VARINT and numerics type conversion

include/pgduckdb/pg/types.hpp

@@ -7,4 +7,5 @@ bool IsArrayType(Oid type_oid);
 bool IsDomainType(Oid type_oid);
 bool IsArrayDomainType(Oid type_oid);
 Oid GetBaseDuckColumnType(Oid attribute_type_oid);
-}
+Datum StringToNumeric(const char *str);
+} // namespace pgduckdb::pg

src/pg/types.cpp

@@ -2,8 +2,10 @@
 #include "pgduckdb/pgduckdb_utils.hpp"
 extern "C" {
 #include "postgres.h"
+#include "utils/fmgrprotos.h"
 #include "utils/lsyscache.h"
 #include "utils/syscache.h"
+#include "utils/numeric.h"
 #include "catalog/pg_type.h"
 #include "executor/tuptable.h"
 }
@@ -53,4 +55,16 @@ GetBaseDuckColumnType(Oid attribute_type_oid) {
 	return PostgresFunctionGuard(GetBaseDuckColumnType_C, attribute_type_oid);
 }
 
+static Datum
+StringToNumeric_C(const char *str) {
+	Datum pg_numeric = DirectFunctionCall3(numeric_in, CStringGetDatum(str), /*typelen=*/ObjectIdGetDatum(InvalidOid),
+	                                       /*typmod=*/Int32GetDatum(-1));
+	return pg_numeric;
+}
+
+Datum
+StringToNumeric(const char *str) {
+	return PostgresFunctionGuard(StringToNumeric_C, str);
+}
+
 } // namespace pgduckdb::pg

src/pgduckdb_types.cpp

@@ -19,16 +19,15 @@ extern "C" {
 #include "access/tupdesc_details.h"
 #include "catalog/pg_type.h"
 #include "executor/tuptable.h"
-#include "utils/builtins.h"
-#include "utils/numeric.h"
-#include "utils/uuid.h"
 #include "utils/array.h"
-#include "fmgr.h"
+#include "utils/builtins.h"
+#include "utils/date.h"
+#include "utils/jsonb.h"
 #include "utils/lsyscache.h"
+#include "utils/numeric.h"
 #include "utils/syscache.h"
-#include "utils/date.h"
 #include "utils/timestamp.h"
-#include "utils/jsonb.h"
+#include "utils/uuid.h"
 }
 
 #include "pgduckdb/pgduckdb_detoast.hpp"
@@ -337,6 +336,17 @@ ConvertNumeric(const duckdb::Value &ddb_value, idx_t scale, NumericVar &result)
 static Datum
 ConvertNumericDatum(const duckdb::Value &value) {
 	auto value_type_id = value.type().id();
+
+	// Special handle duckdb VARINT type.
+	if (value.type().id() == duckdb::LogicalTypeId::VARINT) {
+		// The performant way to handle the translation is to parse VARINT out, here we leverage string conversion and
+		// parsing mainly for code simplicity.
+		const std::string value_str = value.ToString();
+		Datum pg_numeric = pgduckdb::pg::StringToNumeric(value_str.c_str());
+		return pg_numeric;
+	}
+
+	// Special handle duckdb DOUBLE TYPE.
 	if (value_type_id == duckdb::LogicalTypeId::DOUBLE) {
 		return ConvertDoubleDatum(value);
 	}
@@ -1112,6 +1122,8 @@ GetPostgresDuckDBType(const duckdb::LogicalType &type) {
 		return NUMERICOID;
 	case duckdb::LogicalTypeId::UUID:
 		return UUIDOID;
+	case duckdb::LogicalTypeId::VARINT:
+		return NUMERICOID;
 	case duckdb::LogicalTypeId::LIST: {
 		const duckdb::LogicalType *duck_type = &type;
 		while (duck_type->id() == duckdb::LogicalTypeId::LIST) {

test/regression/expected/test_all_types.out

@@ -4,7 +4,6 @@ SELECT * FROM duckdb.query($$
 FROM test_all_types()
 SELECT * exclude(
     tinyint, -- PG14 outputs this differently currently
-    varint,
     TIME,
     time_tz,
     bit,
@@ -28,7 +27,7 @@ SELECT * exclude(
     nested_int_array, -- The nested array has different lengths, which is not possible in PG
 )
 $$)
--[ RECORD 1 ]---+-------------------------------------------------------------------------------------------------------------------------------------
+-[ RECORD 1 ]---+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
 bool            | f
 smallint        | -32768
 int             | -2147483648
@@ -39,6 +38,7 @@ utinyint        | 0
 usmallint       | 0
 uint            | 0
 ubigint         | 0
+varint          | -179769313486231570814527423731704356798070567525844996598917476803157260780028538760589558632766878171540458953514382464234321326889464182768467546703537516986049910576551282076245490090389328944075868508455133942304583236903222948165808559332123348274797826204144723168738177180919299881250404026184124858368
 date            | 07-14-5881580
 timestamp       | Sun Jan 10 08:01:49.551616 294247
 timestamp_s     | Sun Jan 10 08:01:49.551616 294247
@@ -60,7 +60,7 @@ double_array    | {}
 date_array      | {}
 timestamp_array | {}
 varchar_array   | {}
--[ RECORD 2 ]---+-------------------------------------------------------------------------------------------------------------------------------------
+-[ RECORD 2 ]---+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
 bool            | t
 smallint        | 32767
 int             | 2147483647
@@ -71,6 +71,7 @@ utinyint        | 255
 usmallint       | 65535
 uint            | 4294967295
 ubigint         | 18446744073709551615
+varint          | 179769313486231570814527423731704356798070567525844996598917476803157260780028538760589558632766878171540458953514382464234321326889464182768467546703537516986049910576551282076245490090389328944075868508455133942304583236903222948165808559332123348274797826204144723168738177180919299881250404026184124858368
 date            | 07-10-5881580
 timestamp       | Sun Jan 10 04:00:54.775806 294247
 timestamp_s     | Sun Jan 10 04:00:54 294247
@@ -92,7 +93,7 @@ double_array    | {42,NaN,Infinity,-Infinity,NULL,-42}
 date_array      | {01-01-1970,07-11-5881580,07-13-5881580,NULL,05-12-2022}
 timestamp_array | {"Thu Jan 01 00:00:00 1970","Sun Jan 10 04:00:54.775807 294247","Sun Jan 10 04:00:54.775809 294247",NULL,"Thu May 12 16:23:45 2022"}
 varchar_array   | {🦆🦆🦆🦆🦆🦆,goose,NULL,""}
--[ RECORD 3 ]---+-------------------------------------------------------------------------------------------------------------------------------------
+-[ RECORD 3 ]---+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
 bool            | 
 smallint        | 
 int             | 
@@ -103,6 +104,7 @@ utinyint        |
 usmallint       | 
 uint            | 
 ubigint         | 
+varint          | 
 date            | 
 timestamp       | 
 timestamp_s     | 

test/regression/sql/test_all_types.sql

@@ -4,7 +4,6 @@ SELECT * FROM duckdb.query($$
 FROM test_all_types()
 SELECT * exclude(
     tinyint, -- PG14 outputs this differently currently
-    varint,
     TIME,
     time_tz,
     bit,