Backport HSEARCH-4494 to branch 6.1 - Fix performance test failure caused by https://lists.jboss.org/pipermail/hibernate-dev/2016-January.txt not being available anymore

integrationtest/backend/tck/src/main/java/org/hibernate/search/integrationtest/backend/tck/work/IndexIndexerLargeDocumentsIT.java

@@ -12,7 +12,6 @@
 
 import java.io.IOException;
 import java.io.UncheckedIOException;
-import java.nio.charset.StandardCharsets;
 import java.util.concurrent.CompletableFuture;
 import java.util.function.IntFunction;
 
@@ -25,14 +24,14 @@
 import org.hibernate.search.engine.backend.work.execution.spi.IndexIndexer;
 import org.hibernate.search.integrationtest.backend.tck.testsupport.configuration.DefaultAnalysisDefinitions;
 import org.hibernate.search.integrationtest.backend.tck.testsupport.util.rule.SearchSetupHelper;
+import org.hibernate.search.util.impl.test.data.TextContent;
 import org.hibernate.search.util.impl.integrationtest.mapper.stub.SimpleMappedIndex;
 import org.hibernate.search.util.impl.test.annotation.TestForIssue;
 
 import org.junit.Before;
 import org.junit.Rule;
 import org.junit.Test;
 
-import com.google.common.io.Resources;
 import org.awaitility.Awaitility;
 
 /**
@@ -51,10 +50,7 @@ public class IndexIndexerLargeDocumentsIT {
 	private static final String GREAT_EXPECTATIONS;
 	static {
 		try {
-			GREAT_EXPECTATIONS = Resources.toString(
-					IndexIndexerLargeDocumentsIT.class.getResource( "/great_expectations.txt" ),
-					StandardCharsets.UTF_8
-			);
+			GREAT_EXPECTATIONS = TextContent.greatExpectations().read();
 		}
 		catch (IOException e) {
 			throw new UncheckedIOException( e );

integrationtest/performance/backend/base/src/main/java/org/hibernate/search/integrationtest/performance/backend/base/testsupport/dataset/DatasetHolder.java

@@ -8,8 +8,6 @@
 
 import java.io.IOException;
 
-import org.hibernate.search.integrationtest.performance.backend.base.testsupport.filesystem.TemporaryFileHolder;
-
 import org.openjdk.jmh.annotations.Level;
 import org.openjdk.jmh.annotations.Param;
 import org.openjdk.jmh.annotations.Scope;
@@ -19,14 +17,14 @@
 @State(Scope.Benchmark)
 public class DatasetHolder {
 
-	@Param({ Datasets.HIBERNATE_DEV_ML_2016_01 })
+	@Param({ Datasets.GREAT_EXPECTATIONS })
 	private String dataset;
 
 	private Dataset datasetInstance;
 
 	@Setup(Level.Trial)
-	public void setup(TemporaryFileHolder temporaryFileHolder) throws IOException {
-		datasetInstance = Datasets.createDataset( dataset, temporaryFileHolder.getCacheDirectory() );
+	public void setup() throws IOException {
+		datasetInstance = Datasets.createDataset( dataset );
 	}
 
 	public Dataset getDataset() {

integrationtest/performance/backend/base/src/main/java/org/hibernate/search/integrationtest/performance/backend/base/testsupport/dataset/Datasets.java

@@ -6,69 +6,36 @@
  */
 package org.hibernate.search.integrationtest.performance.backend.base.testsupport.dataset;
 
-import java.io.BufferedReader;
 import java.io.IOException;
-import java.io.InputStream;
-import java.net.URI;
-import java.nio.channels.Channels;
-import java.nio.channels.FileChannel;
-import java.nio.channels.ReadableByteChannel;
-import java.nio.file.Files;
-import java.nio.file.Path;
-import java.nio.file.StandardOpenOption;
 import java.util.List;
 
-public final class Datasets {
+import org.hibernate.search.util.impl.test.data.TextContent;
 
-	public static final String CONSTANT_TEXT = "constant-text";
+import com.google.common.io.CharSource;
 
-	public static final String HIBERNATE_DEV_ML_2016_01 = "hibernate-dev-ml-2016-01";
+public final class Datasets {
 
-	private static final String DATASET_CACHE_DIRECTORY = "dataset";
+	public static final String CONSTANT_TEXT = "constant-text";
 
-	private static final URI HIBERNATE_DEV_MAILING_LIST_URI =
-			URI.create( "https://lists.jboss.org/pipermail/hibernate-dev/2016-January.txt" );
+	public static final String GREAT_EXPECTATIONS = "hibernate-dev-ml-2016-01";
 
 	private Datasets() {
 	}
 
-	public static Dataset createDataset(String name, Path cacheDirectory)
+	public static Dataset createDataset(String name)
 			throws IOException {
 		switch ( name ) {
 			case CONSTANT_TEXT:
 				return new ConstantDataset();
-			case HIBERNATE_DEV_ML_2016_01:
-				Path path = fetch( name, HIBERNATE_DEV_MAILING_LIST_URI, cacheDirectory );
-				return new SampleDataset( parseMailingListDigest( path ) );
+			case GREAT_EXPECTATIONS:
+				return new SampleDataset( parseSimple( TextContent.greatExpectations() ) );
 			default:
 				throw new IllegalArgumentException( "Unknown dataset: " + name );
 		}
 	}
 
-	private static Path fetch(String name, URI uri, Path cacheDirectory) throws IOException {
-		Path datasetCacheDirectory = cacheDirectory.resolve( DATASET_CACHE_DIRECTORY );
-		if ( !Files.exists( datasetCacheDirectory ) ) {
-			Files.createDirectory( datasetCacheDirectory );
-		}
-
-		Path outputPath = datasetCacheDirectory.resolve( name );
-		if ( Files.exists( outputPath ) ) {
-			return outputPath;
-		}
-
-		try ( InputStream input = uri.toURL().openStream();
-				ReadableByteChannel inputChannel = Channels.newChannel( input );
-				FileChannel outputChannel = FileChannel.open(
-						outputPath, StandardOpenOption.WRITE, StandardOpenOption.CREATE_NEW ) ) {
-			outputChannel.transferFrom( inputChannel, 0, Long.MAX_VALUE );
-		}
-		return outputPath;
-	}
-
-	private static List<SampleDataset.DataSample> parseMailingListDigest(Path path) throws IOException {
-		try ( BufferedReader reader = Files.newBufferedReader( path ) ) {
-			return new MailingListDigestParser( reader ).parse();
-		}
+	private static List<SampleDataset.DataSample> parseSimple(CharSource charSource) throws IOException {
+		return charSource.readLines( new SimpleDataSampleParser() );
 	}
 
 }

integrationtest/performance/backend/base/src/main/java/org/hibernate/search/integrationtest/performance/backend/base/testsupport/dataset/SampleDataset.java

@@ -48,9 +48,9 @@ public void populate(MappedIndex index, DocumentElement documentElement, long do
 
 	public static class DataSample {
 
-		private final String shortText;
-		private final String longText;
-		private final int numeric;
+		final String shortText;
+		final String longText;
+		final int numeric;
 
 		public DataSample(String shortText, String longText, int numeric) {
 			// Just in case something turned wrong during sample generation

integrationtest/performance/backend/base/src/main/java/org/hibernate/search/integrationtest/performance/backend/base/testsupport/dataset/SimpleDataSampleParser.java

@@ -0,0 +1,65 @@
+/*
+ * Hibernate Search, full-text search for your domain model
+ *
+ * License: GNU Lesser General Public License (LGPL), version 2.1 or later
+ * See the lgpl.txt file in the root directory or <http://www.gnu.org/licenses/lgpl-2.1.html>.
+ */
+package org.hibernate.search.integrationtest.performance.backend.base.testsupport.dataset;
+
+import java.util.ArrayList;
+import java.util.List;
+
+import com.google.common.io.LineProcessor;
+
+final class SimpleDataSampleParser implements LineProcessor<List<SampleDataset.DataSample>> {
+
+	private static final int LINES_PER_SAMPLE = 10;
+
+	private final List<SampleDataset.DataSample> result = new ArrayList<>();
+
+	private int currentSampleId = 0;
+	private int currentLineCount = 0;
+	private String currentSampleFirstLine = null;
+	private final StringBuilder currentSampleWholeText = new StringBuilder();
+
+	@Override
+	public boolean processLine(String line) {
+		if ( line.isEmpty() ) {
+			return true;
+		}
+		if ( currentLineCount == 0 ) {
+			currentSampleFirstLine = line;
+		}
+		else {
+			currentSampleWholeText.append( "\n" );
+		}
+		currentSampleWholeText.append( line );
+		++currentLineCount;
+		if ( currentLineCount >= LINES_PER_SAMPLE ) {
+			pushCurrentSample();
+		}
+		return true;
+	}
+
+	@Override
+	public List<SampleDataset.DataSample> getResult() {
+		pushCurrentSample();
+		return result;
+	}
+
+	private void pushCurrentSample() {
+		if ( currentLineCount == 0 ) {
+			return;
+		}
+
+		result.add( new SampleDataset.DataSample(
+				currentSampleFirstLine, currentSampleWholeText.toString(), currentSampleId
+		) );
+		++this.currentSampleId;
+
+		currentLineCount = 0;
+		currentSampleFirstLine = null;
+		currentSampleWholeText.setLength( 0 );
+	}
+
+}

integrationtest/performance/backend/base/src/main/java/org/hibernate/search/integrationtest/performance/backend/base/testsupport/filesystem/TemporaryFileHolder.java

@@ -24,13 +24,6 @@
 @State(Scope.Benchmark)
 public class TemporaryFileHolder {
 
-	/**
-	 * Set this system property to an alternative path if you want
-	 * cache files (downloads) to be stored in a specific place
-	 * (other than "/tmp/" + TEST_DIR_PREFIX + "cache-path").
-	 */
-	private static final String CACHE_PATH_PROPERTY_KEY = "cache-path";
-
 	/**
 	 * Set this system property to an alternative path if you want
 	 * index files to be stored in a specific place
@@ -51,10 +44,6 @@ public void cleanUp() throws IOException {
 		deleteAll( toCleanUp );
 	}
 
-	public Path getCacheDirectory() throws IOException {
-		return getTemporaryDirectory( CACHE_PATH_PROPERTY_KEY, false );
-	}
-
 	public Path getIndexesDirectory() throws IOException {
 		return getTemporaryDirectory( INDEXES_PATH_PROPERTY_KEY, true );
 	}