Re: Gram.y patches for better parenthesis handling.

From: Larry Rosenman <ler(at)lerctr(dot)org>
To: PGSQL Hackers List <pgsql-hackers(at)hub(dot)org>
Cc: pgman(at)candle(dot)pha(dot)pa(dot)us, tgl(at)sss(dot)pgh(dot)pa(dot)us
Subject: Re: Gram.y patches for better parenthesis handling.
Date: 2000-10-28 16:38:53
Message-ID: 20001028113853.A25462@lerami.lerctr.org
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Err, with Tom's objections, why was this applied?
* Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us> [001028 11:34]:
> Applied. Thanks.
>
>
> > Okay, here's my attempt at fixing the problems with parentheses in
> > subqueries. It passes the normal 'runcheck' tests, and I've tried
> > a few simple things like
> > select 1 as foo union (((((select 2))))) order by foo;
> >
> > There are a few things that it doesn't do that have been talked
> > about here at least a little:
> >
> > 1) It doesn't allow things like "IN(((select 1)))" -- the select
> > here has to be at the top level. This is not new.
> >
> > 2) It does NOT preserve the odd syntax I found when I started looking
> > at this, where a SELECT statement could begin with parentheses. Thus,
> > (SELECT a from foo) order by a;
> > fails.
> >
> > I have preserved the ability, used in the regression tests, to
> > have a single select statement in what appears to be a RuleActionMulti
> > (but wasn't -- the parens were part of select_clause syntax).
> > In my version, this is a special form.
> >
> > This may cause some discussion: I have differentiated the two kinds
> > of RuleActionMulti. Perhaps nobody knew there were two kinds, because
> > I don't think the second form appears in the regression tests. This
> > one uses square brackets instead of parentheses, but originally was
> > otherwise the same as the one in parentheses. In this version of
> > gram.y, the square bracket form treats SELECT statements the same
> > as the other allowed statements. As discussed before on this list,
> > psql cannot make sense out of the results of such a thing, but an
> > application might. And I have designs on just such an application.
> >
> > ++ kevin
> >
> >
> >
> > --
> > Kevin O'Gorman (805) 650-6274 mailto:kogorman(at)pacbell(dot)net
> > Permanent e-mail forwarder: mailto:Kevin.O'Gorman(dot)64(at)Alum(dot)Dartmouth(dot)org
> > At school: mailto:kogorman(at)cs(dot)ucsb(dot)edu
> > Web: http://www.cs.ucsb.edu/~kogorman/index.html
> > Web: http://trixie.kosman.via.ayuda.com/~kevin/index.html
> >
> > "There is a freedom lying beyond circumstance,
> > derived from the direct intuition that life can
> > be grounded upon its absorption in what is
> > changeless amid change"
> > -- Alfred North Whitehead
>
> > --- gram.y.orig Thu Oct 26 13:13:04 2000
> > +++ gram.y Fri Oct 27 17:37:58 2000
> > @@ -124,14 +124,15 @@
> > DropGroupStmt, DropPLangStmt, DropSchemaStmt, DropStmt, DropTrigStmt,
> > DropUserStmt, DropdbStmt, ExplainStmt, ExtendStmt, FetchStmt,
> > GrantStmt, IndexStmt, InsertStmt, ListenStmt, LoadStmt, LockStmt,
> > - NotifyStmt, OptimizableStmt, ProcedureStmt, ReindexStmt,
> > + NotifyStmt, OptimizableStmt, ProcedureStmt
> > + QualifiedSelectStmt, ReindexStmt,
> > RemoveAggrStmt, RemoveFuncStmt, RemoveOperStmt, RemoveStmt,
> > RenameStmt, RevokeStmt, RuleActionStmt, RuleActionStmtOrEmpty,
> > RuleStmt, SelectStmt, SetSessionStmt, TransactionStmt, TruncateStmt,
> > UnlistenStmt, UpdateStmt, VacuumStmt, VariableResetStmt,
> > VariableSetStmt, VariableShowStmt, ViewStmt
> >
> > -%type <node> select_clause, select_subclause
> > +%type <node> subquery, simple_select, select_head, set_select
> >
> > %type <list> SessionList
> > %type <node> SessionClause
> > @@ -174,19 +175,20 @@
> > result, OptTempTableName, relation_name_list, OptTableElementList,
> > OptUnder, OptInherit, definition, opt_distinct,
> > opt_with, func_args, func_args_list, func_as,
> > - oper_argtypes, RuleActionList, RuleActionMulti,
> > + oper_argtypes, RuleActionList, RuleActionMulti,
> > + RuleActionOrSelectMulti, RuleActions, RuleActionBracket,
> > opt_column_list, columnList, opt_va_list, va_list,
> > sort_clause, sortby_list, index_params, index_list, name_list,
> > from_clause, from_list, opt_array_bounds,
> > expr_list, attrs, target_list, update_target_list,
> > def_list, opt_indirection, group_clause, TriggerFuncArgs,
> > - opt_select_limit
> > + opt_select_limit, select_limit
> >
> > %type <typnam> func_arg, func_return, aggr_argtype
> >
> > %type <boolean> opt_arg, TriggerForOpt, TriggerForType, OptTemp
> >
> > -%type <list> for_update_clause, update_list
> > +%type <list> opt_for_update_clause, for_update_clause, update_list
> > %type <boolean> opt_all
> > %type <boolean> opt_table
> > %type <boolean> opt_chain, opt_trans
> > @@ -2689,7 +2691,7 @@
> > RuleStmt: CREATE RULE name AS
> > { QueryIsRule=TRUE; }
> > ON event TO event_object where_clause
> > - DO opt_instead RuleActionList
> > + DO opt_instead RuleActions
> > {
> > RuleStmt *n = makeNode(RuleStmt);
> > n->rulename = $3;
> > @@ -2702,17 +2704,42 @@
> > }
> > ;
> >
> > -RuleActionList: NOTHING { $$ = NIL; }
> > - | SelectStmt { $$ = makeList1($1); }
> > - | RuleActionStmt { $$ = makeList1($1); }
> > - | '[' RuleActionMulti ']' { $$ = $2; }
> > - | '(' RuleActionMulti ')' { $$ = $2; }
> > +RuleActions: NOTHING { $$ = NIL; }
> > + | RuleActionStmt { $$ = makeList1($1); }
> > + | SelectStmt { $$ = makeList1($1); }
> > + | RuleActionList
> > + | RuleActionBracket
> > + ;
> > +
> > +/* LEGACY: Version 7.0 did not like SELECT statements in these lists,
> > + * but because of an oddity in the syntax for select_clause, allowed
> > + * certain forms like "DO INSTEAD (select 1)", and this is used in
> > + * the regression tests.
> > + * Here, we're allowing just one SELECT in parentheses, to preserve
> > + * any such expectations, and make the regression tests work.
> > + * ++ KO'G
> > + */
> > +RuleActionList: '(' RuleActionMulti ')' { $$ = $2; }
> > + | '(' SelectStmt ')' { $$ = makeList1($2); }
> > + ;
> > +
> > +/* An undocumented feature, bracketed lists are allowed to contain
> > + * SELECT statements on the same basis as the others. Before this,
> > + * they were the same as parenthesized lists, and did not allow
> > + * SelectStmts. Anybody know why they were here originally? Or if
> > + * they're in the regression tests at all?
> > + * ++ KO'G
> > + */
> > +RuleActionBracket: '[' RuleActionOrSelectMulti ']' { $$ = $2; }
> > ;
> >
> > /* the thrashing around here is to discard "empty" statements... */
> > RuleActionMulti: RuleActionMulti ';' RuleActionStmtOrEmpty
> > { if ($3 != (Node *) NULL)
> > - $$ = lappend($1, $3);
> > + if ($1 != NIL)
> > + $$ = lappend($1, $3);
> > + else
> > + $$ = makeList1($3);
> > else
> > $$ = $1;
> > }
> > @@ -2724,6 +2751,31 @@
> > }
> > ;
> >
> > +RuleActionOrSelectMulti: RuleActionOrSelectMulti ';' RuleActionStmtOrEmpty
> > + { if ($3 != (Node *) NULL)
> > + if ($1 != NIL)
> > + $$ = lappend($1, $3);
> > + else
> > + $$ = makeList1($3);
> > + else
> > + $$ = $1;
> > + }
> > + | RuleActionOrSelectMulti ';' SelectStmt
> > + { if ($1 != NIL)
> > + $$ = lappend($1, $3);
> > + else
> > + $$ = makeList1($3);
> > + }
> > + | RuleActionStmtOrEmpty
> > + { if ($1 != (Node *) NULL)
> > + $$ = makeList1($1);
> > + else
> > + $$ = NIL;
> > + }
> > + | SelectStmt { $$ = makeList1($1); }
> > + ;
> > +
> > +
> > RuleActionStmt: InsertStmt
> > | UpdateStmt
> > | DeleteStmt
> > @@ -3289,7 +3341,12 @@
> > * However, this is not checked by the grammar; parse analysis must check it.
> > */
> >
> > -SelectStmt: select_clause sort_clause for_update_clause opt_select_limit
> > +SelectStmt: QualifiedSelectStmt
> > + | select_head
> > + ;
> > +
> > +QualifiedSelectStmt:
> > + select_head sort_clause opt_for_update_clause opt_select_limit
> > {
> > SelectStmt *n = findLeftmostSelect($1);
> >
> > @@ -3299,34 +3356,35 @@
> > n->limitCount = nth(1, $4);
> > $$ = $1;
> > }
> > - ;
> > -
> > -/* This rule parses Select statements that can appear within set operations,
> > - * including UNION, INTERSECT and EXCEPT. '(' and ')' can be used to specify
> > - * the ordering of the set operations. Without '(' and ')' we want the
> > - * operations to be ordered per the precedence specs at the head of this file.
> > - *
> > - * Since parentheses around SELECTs also appear in the expression grammar,
> > - * there is a parse ambiguity if parentheses are allowed at the top level of a
> > - * select_clause: are the parens part of the expression or part of the select?
> > - * We separate select_clause into two levels to resolve this: select_clause
> > - * can have top-level parentheses, select_subclause cannot.
> > - *
> > - * Note that sort clauses cannot be included at this level --- a sort clause
> > - * can only appear at the end of the complete Select, and it will be handled
> > - * by the topmost SelectStmt rule. Likewise FOR UPDATE and LIMIT.
> > - */
> > -select_clause: '(' select_subclause ')'
> > + | select_head for_update_clause opt_select_limit
> > {
> > - $$ = $2;
> > + SelectStmt *n = findLeftmostSelect($1);
> > +
> > + n->sortClause = NULL;
> > + n->forUpdate = $2;
> > + n->limitOffset = nth(0, $3);
> > + n->limitCount = nth(1, $3);
> > + $$ = $1;
> > }
> > - | select_subclause
> > + | select_head select_limit
> > {
> > - $$ = $1;
> > + SelectStmt *n = findLeftmostSelect($1);
> > +
> > + n->sortClause = NULL;
> > + n->forUpdate = NULL;
> > + n->limitOffset = nth(0, $2);
> > + n->limitCount = nth(1, $2);
> > + $$ = $1;
> > }
> > ;
> >
> > -select_subclause: SELECT opt_distinct target_list
> > +subquery: '(' subquery ')' { $$ = $2; }
> > + | '(' QualifiedSelectStmt ')' { $$ = $2; }
> > + | '(' set_select ')' { $$ = $2; }
> > + | simple_select { $$ = $1; }
> > + ;
> > +
> > +simple_select: SELECT opt_distinct target_list
> > result from_clause where_clause
> > group_clause having_clause
> > {
> > @@ -3341,7 +3399,13 @@
> > n->havingClause = $8;
> > $$ = (Node *)n;
> > }
> > - | select_clause UNION opt_all select_clause
> > + ;
> > +
> > +select_head: simple_select { $$ = $1; }
> > + | set_select { $$ = $1; }
> > + ;
> > +
> > +set_select: select_head UNION opt_all subquery
> > {
> > SetOperationStmt *n = makeNode(SetOperationStmt);
> > n->op = SETOP_UNION;
> > @@ -3350,7 +3414,7 @@
> > n->rarg = $4;
> > $$ = (Node *) n;
> > }
> > - | select_clause INTERSECT opt_all select_clause
> > + | select_head INTERSECT opt_all subquery
> > {
> > SetOperationStmt *n = makeNode(SetOperationStmt);
> > n->op = SETOP_INTERSECT;
> > @@ -3359,7 +3423,7 @@
> > n->rarg = $4;
> > $$ = (Node *) n;
> > }
> > - | select_clause EXCEPT opt_all select_clause
> > + | select_head EXCEPT opt_all subquery
> > {
> > SetOperationStmt *n = makeNode(SetOperationStmt);
> > n->op = SETOP_EXCEPT;
> > @@ -3424,7 +3488,6 @@
> > ;
> >
> > sort_clause: ORDER BY sortby_list { $$ = $3; }
> > - | /*EMPTY*/ { $$ = NIL; }
> > ;
> >
> > sortby_list: sortby { $$ = makeList1($1); }
> > @@ -3446,7 +3509,7 @@
> > ;
> >
> >
> > -opt_select_limit: LIMIT select_limit_value ',' select_offset_value
> > +select_limit: LIMIT select_limit_value ',' select_offset_value
> > { $$ = makeList2($4, $2); }
> > | LIMIT select_limit_value OFFSET select_offset_value
> > { $$ = makeList2($4, $2); }
> > @@ -3456,6 +3519,9 @@
> > { $$ = makeList2($2, $4); }
> > | OFFSET select_offset_value
> > { $$ = makeList2($2, NULL); }
> > + ;
> > +
> > +opt_select_limit: select_limit { $$ = $1; }
> > | /* EMPTY */
> > { $$ = makeList2(NULL, NULL); }
> > ;
> > @@ -3555,6 +3621,9 @@
> >
> > for_update_clause: FOR UPDATE update_list { $$ = $3; }
> > | FOR READ ONLY { $$ = NULL; }
> > + ;
> > +
> > +opt_for_update_clause: for_update_clause { $$ = $1; }
> > | /* EMPTY */ { $$ = NULL; }
> > ;
> >
> > @@ -3598,7 +3667,7 @@
> > $1->name = $2;
> > $$ = (Node *) $1;
> > }
> > - | '(' select_subclause ')' alias_clause
> > + | '(' SelectStmt ')' alias_clause
> > {
> > RangeSubselect *n = makeNode(RangeSubselect);
> > n->subquery = $2;
> > @@ -4134,7 +4203,7 @@
> > * Define row_descriptor to allow yacc to break the reduce/reduce conflict
> > * with singleton expressions.
> > */
> > -row_expr: '(' row_descriptor ')' IN '(' select_subclause ')'
> > +row_expr: '(' row_descriptor ')' IN '(' SelectStmt ')'
> > {
> > SubLink *n = makeNode(SubLink);
> > n->lefthand = $2;
> > @@ -4144,7 +4213,7 @@
> > n->subselect = $6;
> > $$ = (Node *)n;
> > }
> > - | '(' row_descriptor ')' NOT IN '(' select_subclause ')'
> > + | '(' row_descriptor ')' NOT IN '(' SelectStmt ')'
> > {
> > SubLink *n = makeNode(SubLink);
> > n->lefthand = $2;
> > @@ -4154,7 +4223,7 @@
> > n->subselect = $7;
> > $$ = (Node *)n;
> > }
> > - | '(' row_descriptor ')' all_Op sub_type '(' select_subclause ')'
> > + | '(' row_descriptor ')' all_Op sub_type '(' SelectStmt ')'
> > {
> > SubLink *n = makeNode(SubLink);
> > n->lefthand = $2;
> > @@ -4167,7 +4236,7 @@
> > n->subselect = $7;
> > $$ = (Node *)n;
> > }
> > - | '(' row_descriptor ')' all_Op '(' select_subclause ')'
> > + | '(' row_descriptor ')' all_Op '(' SelectStmt ')'
> > {
> > SubLink *n = makeNode(SubLink);
> > n->lefthand = $2;
> > @@ -4498,7 +4567,7 @@
> > $$ = n;
> > }
> > }
> > - | a_expr all_Op sub_type '(' select_subclause ')'
> > + | a_expr all_Op sub_type '(' SelectStmt ')'
> > {
> > SubLink *n = makeNode(SubLink);
> > n->lefthand = makeList1($1);
> > @@ -4894,7 +4963,7 @@
> > n->agg_distinct = FALSE;
> > $$ = (Node *)n;
> > }
> > - | '(' select_subclause ')'
> > + | '(' SelectStmt ')'
> > {
> > SubLink *n = makeNode(SubLink);
> > n->lefthand = NIL;
> > @@ -4904,7 +4973,7 @@
> > n->subselect = $2;
> > $$ = (Node *)n;
> > }
> > - | EXISTS '(' select_subclause ')'
> > + | EXISTS '(' SelectStmt ')'
> > {
> > SubLink *n = makeNode(SubLink);
> > n->lefthand = NIL;
> > @@ -5003,7 +5072,7 @@
> > { $$ = $1; }
> > ;
> >
> > -in_expr: select_subclause
> > +in_expr: SelectStmt
> > {
> > SubLink *n = makeNode(SubLink);
> > n->subselect = $1;
>
>
> --
> Bruce Momjian | http://candle.pha.pa.us
> pgman(at)candle(dot)pha(dot)pa(dot)us | (610) 853-3000
> + If your life is a hard drive, | 830 Blythe Avenue
> + Christ can be your backup. | Drexel Hill, Pennsylvania 19026
--
Larry Rosenman http://www.lerctr.org/~ler
Phone: +1 972-414-9812 (voice) Internet: ler(at)lerctr(dot)org
US Mail: 1905 Steamboat Springs Drive, Garland, TX 75044-6749

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2000-10-28 16:41:19 Re: Gram.y patches for better parenthesis handling.
Previous Message Tom Lane 2000-10-28 16:35:29 Re: pgsql/src/backend/nodes (copyfuncs.c outfuncs.c print.c)