paleolimbot commented on code in PR #733:
URL: https://github.com/apache/sedona-db/pull/733#discussion_r2981973576
##########
python/sedonadb/tests/test_sjoin.py:
##########
@@ -68,6 +69,96 @@ def test_spatial_join(join_type, on):
eng_postgis.assert_query_result(sql, sedonadb_results)
+def _plan_text(df):
+ query_plan = df.to_pandas()
+ return "\n".join(query_plan.iloc[:, 1].astype(str).tolist())
+
+
+def _spatial_join_side_file_names(plan_text):
+ """Extract the left/right parquet file names used by `SpatialJoinExec`.
+
+ Example input:
+ SpatialJoinExec: join_type=Inner, on=ST_intersects(geo_right@0,
geo_left@0)
+ ProjectionExec: expr=[geometry@0 as geo_right]
+ DataSourceExec: file_groups={1 group:
[[.../natural-earth_countries_geo.parquet]]}, projection=[geometry],
file_type=parquet
+ ProbeShuffleExec: partitioning=RoundRobinBatch(1)
+ ProjectionExec: expr=[geometry@0 as geo_left]
+ DataSourceExec: file_groups={1 group:
[[.../natural-earth_cities_geo.parquet]]}, projection=[geometry],
file_type=parquet
+
+ Example output:
+ ["natural-earth_countries_geo", "natural-earth_cities_geo"]
+ """
+ spatial_join_idx = plan_text.find("SpatialJoinExec:")
+ assert spatial_join_idx != -1, plan_text
+
+ file_names = re.findall(
+ r"DataSourceExec:.*?/([^/\]]+)\.parquet", plan_text[spatial_join_idx:]
+ )
+ assert len(file_names) >= 2, plan_text
+ return file_names[:2]
+
+
+def test_spatial_join_reordering_can_be_disabled_e2e(geoarrow_data):
+ path_left = (
+ geoarrow_data / "natural-earth" / "files" /
"natural-earth_cities_geo.parquet"
+ )
+ path_right = (
+ geoarrow_data
+ / "natural-earth"
+ / "files"
+ / "natural-earth_countries_geo.parquet"
+ )
+ assert path_left.exists(), f"Missing test asset: {path_left}"
+ assert path_right.exists(), f"Missing test asset: {path_right}"
Review Comment:
I believe we have `sedonadb.testing.skip_if_not_exists()` for this (so the
tests can run without the submodule)
##########
python/sedonadb/tests/test_sjoin.py:
##########
@@ -68,6 +69,96 @@ def test_spatial_join(join_type, on):
eng_postgis.assert_query_result(sql, sedonadb_results)
+def _plan_text(df):
+ query_plan = df.to_pandas()
+ return "\n".join(query_plan.iloc[:, 1].astype(str).tolist())
+
+
+def _spatial_join_side_file_names(plan_text):
Review Comment:
I opened https://github.com/apache/sedona-db/issues/734 so that we can make
hopefully make this easier some day.
##########
python/sedonadb/tests/test_sjoin.py:
##########
@@ -68,6 +69,96 @@ def test_spatial_join(join_type, on):
eng_postgis.assert_query_result(sql, sedonadb_results)
+def _plan_text(df):
+ query_plan = df.to_pandas()
+ return "\n".join(query_plan.iloc[:, 1].astype(str).tolist())
+
+
+def _spatial_join_side_file_names(plan_text):
+ """Extract the left/right parquet file names used by `SpatialJoinExec`.
+
+ Example input:
+ SpatialJoinExec: join_type=Inner, on=ST_intersects(geo_right@0,
geo_left@0)
+ ProjectionExec: expr=[geometry@0 as geo_right]
+ DataSourceExec: file_groups={1 group:
[[.../natural-earth_countries_geo.parquet]]}, projection=[geometry],
file_type=parquet
+ ProbeShuffleExec: partitioning=RoundRobinBatch(1)
+ ProjectionExec: expr=[geometry@0 as geo_left]
+ DataSourceExec: file_groups={1 group:
[[.../natural-earth_cities_geo.parquet]]}, projection=[geometry],
file_type=parquet
+
+ Example output:
+ ["natural-earth_countries_geo", "natural-earth_cities_geo"]
+ """
+ spatial_join_idx = plan_text.find("SpatialJoinExec:")
+ assert spatial_join_idx != -1, plan_text
+
+ file_names = re.findall(
+ r"DataSourceExec:.*?/([^/\]]+)\.parquet", plan_text[spatial_join_idx:]
+ )
+ assert len(file_names) >= 2, plan_text
+ return file_names[:2]
+
+
+def test_spatial_join_reordering_can_be_disabled_e2e(geoarrow_data):
Review Comment:
Definitely ok! (Parquet tests use some of the larger files to ensure
everything works with realistic input)
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
