pre-commit add Python black hook; format Python with black

Standardizes all the Python code to black format https://black.readthedocs.io/en/stable/integrations/source_version_control.html https://www.flake8rules.com/
apache · Sep 19, 2024 · 5f766cd · 5f766cd
1 parent d04a8e0
commit 5f766cd
Show file tree

Hide file tree

Showing 148 changed files with 8,857 additions and 4,162 deletions.
diff --git a/.github/linters/ruff.toml b/.github/linters/ruff.toml
@@ -39,8 +39,8 @@ target-version = "py38"
 # Enable Pyflakes (`F`) and a subset of the pycodestyle (`E`)  codes by default.
 # Unlike Flake8, Ruff doesn't enable pycodestyle warnings (`W`) or
 # McCabe complexity (`C901`) by default.
-select = ["E4", "E7", "E9", "F"]
-ignore = ["E721", "E722", "E731", "F401", "F402", "F403", "F405", "F811", "F821", "F822", "F841", "F901"]
+select = ["E3", "E4", "E5", "E7", "E9", "F"]
+ignore = ["E501", "E721", "E722", "E731", "F401", "F402", "F403", "F405", "F811", "F821", "F822", "F841", "F901"]
 
 # Allow fix for all enabled rules (when `--fix`) is provided.
 fixable = ["ALL"]

diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
@@ -10,6 +10,10 @@ repos:
     hooks:
       - id: identity
       - id: check-hooks-apply
+  - repo: https://github.com/psf/black-pre-commit-mirror
+    rev: 24.8.0
+    hooks:
+      - id: black
   - repo: https://github.com/codespell-project/codespell
     rev: v2.3.0
     hooks:

diff --git a/docs/usecases/ApacheSedonaCore.ipynb b/docs/usecases/ApacheSedonaCore.ipynb
@@ -40,7 +40,7 @@
     "from shapely.geometry import Polygon\n",
     "\n",
     "from sedona.spark import *\n",
-    "from sedona.core.geom.envelope import Envelope\n"
+    "from sedona.core.geom.envelope import Envelope"
    ]
   },
   {
@@ -182,12 +182,16 @@
     }
    ],
    "source": [
-    "config = SedonaContext.builder() .\\\n",
-    "    config('spark.jars.packages',\n",
-    "           'org.apache.sedona:sedona-spark-3.4_2.12:1.6.0,'\n",
-    "           'org.datasyslab:geotools-wrapper:1.6.0-28.2,'\n",
-    "           'uk.co.gresearch.spark:spark-extension_2.12:2.11.0-3.4'). \\\n",
-    "    getOrCreate()\n",
+    "config = (\n",
+    "    SedonaContext.builder()\n",
+    "    .config(\n",
+    "        \"spark.jars.packages\",\n",
+    "        \"org.apache.sedona:sedona-spark-3.4_2.12:1.6.0,\"\n",
+    "        \"org.datasyslab:geotools-wrapper:1.6.0-28.2,\"\n",
+    "        \"uk.co.gresearch.spark:spark-extension_2.12:2.11.0-3.4\",\n",
+    "    )\n",
+    "    .getOrCreate()\n",
+    ")\n",
     "\n",
     "sedona = SedonaContext.create(config)"
    ]
@@ -560,7 +564,9 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "point_rdd_to_geo = point_rdd.rawSpatialRDD.map(lambda x: [x.geom, *x.getUserData().split(\"\\t\")])"
+    "point_rdd_to_geo = point_rdd.rawSpatialRDD.map(\n",
+    "    lambda x: [x.geom, *x.getUserData().split(\"\\t\")]\n",
+    ")"
    ]
   },
   {
@@ -578,7 +584,9 @@
    ],
    "source": [
     "point_gdf = gpd.GeoDataFrame(\n",
-    "    point_rdd_to_geo.collect(), columns=[\"geom\", \"attr1\", \"attr2\", \"attr3\"], geometry=\"geom\"\n",
+    "    point_rdd_to_geo.collect(),\n",
+    "    columns=[\"geom\", \"attr1\", \"attr2\", \"attr3\"],\n",
+    "    geometry=\"geom\",\n",
     ")"
    ]
   },
@@ -696,9 +704,9 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "spatial_df = Adapter.\\\n",
-    "    toDf(point_rdd, [\"attr1\", \"attr2\", \"attr3\"], sedona).\\\n",
-    "    createOrReplaceTempView(\"spatial_df\")\n",
+    "spatial_df = Adapter.toDf(\n",
+    "    point_rdd, [\"attr1\", \"attr2\", \"attr3\"], sedona\n",
+    ").createOrReplaceTempView(\"spatial_df\")\n",
     "\n",
     "spatial_gdf = sedona.sql(\"Select attr1, attr2, attr3, geometry as geom from spatial_df\")"
    ]
@@ -995,10 +1003,16 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "rectangle_rdd = RectangleRDD(sc, \"data/zcta510-small.csv\", FileDataSplitter.CSV, True, 11)\n",
+    "rectangle_rdd = RectangleRDD(\n",
+    "    sc, \"data/zcta510-small.csv\", FileDataSplitter.CSV, True, 11\n",
+    ")\n",
     "point_rdd = PointRDD(sc, \"data/arealm-small.csv\", 1, FileDataSplitter.CSV, False, 11)\n",
-    "polygon_rdd = PolygonRDD(sc, \"data/primaryroads-polygon.csv\", FileDataSplitter.CSV, True, 11)\n",
-    "linestring_rdd = LineStringRDD(sc, \"data/primaryroads-linestring.csv\", FileDataSplitter.CSV, True)"
+    "polygon_rdd = PolygonRDD(\n",
+    "    sc, \"data/primaryroads-polygon.csv\", FileDataSplitter.CSV, True, 11\n",
+    ")\n",
+    "linestring_rdd = LineStringRDD(\n",
+    "    sc, \"data/primaryroads-linestring.csv\", FileDataSplitter.CSV, True\n",
+    ")"
    ]
   },
   {
@@ -1298,7 +1312,7 @@
     "schema = StructType(\n",
     "    [\n",
     "        StructField(\"geom_left\", GeometryType(), False),\n",
-    "        StructField(\"geom_right\", GeometryType(), False)\n",
+    "        StructField(\"geom_right\", GeometryType(), False),\n",
     "    ]\n",
     ")"
    ]
@@ -1497,7 +1511,9 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "spatial_join_result_non_flat = JoinQuery.SpatialJoinQuery(point_rdd, rectangle_rdd, False, True)"
+    "spatial_join_result_non_flat = JoinQuery.SpatialJoinQuery(\n",
+    "    point_rdd, rectangle_rdd, False, True\n",
+    ")"
    ]
   },
   {
@@ -1507,7 +1523,9 @@
    "outputs": [],
    "source": [
     "# number of point for each polygon\n",
-    "number_of_points = spatial_join_result_non_flat.map(lambda x: [x[0].geom, x[1].__len__()])"
+    "number_of_points = spatial_join_result_non_flat.map(\n",
+    "    lambda x: [x[0].geom, x[1].__len__()]\n",
+    ")"
    ]
   },
   {
@@ -1516,10 +1534,12 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "schema = StructType([\n",
-    "    StructField(\"geometry\", GeometryType(), False),\n",
-    "    StructField(\"number_of_points\", LongType(), False)\n",
-    "])"
+    "schema = StructType(\n",
+    "    [\n",
+    "        StructField(\"geometry\", GeometryType(), False),\n",
+    "        StructField(\"number_of_points\", LongType(), False),\n",
+    "    ]\n",
+    ")"
    ]
   },
   {
@@ -1650,10 +1670,14 @@
    ],
    "source": [
     "polygon = Polygon(\n",
-    "    [(-84.237756, 33.904859), (-84.237756, 34.090426),\n",
-    "     (-83.833011, 34.090426), (-83.833011, 33.904859),\n",
-    "     (-84.237756, 33.904859)\n",
-    "    ])\n",
+    "    [\n",
+    "        (-84.237756, 33.904859),\n",
+    "        (-84.237756, 34.090426),\n",
+    "        (-83.833011, 34.090426),\n",
+    "        (-83.833011, 33.904859),\n",
+    "        (-84.237756, 33.904859),\n",
+    "    ]\n",
+    ")\n",
     "polygons_nearby = KNNQuery.SpatialKnnQuery(polygon_rdd, polygon, 5, False)"
    ]
   },
@@ -1737,7 +1761,9 @@
    "source": [
     "query_envelope = Envelope(-85.01, -60.01, 34.01, 50.01)\n",
     "\n",
-    "result_range_query = RangeQuery.SpatialRangeQuery(linestring_rdd, query_envelope, False, False)"
+    "result_range_query = RangeQuery.SpatialRangeQuery(\n",
+    "    linestring_rdd, query_envelope, False, False\n",
+    ")"
    ]
   },
   {
@@ -1835,9 +1861,7 @@
    ],
    "source": [
     "sedona.createDataFrame(\n",
-    "    result_range_query.map(lambda x: [x.geom]),\n",
-    "    schema,\n",
-    "    verifySchema=False\n",
+    "    result_range_query.map(lambda x: [x.geom]), schema, verifySchema=False\n",
     ").show(5, True)"
    ]
   },
@@ -2325,7 +2349,9 @@
    "source": [
     "query_envelope = Envelope(-85.01, -60.01, 34.01, 50.01)\n",
     "\n",
-    "result_range_query = RangeQueryRaw.SpatialRangeQuery(linestring_rdd, query_envelope, False, False)"
+    "result_range_query = RangeQueryRaw.SpatialRangeQuery(\n",
+    "    linestring_rdd, query_envelope, False, False\n",
+    ")"
    ]
   },
   {