feat: support loading single Shapefile from folder

Support for identifying and loading a single Shapefile if the source set for the ShapeSchemaReader is a folder. ING-4416
halestudio · Aug 6, 2024 · 5f51fcb · 5f51fcb
1 parent 34f8fd9
commit 5f51fcb
Show file tree

Hide file tree

Showing 2 changed files with 70 additions and 2 deletions.
diff --git a/...ihumboldt.hale.io.shp.test/src/eu/esdihumboldt/hale/io/shp/ShapeInstanceReaderTest.groovy b/...ihumboldt.hale.io.shp.test/src/eu/esdihumboldt/hale/io/shp/ShapeInstanceReaderTest.groovy
@@ -21,8 +21,10 @@ import static org.junit.Assert.assertNotNull
 import static org.junit.Assert.assertTrue
 
 import java.nio.charset.StandardCharsets
+import java.nio.file.Files
 import java.util.function.Consumer
 
+import org.apache.commons.io.IOUtils
 import org.junit.Test
 import org.locationtech.jts.geom.Geometry
 
@@ -47,6 +49,9 @@ import groovy.transform.TypeCheckingMode
 @CompileStatic
 class ShapeInstanceReaderTest {
 
+	/**
+	 * Test reading Shapefile instances using the Shapefile as schema.
+	 */
 	@Test
 	void testReadShapeInstances() {
 		Schema xmlSchema = TestUtil.loadSchema(getClass().getClassLoader().getResource("testdata/arokfnp/ikg.shp").toURI())
@@ -63,6 +68,38 @@ class ShapeInstanceReaderTest {
 		validateArokFnpIkg(list, 'the_geom')
 	}
 
+	/**
+	 * Test reading a single Shapefile from a folder.
+	 */
+	@Test
+	void testReadFromFolder() {
+		Schema xmlSchema = TestUtil.loadSchema(getClass().getClassLoader().getResource("testdata/arokfnp/ikg.shp").toURI())
+
+		File tempDir = Files.createTempDirectory("read-shape").toFile()
+		try {
+			def ext = ['shp', 'dbf', 'prj', 'shx']
+			ext.each {
+				IOUtils.copy(getClass().getClassLoader().getResource("testdata/arokfnp/ikg.$it"), new File(tempDir, "ikg.$it"))
+			}
+
+			InstanceCollection instances = loadInstances(xmlSchema, tempDir.toURI())
+
+			assertNotNull(instances)
+			List<Instance> list = instances.toList()
+
+			// test count
+			assertThat(list).hasSize(14)
+
+			// instance validation
+			validateArokFnpIkg(list, 'the_geom')
+		} finally {
+			tempDir.deleteDir()
+		}
+	}
+
+	/**
+	 * Test reading Shapefile instances using an XML schema.
+	 */
 	@Test
 	void testReadXsdInstances() {
 		Schema xmlSchema = TestUtil.loadSchema(getClass().getClassLoader().getResource("testdata/arokfnp/arok-fnp.xsd").toURI())

diff --git a/...oldt.hale.io.shp/src/eu/esdihumboldt/hale/io/shp/reader/internal/ShapeInstanceReader.java b/...oldt.hale.io.shp/src/eu/esdihumboldt/hale/io/shp/reader/internal/ShapeInstanceReader.java
@@ -15,7 +15,9 @@
  */
 package eu.esdihumboldt.hale.io.shp.reader.internal;
 
+import java.io.File;
 import java.io.IOException;
+import java.net.URI;
 import java.nio.charset.Charset;
 import java.text.MessageFormat;
 import java.util.Collection;
@@ -36,6 +38,7 @@
 import eu.esdihumboldt.hale.common.core.io.report.IOReport;
 import eu.esdihumboldt.hale.common.core.io.report.IOReporter;
 import eu.esdihumboldt.hale.common.core.io.report.impl.IOMessageImpl;
+import eu.esdihumboldt.hale.common.core.io.supplier.DefaultInputSupplier;
 import eu.esdihumboldt.hale.common.core.parameter.AbstractParameterValueDescriptor;
 import eu.esdihumboldt.hale.common.instance.io.InstanceReader;
 import eu.esdihumboldt.hale.common.instance.io.impl.AbstractInstanceReader;
@@ -98,7 +101,34 @@ protected IOReport execute(ProgressIndicator progress, IOReporter reporter)
 			throws IOProviderConfigurationException, IOException {
 		progress.begin(Messages.getString("ShapeSchemaProvider.1"), ProgressIndicator.UNKNOWN); //$NON-NLS-1$
 
-		ShapefileDataStore store = new ShapefileDataStore(getSource().getLocation().toURL());
+		URI loc = getSource().getLocation();
+		try {
+			File file = new File(loc);
+
+			// special handling for directory as source -> load single Shapefile
+			if (file.exists() && file.isDirectory()) {
+				File[] candidates = file
+						.listFiles(f -> f.isFile() && f.getName().toLowerCase().endsWith(".shp"));
+				if (candidates != null && candidates.length > 0) {
+					// use first Shapefile found in folder
+					loc = candidates[0].toURI();
+
+					if (candidates.length > 1) {
+						reporter.warn(
+								"Picked file {} to load from folder, other Shapefiles in folder found but ignored",
+								candidates[0].getName());
+					}
+					else {
+						reporter.info("Picked file {} to load from folder",
+								candidates[0].getName());
+					}
+				}
+			}
+		} catch (IllegalArgumentException e) {
+			// ignore
+		}
+
+		ShapefileDataStore store = new ShapefileDataStore(loc.toURL());
 		store.setCharset(getCharset());
 
 		progress.setCurrentTask("Extracting shape instances");
@@ -136,7 +166,8 @@ protected IOReport execute(ProgressIndicator progress, IOReporter reporter)
 			reporter.info(new IOMessageImpl(
 					"No type name supplied as parameter, trying to auto-detect the schema type.",
 					null));
-			TypeDefinition dataType = ShapeSchemaReader.readShapeType(getSource());
+			TypeDefinition dataType = ShapeSchemaReader
+					.readShapeType(new DefaultInputSupplier(loc));
 			if (dataType == null) {
 				throw new IOException("Could not read shapefile structure information");
 			}