refactor(integration): merge file-size tests into FileSizeComparisonIntegrationTest

Three-tier structure: CSV (baseline) → JNI/Rust (oracle) → Java (SUT).
Drops separate WriteFileSizeIntegrationTest and VortexVsCsvFileSizeTest.
Both vortex writers now use cascading(3) for fair comparison.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Author: dfa1

…ration/WriteFileSizeIntegrationTest.java …n/FileSizeComparisonIntegrationTest.javaintegration/src/test/java/io/github/dfa1/vortex/integration/WriteFileSizeIntegrationTest.java renamed to integration/src/test/java/io/github/dfa1/vortex/integration/FileSizeComparisonIntegrationTest.java integration/src/test/java/io/github/dfa1/vortex/integration/WriteFileSizeIntegrationTest.java renamed to integration/src/test/java/io/github/dfa1/vortex/integration/FileSizeComparisonIntegrationTest.java

@@ -5,7 +5,6 @@
 import dev.vortex.jni.NativeLoader;
 import io.github.dfa1.vortex.core.DType;
 import io.github.dfa1.vortex.core.PType;
-import io.github.dfa1.vortex.core.array.DoubleArray;
 import io.github.dfa1.vortex.core.array.LongArray;
 import io.github.dfa1.vortex.encoding.EncodingRegistry;
 import io.github.dfa1.vortex.io.VortexReader;
@@ -18,6 +17,7 @@
 import org.apache.arrow.vector.BigIntVector;
 import org.apache.arrow.vector.DateDayVector;
 import org.apache.arrow.vector.Float8Vector;
+import org.apache.arrow.vector.VarCharVector;
 import org.apache.arrow.vector.VectorSchemaRoot;
 import org.apache.arrow.vector.types.DateUnit;
 import org.apache.arrow.vector.types.FloatingPointPrecision;
@@ -27,39 +27,33 @@
 import org.junit.jupiter.api.Test;
 import org.junit.jupiter.api.io.TempDir;
 
+import java.io.BufferedWriter;
 import java.io.IOException;
 import java.nio.channels.FileChannel;
+import java.nio.charset.StandardCharsets;
 import java.nio.file.Files;
 import java.nio.file.Path;
 import java.nio.file.StandardOpenOption;
+import java.time.LocalDate;
 import java.util.HashMap;
 import java.util.List;
 import java.util.Map;
 
 import static org.assertj.core.api.Assertions.assertThat;
 
-/// Verifies that Java and JNI writers produce valid, readable files for the same numeric OHLC dataset,
-/// and captures file sizes for regression visibility.
+/// File-size comparison: CSV (baseline) vs JNI/Rust writer (oracle) vs Java writer (SUT).
 ///
-/// Schema: date(I32), open/high/low/close(F64), volume(I64) — 6 columns, no strings.
-class WriteFileSizeIntegrationTest {
+/// Schema: symbol(Utf8), date(I32), open/high/low/close(F64), volume(I64) — 7 columns.
+/// Both Vortex writers use cascading(3) to match the Rust default compression pipeline.
+class FileSizeComparisonIntegrationTest {
 
 	private static final int TOTAL_ROWS = 100_000;
 	private static final int BATCH_SIZE = 50_000;
 
-	private static final ArrowType F64_TYPE = new ArrowType.FloatingPoint(FloatingPointPrecision.DOUBLE);
-	private static final Schema JNI_SCHEMA = new Schema(List.of(
-			Field.notNullable("date", new ArrowType.Date(DateUnit.DAY)),
-			Field.notNullable("open", F64_TYPE),
-			Field.notNullable("high", F64_TYPE),
-			Field.notNullable("low", F64_TYPE),
-			Field.notNullable("close", F64_TYPE),
-			Field.notNullable("volume", new ArrowType.Int(64, true))
-	));
-
 	private static final DType.Struct JAVA_SCHEMA = new DType.Struct(
-			List.of("date", "open", "high", "low", "close", "volume"),
+			List.of("symbol", "date", "open", "high", "low", "close", "volume"),
 			List.of(
+					new DType.Utf8(false),
 					new DType.Primitive(PType.I32, false),
 					new DType.Primitive(PType.F64, false),
 					new DType.Primitive(PType.F64, false),
@@ -70,42 +64,71 @@ class WriteFileSizeIntegrationTest {
 			false
 	);
 
+	private static final Schema JNI_SCHEMA = new Schema(List.of(
+			Field.notNullable("symbol", new ArrowType.Utf8()),
+			Field.notNullable("date", new ArrowType.Date(DateUnit.DAY)),
+			Field.notNullable("open", new ArrowType.FloatingPoint(FloatingPointPrecision.DOUBLE)),
+			Field.notNullable("high", new ArrowType.FloatingPoint(FloatingPointPrecision.DOUBLE)),
+			Field.notNullable("low", new ArrowType.FloatingPoint(FloatingPointPrecision.DOUBLE)),
+			Field.notNullable("close", new ArrowType.FloatingPoint(FloatingPointPrecision.DOUBLE)),
+			Field.notNullable("volume", new ArrowType.Int(64, true))
+	));
+
 	private static final Session SESSION = Session.create();
 	private static final BufferAllocator ALLOCATOR = ArrowAllocation.rootAllocator();
 
 	static {
 		NativeLoader.loadJni();
 	}
 
-
 	// ── Writers ───────────────────────────────────────────────────────────────
 
-	private static void writeJava(Path file, List<OhlcGenerator.OhlcBatch> batches) throws IOException {
+	private static Path writeCsv(Path dir, List<OhlcGenerator.OhlcBatch> batches) throws IOException {
+		Path file = dir.resolve("ohlc.csv");
+		try (BufferedWriter csv = Files.newBufferedWriter(file)) {
+			csv.write("symbol,date,open,high,low,close,volume\n");
+			for (OhlcGenerator.OhlcBatch b : batches) {
+				for (int i = 0; i < b.dates().length; i++) {
+					csv.write(b.symbols()[i] + "," + LocalDate.ofEpochDay(b.dates()[i]) + ","
+							+ b.open()[i] + "," + b.high()[i] + ","
+							+ b.low()[i] + "," + b.close()[i] + "," + b.volume()[i] + "\n");
+				}
+			}
+		}
+		return file;
+	}
+
+	private static Path writeJava(Path dir, List<OhlcGenerator.OhlcBatch> batches) throws IOException {
+		Path file = dir.resolve("ohlc-java.vtx");
 		try (FileChannel ch = FileChannel.open(file, StandardOpenOption.CREATE, StandardOpenOption.WRITE);
 		     VortexWriter writer = VortexWriter.create(ch, JAVA_SCHEMA, WriteOptions.cascading(3))) {
 			for (OhlcGenerator.OhlcBatch b : batches) {
 				writer.writeChunk(Map.of(
-						"date", b.dates(), "open", b.open(), "high", b.high(),
-						"low", b.low(), "close", b.close(), "volume", b.volume()
-				));
+						"symbol", b.symbols(), "date", b.dates(),
+						"open", b.open(), "high", b.high(),
+						"low", b.low(), "close", b.close(), "volume", b.volume()));
 			}
 		}
+		return file;
 	}
 
-	private static void writeJni(Path file, List<OhlcGenerator.OhlcBatch> batches) throws IOException {
+	private static Path writeJni(Path dir, List<OhlcGenerator.OhlcBatch> batches) throws IOException {
+		Path file = dir.resolve("ohlc-jni.vtx");
 		String uri = file.toAbsolutePath().toUri().toString();
 		try (dev.vortex.api.VortexWriter writer = dev.vortex.api.VortexWriter.create(
 				SESSION, uri, JNI_SCHEMA, new HashMap<>(), ALLOCATOR)) {
 			for (OhlcGenerator.OhlcBatch b : batches) {
 				try (VectorSchemaRoot root = VectorSchemaRoot.create(JNI_SCHEMA, ALLOCATOR)) {
-					DateDayVector dateVec  = (DateDayVector) root.getVector("date");
-					Float8Vector  openVec  = (Float8Vector)  root.getVector("open");
-					Float8Vector  highVec  = (Float8Vector)  root.getVector("high");
-					Float8Vector  lowVec   = (Float8Vector)  root.getVector("low");
-					Float8Vector  closeVec = (Float8Vector)  root.getVector("close");
-					BigIntVector  volVec   = (BigIntVector)  root.getVector("volume");
+					VarCharVector symbolVec = (VarCharVector) root.getVector("symbol");
+					DateDayVector dateVec   = (DateDayVector) root.getVector("date");
+					Float8Vector  openVec   = (Float8Vector)  root.getVector("open");
+					Float8Vector  highVec   = (Float8Vector)  root.getVector("high");
+					Float8Vector  lowVec    = (Float8Vector)  root.getVector("low");
+					Float8Vector  closeVec  = (Float8Vector)  root.getVector("close");
+					BigIntVector  volVec    = (BigIntVector)  root.getVector("volume");
 
 					int n = b.dates().length;
+					symbolVec.allocateNew();
 					dateVec.allocateNew(n);
 					openVec.allocateNew(n);
 					highVec.allocateNew(n);
@@ -114,6 +137,7 @@ private static void writeJni(Path file, List<OhlcGenerator.OhlcBatch> batches) t
 					volVec.allocateNew(n);
 
 					for (int i = 0; i < n; i++) {
+						symbolVec.setSafe(i, b.symbols()[i].getBytes(StandardCharsets.UTF_8));
 						dateVec.setSafe(i, b.dates()[i]);
 						openVec.setSafe(i, b.open()[i]);
 						highVec.setSafe(i, b.high()[i]);
@@ -131,77 +155,49 @@ private static void writeJni(Path file, List<OhlcGenerator.OhlcBatch> batches) t
 				}
 			}
 		}
+		return file;
 	}
 
-	// ── Tests ─────────────────────────────────────────────────────────────────
+	// ── Test ──────────────────────────────────────────────────────────────────
 
 	@Test
-	void javaFile_isReadable_withCorrectRowCount(@TempDir Path tmp) throws IOException {
+	void fileSizeComparison(@TempDir Path tmp) throws IOException {
 		// Given
-		Path file = tmp.resolve("ohlc-java.vtx");
 		List<OhlcGenerator.OhlcBatch> batches = OhlcGenerator.generate(TOTAL_ROWS, BATCH_SIZE);
 
 		// When
-		writeJava(file, batches);
+		Path csvFile  = writeCsv(tmp, batches);
+		Path jniFile  = writeJni(tmp, batches);
+		Path javaFile = writeJava(tmp, batches);
 
-		// Then — readable and correct row count
-		long totalRows = 0L;
-		try (VortexReader reader = VortexReader.open(file, EncodingRegistry.loadAll())) {
-			var iter = reader.scan(io.github.dfa1.vortex.scan.ScanOptions.columns("volume"));
-			while (iter.hasNext()) {
-				totalRows += iter.next().<LongArray>column("volume").length();
-			}
-		}
-		assertThat(totalRows).isEqualTo(TOTAL_ROWS);
-		System.out.printf("[WriteFileSizeIntegrationTest] Java file: %,d bytes (%.2f MB)%n",
-				Files.size(file), Files.size(file) / 1_048_576.0);
-	}
+		long csvSize  = Files.size(csvFile);
+		long jniSize  = Files.size(jniFile);
+		long javaSize = Files.size(javaFile);
 
-	@Test
-	void jniFile_isReadable_withCorrectRowCount(@TempDir Path tmp) throws IOException {
-		// Given
-		Path file = tmp.resolve("ohlc-jni.vtx");
-		List<OhlcGenerator.OhlcBatch> batches = OhlcGenerator.generate(TOTAL_ROWS, BATCH_SIZE);
+		// Then — report
+		System.out.printf(
+				"[FileSizeComparison] %,d rows  CSV=%,d bytes (%.1f MB)  JNI=%,d bytes (%.1f MB)  Java=%,d bytes (%.1f MB)  Java/JNI=%.2fx  CSV/Java=%.1fx%n",
+				TOTAL_ROWS,
+				csvSize,  csvSize  / 1_048_576.0,
+				jniSize,  jniSize  / 1_048_576.0,
+				javaSize, javaSize / 1_048_576.0,
+				(double) javaSize / jniSize,
+				(double) csvSize  / javaSize);
 
-		// When
-		writeJni(file, batches);
+		// Then — Java beats CSV
+		assertThat(javaSize).isLessThan(csvSize);
 
-		// Then — readable and correct row count
+		// Then — Java within 2x of JNI (both use cascading(3))
+		assertThat(javaSize).isLessThan(jniSize * 2);
+
+		// Then — Java file is readable with correct row count
 		long totalRows = 0L;
-		try (VortexReader reader = VortexReader.open(file, EncodingRegistry.loadAll())) {
-			var iter = reader.scan(io.github.dfa1.vortex.scan.ScanOptions.columns("close"));
+		try (VortexReader reader = VortexReader.open(javaFile, EncodingRegistry.loadAll())) {
+			var iter = reader.scan(io.github.dfa1.vortex.scan.ScanOptions.columns("volume"));
 			while (iter.hasNext()) {
-				totalRows += iter.next().<DoubleArray>column("close").length();
+				totalRows += iter.next().<LongArray>column("volume").length();
 			}
 		}
 		assertThat(totalRows).isEqualTo(TOTAL_ROWS);
 	}
-
-	@Test
-	void fileSizeComparison(@TempDir Path tmp) throws IOException {
-		// Given
-		Path javaFile = tmp.resolve("ohlc-java.vtx");
-		Path jniFile  = tmp.resolve("ohlc-jni.vtx");
-		List<OhlcGenerator.OhlcBatch> batches = OhlcGenerator.generate(TOTAL_ROWS, BATCH_SIZE);
-
-		// When
-		writeJava(javaFile, batches);
-		writeJni(jniFile, batches);
-
-		long javaSize = Files.size(javaFile);
-		long jniSize  = Files.size(jniFile);
-
-		// Then
-		System.out.printf("[WriteFileSizeIntegrationTest] %d rows: Java=%,d bytes (%.1f MB)  JNI=%,d bytes (%.1f MB)  ratio=%.2fx%n",
-				TOTAL_ROWS,
-				javaSize, javaSize / 1_048_576.0,
-				jniSize,  jniSize  / 1_048_576.0,
-				(double) javaSize / jniSize);
-
-		assertThat(javaSize).isPositive();
-		assertThat(jniSize).isPositive();
-		// Both writers use cascading(3): ALP+bitpacking for F64, delta+bitpacking for I64.
-		// Bound < 2× catches regressions while tolerating minor encoding strategy differences.
-		assertThat(javaSize).isLessThan(jniSize * 2);
-	}
 }