Fix issue when loading geoparquet file.

apache · Nov 15, 2024 · 95bd4bb · 95bd4bb
1 parent 8cf6935
commit 95bd4bb
Show file tree

Hide file tree

Showing 3 changed files with 27 additions and 1 deletion.
diff --git a/spark/spark-3.3/src/test/scala/org/apache/sedona/sql/geoparquetIOTests.scala b/spark/spark-3.3/src/test/scala/org/apache/sedona/sql/geoparquetIOTests.scala
@@ -54,6 +54,7 @@ class geoparquetIOTests extends TestBaseScala with BeforeAndAfterAll {
   val legacyparquetdatalocation: String =
     resourceFolder + "parquet/legacy-parquet-nested-columns.snappy.parquet"
   val geoparquetoutputlocation: String = resourceFolder + "geoparquet/geoparquet_output/"
+  val overtureBBOX: String = resourceFolder + "geoparquet/overture/bbox.geoparquet"
 
   override def afterAll(): Unit = FileUtils.deleteDirectory(new File(geoparquetoutputlocation))
 
@@ -732,6 +733,18 @@ class geoparquetIOTests extends TestBaseScala with BeforeAndAfterAll {
     }
   }
 
+  describe("loading one file geoparquet and filtering") {
+    it("should not fail when bbox is not available in geoparquet metadata") {
+      val numberOfRecords = sparkSession.read
+        .format("geoparquet")
+        .load(overtureBBOX)
+        .where("ST_Intersects(geometry, ST_PolygonFromEnvelope(0, 0, 1, 1))")
+        .count()
+
+      assert(numberOfRecords == 9)
+    }
+  }
+
   def validateGeoParquetMetadata(path: String)(body: org.json4s.JValue => Unit): Unit = {
     val parquetFiles = new File(path).listFiles().filter(_.getName.endsWith(".parquet"))
     parquetFiles.foreach { filePath =>

diff --git a/spark/spark-3.4/src/test/scala/org/apache/sedona/sql/geoparquetIOTests.scala b/spark/spark-3.4/src/test/scala/org/apache/sedona/sql/geoparquetIOTests.scala
@@ -54,6 +54,7 @@ class geoparquetIOTests extends TestBaseScala with BeforeAndAfterAll {
   val legacyparquetdatalocation: String =
     resourceFolder + "parquet/legacy-parquet-nested-columns.snappy.parquet"
   val geoparquetoutputlocation: String = resourceFolder + "geoparquet/geoparquet_output/"
+  val overtureBBOX: String = resourceFolder + "geoparquet/overture/bbox.geoparquet"
 
   override def afterAll(): Unit = FileUtils.deleteDirectory(new File(geoparquetoutputlocation))
 
@@ -758,6 +759,18 @@ class geoparquetIOTests extends TestBaseScala with BeforeAndAfterAll {
     }
   }
 
+  describe("loading one file geoparquet and filtering") {
+    it("should not fail when bbox is not available in geoparquet metadata") {
+      val numberOfRecords = sparkSession.read
+        .format("geoparquet")
+        .load(overtureBBOX)
+        .where("ST_Intersects(geometry, ST_PolygonFromEnvelope(0, 0, 1, 1))")
+        .count()
+
+      assert(numberOfRecords == 9)
+    }
+  }
+
   def validateGeoParquetMetadata(path: String)(body: org.json4s.JValue => Unit): Unit = {
     val parquetFiles = new File(path).listFiles().filter(_.getName.endsWith(".parquet"))
     parquetFiles.foreach { filePath =>

diff --git a/spark/spark-3.5/src/test/scala/org/apache/sedona/sql/geoparquetIOTests.scala b/spark/spark-3.5/src/test/scala/org/apache/sedona/sql/geoparquetIOTests.scala
@@ -762,7 +762,7 @@ class geoparquetIOTests extends TestBaseScala with BeforeAndAfterAll {
   }
 
   describe("loading one file geoparquet and filtering") {
-    it("filtering one file geoparquet") {
+    it("should not fail when bbox is not available in geoparquet metadata") {
       val numberOfRecords = sparkSession.read
         .format("geoparquet")
         .load(overtureBBOX)