Define the unittests using pytest

.github/workflows/python_test.yaml

@@ -53,7 +53,7 @@ jobs:
           pip install -r requirements.txt
           maturin develop
 
-          python -m unittest discover tests
+          pytest -v .
         env:
           CARGO_HOME: "/home/runner/.cargo"
           CARGO_TARGET_DIR: "/home/runner/target"

dev/release/rat_exclude_files.txt

@@ -105,3 +105,4 @@ benchmarks/queries/q*.sql
 ballista/rust/scheduler/testdata/*
 ballista/ui/scheduler/yarn.lock
 python/rust-toolchain
+python/requirements*.txt

python/requirements.in

@@ -17,3 +17,4 @@
 maturin
 toml
 pyarrow
+pytest

python/requirements.txt

@@ -1,25 +1,17 @@
-# Licensed to the Apache Software Foundation (ASF) under one
-# or more contributor license agreements.  See the NOTICE file
-# distributed with this work for additional information
-# regarding copyright ownership.  The ASF licenses this file
-# to you under the Apache License, Version 2.0 (the
-# "License"); you may not use this file except in compliance
-# with the License.  You may obtain a copy of the License at
-#
-#   http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing,
-# software distributed under the License is distributed on an
-# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
-# KIND, either express or implied.  See the License for the
-# specific language governing permissions and limitations
-# under the License.
 #
 # This file is autogenerated by pip-compile
 # To update, run:
 #
-#    pip-compile --generate-hashes
+#    pip-compile --generate-hashes requirements.in
 #
+attrs==21.2.0 \
+    --hash=sha256:149e90d6d8ac20db7a955ad60cf0e6881a3f20d37096140088356da6c716b0b1 \
+    --hash=sha256:ef6aaac3ca6cd92904cdd0d83f629a15f18053ec84e6432106f7a4d04ae4f5fb
+    # via pytest
+iniconfig==1.1.1 \
+    --hash=sha256:011e24c64b7f47f6ebd835bb12a743f2fbe9a26d4cecaa7f53bc4f35ee9da8b3 \
+    --hash=sha256:bc3af051d7d14b2ee5ef9969666def0cd1a000e121eaea580d4a313df4b37f32
+    # via pytest
 maturin==0.10.6 \
     --hash=sha256:0e81496f70a4805e6ea7dda7b0425246c111ccb119a2e22c64abeff131f4dd21 \
     --hash=sha256:3b5d5429bc05a816824420d99973f0cab39d8e274f6c3647bfd9afd95a030304 \
@@ -59,6 +51,18 @@ numpy==1.20.3 \
     --hash=sha256:f1452578d0516283c87608a5a5548b0cdde15b99650efdfd85182102ef7a7c17 \
     --hash=sha256:f39a995e47cb8649673cfa0579fbdd1cdd33ea497d1728a6cb194d6252268e48
     # via pyarrow
+packaging==20.9 \
+    --hash=sha256:5b327ac1320dc863dca72f4514ecc086f31186744b84a230374cc1fd776feae5 \
+    --hash=sha256:67714da7f7bc052e064859c05c595155bd1ee9f69f76557e21f051443c20947a
+    # via pytest
+pluggy==0.13.1 \
+    --hash=sha256:15b2acde666561e1298d71b523007ed7364de07029219b604cf808bfa1c765b0 \
+    --hash=sha256:966c145cd83c96502c3c3868f50408687b38434af77734af1e9ca461a4081d2d
+    # via pytest
+py==1.10.0 \
+    --hash=sha256:21b81bda15b66ef5e1a777a21c4dcd9c20ad3efd0b3f817e7a809035269e1bd3 \
+    --hash=sha256:3b80836aa6d1feeaa108e046da6423ab8f6ceda6468545ae8d02d9d58d18818a
+    # via pytest
 pyarrow==4.0.1 \
     --hash=sha256:04be0f7cb9090bd029b5b53bed628548fef569e5d0b5c6cd7f6d0106dbbc782d \
     --hash=sha256:0fde9c7a3d5d37f3fe5d18c4ed015e8f585b68b26d72a10d7012cad61afe43ff \
@@ -86,9 +90,18 @@ pyarrow==4.0.1 \
     --hash=sha256:fa7b165cfa97158c1e6d15c68428317b4f4ae786d1dc2dbab43f1328c1eb43aa \
     --hash=sha256:fe976695318560a97c6d31bba828eeca28c44c6f6401005e54ba476a28ac0a10
     # via -r requirements.in
+pyparsing==2.4.7 \
+    --hash=sha256:c203ec8783bf771a155b207279b9bccb8dea02d8f0c9e5f8ead507bc3246ecc1 \
+    --hash=sha256:ef9d7589ef3c200abe66653d3f1ab1033c3c419ae9b9bdb1240a85b024efc88b
+    # via packaging
+pytest==6.2.4 \
+    --hash=sha256:50bcad0a0b9c5a72c8e4e7c9855a3ad496ca6a881a3641b4260605450772c54b \
+    --hash=sha256:91ef2131a9bd6be8f76f1f08eac5c5317221d6ad1e143ae03894b862e8976890
+    # via -r requirements.in
 toml==0.10.2 \
     --hash=sha256:806143ae5bfb6a3c6e736a764057db0e6a0e05e338b5630894a5f779cabb4f9b \
     --hash=sha256:b3bda1d108d5dd99f4a20d24d9c348e91c4db7ab1b749200bded2f839ccbe68f
     # via
     #   -r requirements.in
     #   maturin
+    #   pytest

python/tests/generic.py

@@ -16,24 +16,30 @@
 # under the License.
 
 import datetime
-import numpy
-import pyarrow
+
+import numpy as np
+import pyarrow as pa
+import pyarrow.parquet as pq
 
 # used to write parquet files
-import pyarrow.parquet
 
 
 def data():
-    data = numpy.concatenate(
-        [numpy.random.normal(0, 0.01, size=50), numpy.random.normal(50, 0.01, size=50)]
+    np.random.seed(1)
+    data = np.concatenate(
+        [
+            np.random.normal(0, 0.01, size=50),
+            np.random.normal(50, 0.01, size=50),
+        ]
     )
-    return pyarrow.array(data)
+    return pa.array(data)
 
 
 def data_with_nans():
-    data = numpy.random.normal(0, 0.01, size=50)
-    mask = numpy.random.randint(0, 2, size=50)
-    data[mask == 0] = numpy.NaN
+    np.random.seed(0)
+    data = np.random.normal(0, 0.01, size=50)
+    mask = np.random.randint(0, 2, size=50)
+    data[mask == 0] = np.NaN
     return data
 
 
@@ -43,8 +49,19 @@ def data_datetime(f):
         datetime.datetime.now() - datetime.timedelta(days=1),
         datetime.datetime.now() + datetime.timedelta(days=1),
     ]
-    return pyarrow.array(
-        data, type=pyarrow.timestamp(f), mask=numpy.array([False, True, False])
+    return pa.array(
+        data, type=pa.timestamp(f), mask=np.array([False, True, False])
+    )
+
+
+def data_date32():
+    data = [
+        datetime.date(2000, 1, 1),
+        datetime.date(1980, 1, 1),
+        datetime.date(2030, 1, 1),
+    ]
+    return pa.array(
+        data, type=pa.date32(), mask=np.array([False, True, False])
     )
 
 
@@ -54,16 +71,16 @@ def data_timedelta(f):
         datetime.timedelta(days=1),
         datetime.timedelta(seconds=1),
     ]
-    return pyarrow.array(
-        data, type=pyarrow.duration(f), mask=numpy.array([False, True, False])
+    return pa.array(
+        data, type=pa.duration(f), mask=np.array([False, True, False])
     )
 
 
 def data_binary_other():
-    return numpy.array([1, 0, 0], dtype="u4")
+    return np.array([1, 0, 0], dtype="u4")
 
 
 def write_parquet(path, data):
-    table = pyarrow.Table.from_arrays([data], names=["a"])
-    pyarrow.parquet.write_table(table, path)
-    return path
+    table = pa.Table.from_arrays([data], names=["a"])
+    pq.write_table(table, path)
+    return str(path)

python/tests/test_df.py

@@ -15,100 +15,98 @@
 # specific language governing permissions and limitations
 # under the License.
 
-import unittest
-
 import pyarrow as pa
-import datafusion
+import pytest
+from datafusion import ExecutionContext
+from datafusion import functions as f
+
+
+@pytest.fixture
+def df():
+    ctx = ExecutionContext()
+
+    # create a RecordBatch and a new DataFrame from it
+    batch = pa.RecordBatch.from_arrays(
+        [pa.array([1, 2, 3]), pa.array([4, 5, 6])],
+        names=["a", "b"],
+    )
 
-f = datafusion.functions
+    return ctx.create_dataframe([[batch]])
 
 
-class TestCase(unittest.TestCase):
-    def _prepare(self):
-        ctx = datafusion.ExecutionContext()
+def test_select(df):
+    df = df.select(
+        f.col("a") + f.col("b"),
+        f.col("a") - f.col("b"),
+    )
 
-        # create a RecordBatch and a new DataFrame from it
-        batch = pa.RecordBatch.from_arrays(
-            [pa.array([1, 2, 3]), pa.array([4, 5, 6])],
-            names=["a", "b"],
-        )
-        return ctx.create_dataframe([[batch]])
+    # execute and collect the first (and only) batch
+    result = df.collect()[0]
 
-    def test_select(self):
-        df = self._prepare()
+    assert result.column(0) == pa.array([5, 7, 9])
+    assert result.column(1) == pa.array([-3, -3, -3])
 
-        df = df.select(
-            f.col("a") + f.col("b"),
-            f.col("a") - f.col("b"),
-        )
 
-        # execute and collect the first (and only) batch
-        result = df.collect()[0]
+def test_filter(df):
+    df = df.select(
+        f.col("a") + f.col("b"),
+        f.col("a") - f.col("b"),
+    ).filter(f.col("a") > f.lit(2))
 
-        self.assertEqual(result.column(0), pa.array([5, 7, 9]))
-        self.assertEqual(result.column(1), pa.array([-3, -3, -3]))
+    # execute and collect the first (and only) batch
+    result = df.collect()[0]
 
-    def test_filter(self):
-        df = self._prepare()
+    assert result.column(0) == pa.array([9])
+    assert result.column(1) == pa.array([-3])
 
-        df = df.select(
-            f.col("a") + f.col("b"),
-            f.col("a") - f.col("b"),
-        ).filter(f.col("a") > f.lit(2))
 
-        # execute and collect the first (and only) batch
-        result = df.collect()[0]
+def test_sort(df):
+    df = df.sort([f.col("b").sort(ascending=False)])
 
-        self.assertEqual(result.column(0), pa.array([9]))
-        self.assertEqual(result.column(1), pa.array([-3]))
+    table = pa.Table.from_batches(df.collect())
+    expected = {"a": [3, 2, 1], "b": [6, 5, 4]}
 
-    def test_sort(self):
-        df = self._prepare()
-        df = df.sort([f.col("b").sort(ascending=False)])
+    assert table.to_pydict() == expected
 
-        table = pa.Table.from_batches(df.collect())
-        expected = {"a": [3, 2, 1], "b": [6, 5, 4]}
-        self.assertEqual(table.to_pydict(), expected)
 
-    def test_limit(self):
-        df = self._prepare()
+def test_limit(df):
+    df = df.limit(1)
 
-        df = df.limit(1)
+    # execute and collect the first (and only) batch
+    result = df.collect()[0]
 
-        # execute and collect the first (and only) batch
-        result = df.collect()[0]
+    assert len(result.column(0)) == 1
+    assert len(result.column(1)) == 1
 
-        self.assertEqual(len(result.column(0)), 1)
-        self.assertEqual(len(result.column(1)), 1)
 
-    def test_udf(self):
-        df = self._prepare()
+def test_udf(df):
+    # is_null is a pa function over arrays
+    udf = f.udf(lambda x: x.is_null(), [pa.int64()], pa.bool_())
 
-        # is_null is a pa function over arrays
-        udf = f.udf(lambda x: x.is_null(), [pa.int64()], pa.bool_())
+    df = df.select(udf(f.col("a")))
+    result = df.collect()[0].column(0)
 
-        df = df.select(udf(f.col("a")))
+    assert result == pa.array([False, False, False])
 
-        self.assertEqual(df.collect()[0].column(0), pa.array([False, False, False]))
 
-    def test_join(self):
-        ctx = datafusion.ExecutionContext()
+def test_join():
+    ctx = ExecutionContext()
 
-        batch = pa.RecordBatch.from_arrays(
-            [pa.array([1, 2, 3]), pa.array([4, 5, 6])],
-            names=["a", "b"],
-        )
-        df = ctx.create_dataframe([[batch]])
+    batch = pa.RecordBatch.from_arrays(
+        [pa.array([1, 2, 3]), pa.array([4, 5, 6])],
+        names=["a", "b"],
+    )
+    df = ctx.create_dataframe([[batch]])
 
-        batch = pa.RecordBatch.from_arrays(
-            [pa.array([1, 2]), pa.array([8, 10])],
-            names=["a", "c"],
-        )
-        df1 = ctx.create_dataframe([[batch]])
+    batch = pa.RecordBatch.from_arrays(
+        [pa.array([1, 2]), pa.array([8, 10])],
+        names=["a", "c"],
+    )
+    df1 = ctx.create_dataframe([[batch]])
 
-        df = df.join(df1, on="a", how="inner")
-        df = df.sort([f.col("a").sort(ascending=True)])
-        table = pa.Table.from_batches(df.collect())
+    df = df.join(df1, on="a", how="inner")
+    df = df.sort([f.col("a").sort(ascending=True)])
+    table = pa.Table.from_batches(df.collect())
 
-        expected = {"a": [1, 2], "c": [8, 10], "b": [4, 5]}
-        self.assertEqual(table.to_pydict(), expected)
+    expected = {"a": [1, 2], "c": [8, 10], "b": [4, 5]}
+    assert table.to_pydict() == expected