feat: rename dbtime and dbdate dtypes to avoid future conflicts with pandas (#32)

Author: tswast

db_dtypes/__init__.py

@@ -34,8 +34,8 @@
 from db_dtypes import core
 
 
-date_dtype_name = "date"
-time_dtype_name = "time"
+date_dtype_name = "dbdate"
+time_dtype_name = "dbtime"
 
 pandas_release = packaging.version.parse(pandas.__version__).release
 

samples/snippets/pandas_date_and_time.py

@@ -14,16 +14,13 @@
 
 
 def pandas_date_and_time():
-    # fmt: off
     # [START bigquery_date_create]
 
     import datetime
     import pandas as pd
     import db_dtypes  # noqa import to register dtypes
 
-    dates = pd.Series(
-        [datetime.date(2021, 9, 17), '2021-9-18'],
-        dtype='date')
+    dates = pd.Series([datetime.date(2021, 9, 17), "2021-9-18"], dtype="dbdate")
 
     # [END bigquery_date_create]
     # [START bigquery_date_as_datetime]
@@ -33,7 +30,7 @@ def pandas_date_and_time():
     # [END bigquery_date_as_datetime]
     # [START bigquery_date_sub]
 
-    dates2 = pd.Series(['2021-1-1', '2021-1-2'], dtype='date')
+    dates2 = pd.Series(["2021-1-1", "2021-1-2"], dtype="dbdate")
     diffs = dates - dates2
 
     # [END bigquery_date_sub]
@@ -46,9 +43,7 @@ def pandas_date_and_time():
     # [END bigquery_date_do]
     # [START bigquery_time_create]
 
-    times = pd.Series(
-        [datetime.time(1, 2, 3, 456789), '12:00:00.6'],
-        dtype='time')
+    times = pd.Series([datetime.time(1, 2, 3, 456789), "12:00:00.6"], dtype="dbtime")
 
     # [END bigquery_time_create]
     # [START bigquery_time_as_timedelta]
@@ -67,7 +62,6 @@ def pandas_date_and_time():
     combined = dates + times
 
     # [END bigquery_combine2_date_time]
-    # fmt: on
 
     return (
         dates,

samples/snippets/pandas_date_and_time_test.py

@@ -35,7 +35,7 @@ def test_pandas_date_and_time():
         combined0,
     ) = pandas_date_and_time()
 
-    assert str(dates.dtype) == "date"
+    assert str(dates.dtype) == "dbdate"
     assert list(dates) == [datetime.date(2021, 9, 17), datetime.date(2021, 9, 18)]
 
     assert np.array_equal(
@@ -45,7 +45,7 @@ def test_pandas_date_and_time():
     assert np.array_equal(after, dates.astype("object") + do)
     assert np.array_equal(before, dates.astype("object") - do)
 
-    assert str(times.dtype) == "time"
+    assert str(times.dtype) == "dbtime"
     assert list(times) == [
         datetime.time(1, 2, 3, 456789),
         datetime.time(12, 0, 0, 600000),

tests/unit/test_arrow.py

@@ -25,14 +25,14 @@
 @pytest.mark.parametrize(
     ("series", "expected"),
     (
-        (pandas.Series([], dtype="date"), pyarrow.array([], type=pyarrow.date32())),
+        (pandas.Series([], dtype="dbdate"), pyarrow.array([], type=pyarrow.date32())),
         (
-            pandas.Series([None, None, None], dtype="date"),
+            pandas.Series([None, None, None], dtype="dbdate"),
             pyarrow.array([None, None, None], type=pyarrow.date32()),
         ),
         (
             pandas.Series(
-                [dt.date(2021, 9, 27), None, dt.date(2011, 9, 27)], dtype="date"
+                [dt.date(2021, 9, 27), None, dt.date(2011, 9, 27)], dtype="dbdate"
             ),
             pyarrow.array(
                 [dt.date(2021, 9, 27), None, dt.date(2011, 9, 27)],
@@ -42,21 +42,25 @@
         (
             pandas.Series(
                 [dt.date(1677, 9, 22), dt.date(1970, 1, 1), dt.date(2262, 4, 11)],
-                dtype="date",
+                dtype="dbdate",
             ),
             pyarrow.array(
                 [dt.date(1677, 9, 22), dt.date(1970, 1, 1), dt.date(2262, 4, 11)],
                 type=pyarrow.date32(),
             ),
         ),
-        (pandas.Series([], dtype="time"), pyarrow.array([], type=pyarrow.time64("ns"))),
         (
-            pandas.Series([None, None, None], dtype="time"),
+            pandas.Series([], dtype="dbtime"),
+            pyarrow.array([], type=pyarrow.time64("ns")),
+        ),
+        (
+            pandas.Series([None, None, None], dtype="dbtime"),
             pyarrow.array([None, None, None], type=pyarrow.time64("ns")),
         ),
         (
             pandas.Series(
-                [dt.time(0, 0, 0, 0), None, dt.time(23, 59, 59, 999_999)], dtype="time"
+                [dt.time(0, 0, 0, 0), None, dt.time(23, 59, 59, 999_999)],
+                dtype="dbtime",
             ),
             pyarrow.array(
                 [dt.time(0, 0, 0, 0), None, dt.time(23, 59, 59, 999_999)],
@@ -70,7 +74,7 @@
                     dt.time(12, 30, 15, 125_000),
                     dt.time(23, 59, 59, 999_999),
                 ],
-                dtype="time",
+                dtype="dbtime",
             ),
             pyarrow.array(
                 [
@@ -91,14 +95,14 @@ def test_to_arrow(series, expected):
 @pytest.mark.parametrize(
     ("series", "expected"),
     (
-        (pandas.Series([], dtype="date"), pyarrow.array([], type=pyarrow.date64())),
+        (pandas.Series([], dtype="dbdate"), pyarrow.array([], type=pyarrow.date64())),
         (
-            pandas.Series([None, None, None], dtype="date"),
+            pandas.Series([None, None, None], dtype="dbdate"),
             pyarrow.array([None, None, None], type=pyarrow.date64()),
         ),
         (
             pandas.Series(
-                [dt.date(2021, 9, 27), None, dt.date(2011, 9, 27)], dtype="date"
+                [dt.date(2021, 9, 27), None, dt.date(2011, 9, 27)], dtype="dbdate"
             ),
             pyarrow.array(
                 [dt.date(2021, 9, 27), None, dt.date(2011, 9, 27)],
@@ -108,21 +112,25 @@ def test_to_arrow(series, expected):
         (
             pandas.Series(
                 [dt.date(1677, 9, 22), dt.date(1970, 1, 1), dt.date(2262, 4, 11)],
-                dtype="date",
+                dtype="dbdate",
             ),
             pyarrow.array(
                 [dt.date(1677, 9, 22), dt.date(1970, 1, 1), dt.date(2262, 4, 11)],
                 type=pyarrow.date64(),
             ),
         ),
-        (pandas.Series([], dtype="time"), pyarrow.array([], type=pyarrow.time32("ms"))),
         (
-            pandas.Series([None, None, None], dtype="time"),
+            pandas.Series([], dtype="dbtime"),
+            pyarrow.array([], type=pyarrow.time32("ms")),
+        ),
+        (
+            pandas.Series([None, None, None], dtype="dbtime"),
             pyarrow.array([None, None, None], type=pyarrow.time32("ms")),
         ),
         (
             pandas.Series(
-                [dt.time(0, 0, 0, 0), None, dt.time(23, 59, 59, 999_000)], dtype="time"
+                [dt.time(0, 0, 0, 0), None, dt.time(23, 59, 59, 999_000)],
+                dtype="dbtime",
             ),
             pyarrow.array(
                 [dt.time(0, 0, 0, 0), None, dt.time(23, 59, 59, 999_000)],
@@ -131,7 +139,8 @@ def test_to_arrow(series, expected):
         ),
         (
             pandas.Series(
-                [dt.time(0, 0, 0, 0), None, dt.time(23, 59, 59, 999_999)], dtype="time"
+                [dt.time(0, 0, 0, 0), None, dt.time(23, 59, 59, 999_999)],
+                dtype="dbtime",
             ),
             pyarrow.array(
                 [dt.time(0, 0, 0, 0), None, dt.time(23, 59, 59, 999_999)],
@@ -145,7 +154,7 @@ def test_to_arrow(series, expected):
                     dt.time(12, 30, 15, 125_000),
                     dt.time(23, 59, 59, 999_999),
                 ],
-                dtype="time",
+                dtype="dbtime",
             ),
             pyarrow.array(
                 [

tests/unit/test_date.py

@@ -38,7 +38,7 @@
     ],
 )
 def test_date_parsing(value, expected):
-    assert pandas.Series([value], dtype="date")[0] == expected
+    assert pandas.Series([value], dtype="dbdate")[0] == expected
 
 
 @pytest.mark.parametrize(
@@ -59,4 +59,4 @@ def test_date_parsing(value, expected):
 )
 def test_date_parsing_errors(value, error):
     with pytest.raises(ValueError, match=error):
-        pandas.Series([value], dtype="date")
+        pandas.Series([value], dtype="dbdate")

tests/unit/test_dtypes.py

@@ -23,39 +23,39 @@
 pandas_release = packaging.version.parse(pd.__version__).release
 
 SAMPLE_RAW_VALUES = dict(
-    date=(datetime.date(2021, 2, 2), "2021-2-3", None),
-    time=(datetime.time(1, 2, 2), "1:2:3.5", None),
+    dbdate=(datetime.date(2021, 2, 2), "2021-2-3", None),
+    dbtime=(datetime.time(1, 2, 2), "1:2:3.5", None),
 )
 SAMPLE_VALUES = dict(
-    date=(
+    dbdate=(
         datetime.date(2021, 2, 2),
         datetime.date(2021, 2, 3),
         datetime.date(2021, 2, 4),
         datetime.date(2021, 2, 5),
     ),
-    time=(
+    dbtime=(
         datetime.time(1, 2, 2),
         datetime.time(1, 2, 3, 500000),
         datetime.time(1, 2, 4, 500000),
         datetime.time(1, 2, 5, 500000),
     ),
 )
 SAMPLE_DT_VALUES = dict(
-    date=(
+    dbdate=(
         "2021-02-02T00:00:00.000000",
         "2021-02-03T00:00:00.000000",
         "2021-02-04T00:00:00.000000",
         "2021-02-05T00:00:00.000000",
     ),
-    time=(
+    dbtime=(
         "1970-01-01T01:02:02.000000",
         "1970-01-01T01:02:03.500000",
         "1970-01-01T01:02:04.500000",
         "1970-01-01T01:02:05.500000",
     ),
 )
 
-for_date_and_time = pytest.mark.parametrize("dtype", ["date", "time"])
+for_date_and_time = pytest.mark.parametrize("dtype", ["dbdate", "dbtime"])
 
 
 def eq_na(a1, a2):
@@ -72,7 +72,7 @@ def register_dtype():
 def _cls(dtype):
     import db_dtypes
 
-    return getattr(db_dtypes, dtype.capitalize() + "Array")
+    return getattr(db_dtypes, dtype[2:].capitalize() + "Array")
 
 
 def _make_one(dtype):
@@ -322,7 +322,7 @@ def test_take(dtype, allow_fill, fill_value):
         if fill_value == 42:
             fill_value = expected_fill = (
                 datetime.date(1971, 4, 2)
-                if dtype == "date"
+                if dtype == "dbdate"
                 else datetime.time(0, 42, 42, 424242)
             )
         else:
@@ -441,7 +441,7 @@ def test_astype_copy(dtype):
     ],
 )
 def test_asdatetime(dtype, same):
-    a = _make_one("date")
+    a = _make_one("dbdate")
     for dt in dtype, np.dtype(dtype) if dtype != "datetime" else dtype:
         if same:
             b = a.astype(dt, copy=False)
@@ -480,7 +480,7 @@ def test_astimedelta(dtype):
         .astype("timedelta64[ns]" if dtype == "timedelta" else dtype)
     )
 
-    a = _cls("time")([t, None])
+    a = _cls("dbtime")([t, None])
     b = a.astype(dtype)
     np.array_equal(b[:1], expect)
     assert pd.isna(b[1]) and str(b[1]) == "NaT"
@@ -526,7 +526,7 @@ def test_min_max_median(dtype):
         if pandas_release >= (1, 2):
             assert (
                 a.median() == datetime.time(1, 2, 4)
-                if dtype == "time"
+                if dtype == "dbtime"
                 else datetime.date(2021, 2, 3)
             )
 
@@ -553,14 +553,14 @@ def test_min_max_median(dtype):
     if pandas_release >= (1, 2):
         assert (
             a.median() == datetime.time(1, 2, 2, 750000)
-            if dtype == "time"
+            if dtype == "dbtime"
             else datetime.date(2021, 2, 2)
         )
 
 
 def test_date_add():
-    dates = _cls("date")(SAMPLE_VALUES["date"])
-    times = _cls("time")(SAMPLE_VALUES["time"])
+    dates = _cls("dbdate")(SAMPLE_VALUES["dbdate"])
+    times = _cls("dbtime")(SAMPLE_VALUES["dbtime"])
     expect = dates.astype("datetime64") + times.astype("timedelta64")
 
     assert np.array_equal(dates + times, expect)
@@ -592,8 +592,8 @@ def test_date_add():
 
 
 def test_date_sub():
-    dates = _cls("date")(SAMPLE_VALUES["date"])
-    dates2 = _cls("date")(
+    dates = _cls("dbdate")(SAMPLE_VALUES["dbdate"])
+    dates2 = _cls("dbdate")(
         (
             datetime.date(2021, 1, 2),
             datetime.date(2021, 1, 3),

tests/unit/test_time.py

@@ -61,7 +61,7 @@
     ],
 )
 def test_time_parsing(value, expected):
-    assert pandas.Series([value], dtype="time")[0] == expected
+    assert pandas.Series([value], dtype="dbtime")[0] == expected
 
 
 @pytest.mark.parametrize(
@@ -81,4 +81,4 @@ def test_time_parsing(value, expected):
 )
 def test_time_parsing_errors(value, error):
     with pytest.raises(ValueError, match=error):
-        pandas.Series([value], dtype="time")
+        pandas.Series([value], dtype="dbtime")