vastdb 1.3.3__tar.gz → 1.3.4__tar.gz

{vastdb-1.3.3 → vastdb-1.3.4}/CHANGELOG.md

@@ -4,6 +4,13 @@ All notable changes to this project will be documented in this file.
 
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
 
+## [1.3.4] (2024-12-11)
+[1.3.4]: https://github.com/vast-data/vastdb_sdk/compare/v1.3.3...v1.3.4
+
+### Added
+ - Support predicate pushdown on unsigned integers
+
+
 ## [1.3.3] (2024-11-18)
 [1.3.3]: https://github.com/vast-data/vastdb_sdk/compare/v1.3.2...v1.3.3
 

{vastdb-1.3.3 → vastdb-1.3.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vastdb
-Version: 1.3.3
+Version: 1.3.4
 Summary: VAST Data SDK
 Home-page: https://github.com/vast-data/vastdb_sdk
 Author: VAST DATA

{vastdb-1.3.3 → vastdb-1.3.4}/setup.py

@@ -29,7 +29,7 @@ setup(
     name='vastdb',
     python_requires='>=3.9.0',
     description='VAST Data SDK',
-    version='1.3.3' + suffix,
+    version='1.3.4' + suffix,
     url='https://github.com/vast-data/vastdb_sdk',
     author='VAST DATA',
     author_email='hello@vastdata.com',

{vastdb-1.3.3 → vastdb-1.3.4}/vastdb/_internal.py

@@ -58,6 +58,10 @@ import vast_flatbuf.org.apache.arrow.computeir.flatbuf.Source as fb_source
 import vast_flatbuf.org.apache.arrow.computeir.flatbuf.StringLiteral as fb_string_lit
 import vast_flatbuf.org.apache.arrow.computeir.flatbuf.TimeLiteral as fb_time_lit
 import vast_flatbuf.org.apache.arrow.computeir.flatbuf.TimestampLiteral as fb_timestamp_lit
+import vast_flatbuf.org.apache.arrow.computeir.flatbuf.UInt8Literal as fb_uint8_lit
+import vast_flatbuf.org.apache.arrow.computeir.flatbuf.UInt16Literal as fb_uint16_lit
+import vast_flatbuf.org.apache.arrow.computeir.flatbuf.UInt32Literal as fb_uint32_lit
+import vast_flatbuf.org.apache.arrow.computeir.flatbuf.UInt64Literal as fb_uint64_lit
 import vast_flatbuf.org.apache.arrow.flatbuf.Binary as fb_binary
 import vast_flatbuf.org.apache.arrow.flatbuf.Bool as fb_bool
 import vast_flatbuf.org.apache.arrow.flatbuf.Date as fb_date
@@ -341,50 +345,57 @@ class Predicate:
         fb_expression.AddImpl(self.builder, offset_call)
         return fb_expression.End(self.builder)
 
+    # see https://github.com/apache/arrow/blob/main/format/Schema.fbs
+    # https://github.com/apache/arrow/blob/apache-arrow-7.0.0/experimental/computeir/Expression.fbs
+    # https://github.com/apache/arrow/blob/apache-arrow-7.0.0/experimental/computeir/Literal.fbs
     def build_literal(self, field: pa.Field, value):
         literal_type: Any
 
-        if field.type.equals(pa.int64()):
-            literal_type = fb_int64_lit
-            literal_impl = LiteralImpl.Int64Literal
+        if field.type.equals(pa.int64()) or field.type.equals(pa.uint64()):
+            is_signed = field.type.equals(pa.int64())
+            literal_type = fb_int64_lit if is_signed else fb_uint64_lit
+            literal_impl = LiteralImpl.Int64Literal if is_signed else LiteralImpl.UInt64Literal
 
             field_type_type = Type.Int
             fb_int.Start(self.builder)
             fb_int.AddBitWidth(self.builder, field.type.bit_width)
-            fb_int.AddIsSigned(self.builder, True)
+            fb_int.AddIsSigned(self.builder, is_signed)
             field_type = fb_int.End(self.builder)
 
             value = int(value)
-        elif field.type.equals(pa.int32()):
-            literal_type = fb_int32_lit
-            literal_impl = LiteralImpl.Int32Literal
+        elif field.type.equals(pa.int32()) or field.type.equals(pa.uint32()):
+            is_signed = field.type.equals(pa.int32())
+            literal_type = fb_int32_lit if is_signed else fb_uint32_lit
+            literal_impl = LiteralImpl.Int32Literal if is_signed else LiteralImpl.UInt32Literal
 
             field_type_type = Type.Int
             fb_int.Start(self.builder)
             fb_int.AddBitWidth(self.builder, field.type.bit_width)
-            fb_int.AddIsSigned(self.builder, True)
+            fb_int.AddIsSigned(self.builder, is_signed)
             field_type = fb_int.End(self.builder)
 
             value = int(value)
-        elif field.type.equals(pa.int16()):
-            literal_type = fb_int16_lit
-            literal_impl = LiteralImpl.Int16Literal
+        elif field.type.equals(pa.int16()) or field.type.equals(pa.uint16()):
+            is_signed = field.type.equals(pa.int16())
+            literal_type = fb_int16_lit if is_signed else fb_uint16_lit
+            literal_impl = LiteralImpl.Int16Literal if is_signed else LiteralImpl.UInt16Literal
 
             field_type_type = Type.Int
             fb_int.Start(self.builder)
             fb_int.AddBitWidth(self.builder, field.type.bit_width)
-            fb_int.AddIsSigned(self.builder, True)
+            fb_int.AddIsSigned(self.builder, is_signed)
             field_type = fb_int.End(self.builder)
 
             value = int(value)
-        elif field.type.equals(pa.int8()):
-            literal_type = fb_int8_lit
-            literal_impl = LiteralImpl.Int8Literal
+        elif field.type.equals(pa.int8()) or field.type.equals(pa.uint8()):
+            is_signed = field.type.equals(pa.int8())
+            literal_type = fb_int8_lit if is_signed else fb_uint8_lit
+            literal_impl = LiteralImpl.Int8Literal if is_signed else LiteralImpl.UInt8Literal
 
             field_type_type = Type.Int
             fb_int.Start(self.builder)
             fb_int.AddBitWidth(self.builder, field.type.bit_width)
-            fb_int.AddIsSigned(self.builder, True)
+            fb_int.AddIsSigned(self.builder, is_signed)
             field_type = fb_int.End(self.builder)
 
             value = int(value)

{vastdb-1.3.3 → vastdb-1.3.4}/vastdb/tests/metrics.py

@@ -29,7 +29,7 @@ class Table:
         self.conn = conn
         self.name = name
         columns = ", ".join(
-            f"{f.name} {_MAP_SQLITE_TYPES[f.type]}"
+            f"{f.name} {_MAP_SQLITE_TYPES[f.type]}"  # type: ignore
             for f in self.fields
         )
         cmd = f"CREATE TABLE {self.name} ({columns})"

{vastdb-1.3.3 → vastdb-1.3.4}/vastdb/tests/test_tables.py

@@ -388,6 +388,33 @@ def test_types(session, clean_bucket_name):
             assert select(t['ts9'] == ts_literal) == expected.filter(pc.field('ts9') == ts_literal)
 
 
+def test_unsigned_filters(session, clean_bucket_name):
+    columns = pa.schema([
+        ('a', pa.uint8()),
+        ('b', pa.uint16()),
+        ('c', pa.uint32()),
+        ('d', pa.uint64()),
+    ])
+
+    expected = pa.table(schema=columns, data=[
+        [1, 2, 3],
+        [11, 22, 33],
+        [111, 222, 333],
+        [1111, 2222, 3333],
+    ])
+
+    with prepare_data(session, clean_bucket_name, 's', 't', expected) as table:
+        def select(predicate):
+            return table.select(predicate=predicate).read_all()
+
+        assert select(True) == expected
+        for t in [table, ibis._]:
+            assert select(t['a'] > 2) == expected.filter(pc.field('a') > 2)
+            assert select(t['b'] > 22) == expected.filter(pc.field('b') > 22)
+            assert select(t['c'] > 222) == expected.filter(pc.field('c') > 222)
+            assert select(t['d'] > 2222) == expected.filter(pc.field('d') > 2222)
+
+
 def test_filters(session, clean_bucket_name):
     columns = pa.schema([
         ('a', pa.int32()),

{vastdb-1.3.3 → vastdb-1.3.4}/vastdb.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vastdb
-Version: 1.3.3
+Version: 1.3.4
 Summary: VAST Data SDK
 Home-page: https://github.com/vast-data/vastdb_sdk
 Author: VAST DATA