (sql_dialect == SQL_DIALECT_TSQL && select_common_type_hook && c->tsql_is_null)
-		newc->coalescetype = (*select_common_type_hook)(pstate, newargs);
+		newc->coalescetype = select_common_type(pstate, newargs, "ISNULL", NULL);
 	else
 		newc->coalescetype = select_common_type(pstate, newargs, "COALESCE", NULL);
 	/* coalescecollid will be set by parse_collate.c */

src/include/parser/analyze.h

@@ -61,8 +61,8 @@ typedef void (*pre_transform_setop_tree_hook_type) (SelectStmt *stmt, SelectStmt
 extern PGDLLIMPORT pre_transform_setop_tree_hook_type pre_transform_setop_tree_hook;
 
 /* Hook for handle target table before transforming from clause */
-typedef void (*post_transform_sort_clause_hook_type) (Query *qry, Query *leftmostQuery);
-extern PGDLLIMPORT post_transform_sort_clause_hook_type post_transform_sort_clause_hook;
+typedef void (*pre_transform_setop_sort_clause_hook_type) (ParseState *pstate, Query *qry, List *sortClause, Query *leftmostQuery);
+extern PGDLLIMPORT pre_transform_setop_sort_clause_hook_type pre_transform_setop_sort_clause_hook;
 
 /* Hooks for handling unquoted string argumentss in T-SQL procedure calls */
 typedef Node* (*call_argument_unquoted_string_hook_type)(Node *arg);

src/include/parser/parse_coerce.h

@@ -129,8 +129,8 @@ typedef Node *(*coerce_string_literal_hook_type) (ParseCallbackState *pcbstate,
  */
 typedef void (*validate_implicit_conversion_from_string_literal_hook_type) (Const *newcon, const char *value);
 
-/* Generic hook to override the behavior of select_common_type(...) */
-typedef Oid (*select_common_type_hook_type) (ParseState *pstate, List *exprs);
+typedef Oid (*select_common_type_hook_type) (ParseState *pstate, List *exprs, const char *context, Node **which_expr);
 extern PGDLLIMPORT select_common_type_hook_type select_common_type_hook;
+typedef int32 (*select_common_typmod_hook_type) (ParseState *pstate, List *exprs, Oid common_type);
 
 #endif							/* PARSE_COERCE_H */