add tests

Author: rambleraptor

pyiceberg/table/snapshots.py

@@ -247,7 +247,7 @@ class Snapshot(IcebergBaseModel):
     first_row_id: Optional[int] = Field(
         alias="first-row-id", default=None, description="assigned to the first row in the first data file in the first manifest"
     )
-    _added_rows: Optional[int] = PrivateAttr()
+    added_rows: Optional[int] = Field(alias="added-rows", default=None, description="The upper bound of the number of rows with assigned row IDs")
 
     def __str__(self) -> str:
         """Return the string representation of the Snapshot class."""
@@ -261,10 +261,6 @@ def manifests(self, io: FileIO) -> List[ManifestFile]:
         """Return the manifests for the given snapshot."""
         return list(_manifests(io, self.manifest_list))
 
-    @property
-    def added_rows(self) -> Optional[int]:
-        return self._added_rows
-
 
 class MetadataLogEntry(IcebergBaseModel):
     metadata_file: str = Field(alias="metadata-file")

tests/integration/test_writes/test_writes.py

@@ -64,7 +64,7 @@
     StringType,
     UUIDType,
 )
-from utils import _create_table
+from utils import TABLE_SCHEMA, _create_table
 
 
 @pytest.fixture(scope="session", autouse=True)
@@ -2490,3 +2490,44 @@ def test_stage_only_overwrite_files(
     assert operations == ["append", "append", "delete", "append", "append"]
 
     assert parent_snapshot_id == [None, first_snapshot, second_snapshot, second_snapshot, second_snapshot]
+
+
+@pytest.mark.integration
+def test_v3_write_and_read(spark: SparkSession, session_catalog: Catalog) -> None:
+    """Test writing to a v3 table and reading with Spark."""
+    identifier = "default.test_v3_write_and_read"
+    tbl = _create_table(session_catalog, identifier, {"format-version": "3"})
+    assert tbl.format_version == 3, f"Expected v3, got: v{tbl.format_version}"
+    assert tbl.metadata.next_row_id is not None, "Expected next_row_id to be initialized"
+    initial_next_row_id = tbl.metadata.next_row_id
+
+    test_data = pa.Table.from_pydict(
+        {
+            "bool": [True, False, True],
+            "string": ["a", "b", "c"],
+            "string_long": ["a_long", "b_long", "c_long"],
+            "int": [1, 2, 3],
+            "long": [11, 22, 33],
+            "float": [1.1, 2.2, 3.3],
+            "double": [1.11, 2.22, 3.33],
+            "timestamp": [datetime(2023, 1, 1, 1, 1, 1), datetime(2023, 2, 2, 2, 2, 2), datetime(2023, 3, 3, 3, 3, 3)],
+            "timestamptz": [
+                datetime(2023, 1, 1, 1, 1, 1, tzinfo=pytz.utc),
+                datetime(2023, 2, 2, 2, 2, 2, tzinfo=pytz.utc),
+                datetime(2023, 3, 3, 3, 3, 3, tzinfo=pytz.utc),
+            ],
+            "date": [date(2023, 1, 1), date(2023, 2, 2), date(2023, 3, 3)],
+            "binary": [b"\x01", b"\x02", b"\x03"],
+            "fixed": [b"1234567890123456", b"1234567890123456", b"1234567890123456"],
+        },
+        schema=TABLE_SCHEMA.as_arrow(),
+    )
+
+    tbl.append(test_data)
+
+    assert (
+        tbl.metadata.next_row_id == initial_next_row_id + len(test_data)
+    ), "Expected next_row_id to be incremented by the number of added rows"
+
+    df = spark.table(identifier)
+    assert df.count() == 3, "Expected 3 rows"

tests/table/test_snapshots.py

@@ -139,15 +139,15 @@ def test_deserialize_snapshot_with_properties(snapshot_with_properties: Snapshot
 def test_snapshot_repr(snapshot: Snapshot) -> None:
     assert (
         repr(snapshot)
-        == """Snapshot(snapshot_id=25, parent_snapshot_id=19, sequence_number=200, timestamp_ms=1602638573590, manifest_list='s3:/a/b/c.avro', summary=Summary(Operation.APPEND), schema_id=3, first_row_id=None)"""
+        == """Snapshot(snapshot_id=25, parent_snapshot_id=19, sequence_number=200, timestamp_ms=1602638573590, manifest_list='s3:/a/b/c.avro', summary=Summary(Operation.APPEND), schema_id=3)"""
     )
     assert snapshot == eval(repr(snapshot))
 
 
 def test_snapshot_with_properties_repr(snapshot_with_properties: Snapshot) -> None:
     assert (
         repr(snapshot_with_properties)
-        == """Snapshot(snapshot_id=25, parent_snapshot_id=19, sequence_number=200, timestamp_ms=1602638573590, manifest_list='s3:/a/b/c.avro', summary=Summary(Operation.APPEND, **{'foo': 'bar'}), schema_id=3, first_row_id=None)"""
+        == """Snapshot(snapshot_id=25, parent_snapshot_id=19, sequence_number=200, timestamp_ms=1602638573590, manifest_list='s3:/a/b/c.avro', summary=Summary(Operation.APPEND, **{'foo': 'bar'}), schema_id=3)"""
     )
     assert snapshot_with_properties == eval(repr(snapshot_with_properties))