Support NULLS FIRST and NULLS LAST in ORDER BY clause (#257)

Author: dey4ss

src/parser/bison_parser.cpp

@@ -318,6 +318,7 @@ union HSQL_STYPE
   hsql::LockingClause* locking_t;
   hsql::OrderDescription* order;
   hsql::OrderType order_type;
+  hsql::NullOrdering null_ordering_t;
   hsql::ReferencesSpecification* references_spec_t;
   hsql::SetOperation* set_operator_t;
   hsql::TableConstraint* table_constraint_t;
@@ -347,7 +348,7 @@ union HSQL_STYPE
 
   // clang-format off
 
-#line 351 "bison_parser.h"
+#line 352 "bison_parser.h"
 
 };
 typedef union HSQL_STYPE HSQL_STYPE;

src/parser/bison_parser.h

@@ -142,6 +142,7 @@
   hsql::LockingClause* locking_t;
   hsql::OrderDescription* order;
   hsql::OrderType order_type;
+  hsql::NullOrdering null_ordering_t;
   hsql::ReferencesSpecification* references_spec_t;
   hsql::SetOperation* set_operator_t;
   hsql::TableConstraint* table_constraint_t;
@@ -176,7 +177,7 @@
  ** Destructor symbols
  *********************************/
 
-%destructor { } <fval> <ival> <bval> <join_type> <order_type> <datetime_field> <column_type_t> <column_constraint_t> <import_type_t> <lock_mode_t> <lock_wait_policy_t> <frame_type>
+%destructor { } <fval> <ival> <bval> <join_type> <order_type> <datetime_field> <column_type_t> <column_constraint_t> <import_type_t> <lock_mode_t> <lock_wait_policy_t> <frame_type> <null_ordering_t>
 %destructor {
   free($$.name);
   free($$.schema);
@@ -270,6 +271,7 @@
 %type <limit>                  opt_limit opt_top
 %type <order>                  order_desc
 %type <order_type>             opt_order_type
+%type <null_ordering_t>        opt_null_ordering
 %type <datetime_field>         datetime_field datetime_field_plural duration_field
 %type <column_t>               column_def
 %type <table_element_t>        table_elem
@@ -953,12 +955,33 @@ order_list : order_desc {
   $$ = $1;
 };
 
-order_desc : expr opt_order_type { $$ = new OrderDescription($2, $1); };
+order_desc : expr opt_order_type opt_null_ordering { $$ = new OrderDescription($2, $1, $3); };
 
 opt_order_type : ASC { $$ = kOrderAsc; }
 | DESC { $$ = kOrderDesc; }
 | /* empty */ { $$ = kOrderAsc; };
 
+opt_null_ordering : /* empty */ { $$ = NullOrdering::Undefined; }
+| IDENTIFIER IDENTIFIER {
+  auto null_ordering = NullOrdering::Undefined;
+  if (strcasecmp($1, "nulls") == 0) {
+    if (strcasecmp($2, "first") == 0) {
+      null_ordering = NullOrdering::First;
+    } else if (strcasecmp($2, "last") == 0) {
+      null_ordering = NullOrdering::Last;
+    }
+  }
+  free($1);
+  free($2);
+
+  if (null_ordering == NullOrdering::Undefined) {
+    yyerror(&yyloc, result, scanner, "Expected NULLS FIRST or NULLS LAST ordering.");
+    YYERROR;
+  }
+
+  $$ = null_ordering;
+};
+
 // TODO: TOP and LIMIT can take more than just int literals.
 
 opt_top : TOP int_literal { $$ = new LimitDescription($2, nullptr); }

src/parser/bison_parser.y

@@ -7,6 +7,7 @@
 
 namespace hsql {
 enum OrderType { kOrderAsc, kOrderDesc };
+enum NullOrdering { Undefined, First, Last };
 
 enum SetType { kSetUnion, kSetIntersect, kSetExcept };
 
@@ -15,11 +16,12 @@ enum RowLockWaitPolicy { NoWait, SkipLocked, None };
 
 // Description of the order by clause within a select statement.
 struct OrderDescription {
-  OrderDescription(OrderType type, Expr* expr);
+  OrderDescription(OrderType type, Expr* expr, NullOrdering null_ordering);
   virtual ~OrderDescription();
 
   OrderType type;
   Expr* expr;
+  NullOrdering null_ordering;
 };
 
 // Description of the limit clause within a select statement.

src/sql/SelectStatement.h

@@ -213,7 +213,8 @@ ShowStatement::~ShowStatement() {
 // SelectStatement.h
 
 // OrderDescription
-OrderDescription::OrderDescription(OrderType type, Expr* expr) : type(type), expr(expr) {}
+OrderDescription::OrderDescription(OrderType type, Expr* expr, NullOrdering null_ordering)
+    : type(type), expr(expr), null_ordering(null_ordering) {}
 
 OrderDescription::~OrderDescription() { delete expr; }
 

src/sql/statements.cpp

@@ -93,3 +93,12 @@
 !SELECT * FROM foo WITH HINT (?);
 !SELECT * FROM foo WITH HINT (CAST(column_a AS INT));
 !SELECT * FROM foo WITH HINT (AVG(another_column));
+# ORDER BY with NULL ordering.
+!SELECT * FROM students ORDER BY name ASC NULL FIRST;
+!SELECT * FROM students ORDER BY name ASC gibberish LAST;
+!SELECT * FROM students ORDER BY name NULLS FIRS;
+!SELECT * FROM students ORDER BY name NULLS;
+!SELECT * FROM students ORDER BY name ASC NULLS;
+!SELECT * FROM students ORDER BY name FIRST;
+!SELECT * FROM students ORDER BY name ASC LAST;
+!SELECT * FROM students ORDER BY name DESC NULLS gibberish;

test/queries/queries-bad.sql

@@ -172,18 +172,24 @@ TEST(SelectGroupDistinctTest) {
 }
 
 TEST(OrderByTest) {
-  TEST_PARSE_SINGLE_SQL("SELECT grade, city FROM students ORDER BY grade, city DESC;", kStmtSelect, SelectStatement,
-                        result, stmt);
+  TEST_PARSE_SINGLE_SQL("SELECT grade, city FROM students ORDER BY grade, city DESC NULLS FIRST, name NULLS LAST;",
+                        kStmtSelect, SelectStatement, result, stmt);
 
   ASSERT_NULL(stmt->whereClause);
   ASSERT_NOTNULL(stmt->order);
 
-  ASSERT_EQ(stmt->order->size(), 2);
+  ASSERT_EQ(stmt->order->size(), 3);
   ASSERT_EQ(stmt->order->at(0)->type, kOrderAsc);
   ASSERT_STREQ(stmt->order->at(0)->expr->name, "grade");
+  ASSERT_EQ(stmt->order->at(0)->null_ordering, NullOrdering::Undefined);
 
   ASSERT_EQ(stmt->order->at(1)->type, kOrderDesc);
   ASSERT_STREQ(stmt->order->at(1)->expr->name, "city");
+  ASSERT_EQ(stmt->order->at(1)->null_ordering, NullOrdering::First);
+
+  ASSERT_EQ(stmt->order->at(2)->type, kOrderAsc);
+  ASSERT_STREQ(stmt->order->at(2)->expr->name, "name");
+  ASSERT_EQ(stmt->order->at(2)->null_ordering, NullOrdering::Last);
 }
 
 TEST(SelectBetweenTest) {