Number parsing (#14)

Author: piotrrzysko

.gitignore

@@ -2,3 +2,4 @@
 .gradle
 build
 profilers
+testdata

README.md

@@ -8,7 +8,6 @@ This implementation is still missing several features available in simdsjon. For
 
 * Support for Unicode characters
 * UTF-8 validation
-* Full support for parsing floats
 * Support for 512-bit vectors
 
 ## Code Sample

build.gradle

@@ -1,10 +1,12 @@
 import me.champeau.jmh.JmhBytecodeGeneratorTask
 import org.gradle.internal.os.OperatingSystem
+import org.ajoberstar.grgit.Grgit
 
 plugins {
     id 'java'
     id 'scala'
     id 'me.champeau.jmh' version '0.7.1'
+    id 'org.ajoberstar.grgit' version '5.2.0'
 }
 
 group = 'org.simdjson'
@@ -37,7 +39,22 @@ dependencies {
     testRuntimeOnly group: 'org.junit.jupiter', name: 'junit-jupiter-engine', version: junitVersion
 }
 
+tasks.register('downloadTestData') {
+    doFirst {
+        def testDataDir = new File("${project.projectDir.getAbsolutePath()}/testdata")
+        if (!testDataDir.exists()) {
+            testDataDir.mkdir()
+        }
+        def numbersTestDataDir = new File("${testDataDir}/parse-number-fxx-test-data")
+        if (!numbersTestDataDir.exists()) {
+            def grgit = Grgit.clone(dir: numbersTestDataDir, uri: 'https://github.com/nigeltao/parse-number-fxx-test-data.git')
+            grgit.close()
+        }
+    }
+}
+
 test {
+    dependsOn downloadTestData
     useJUnitPlatform()
     jvmArgs += [
             '--add-modules', 'jdk.incubator.vector',

src/jmh/java/org/simdjson/NumberParserBenchmark.java

@@ -0,0 +1,49 @@
+package org.simdjson;
+
+import org.openjdk.jmh.annotations.Benchmark;
+import org.openjdk.jmh.annotations.BenchmarkMode;
+import org.openjdk.jmh.annotations.Level;
+import org.openjdk.jmh.annotations.Mode;
+import org.openjdk.jmh.annotations.OutputTimeUnit;
+import org.openjdk.jmh.annotations.Param;
+import org.openjdk.jmh.annotations.Scope;
+import org.openjdk.jmh.annotations.Setup;
+import org.openjdk.jmh.annotations.State;
+
+import java.io.IOException;
+import java.util.concurrent.TimeUnit;
+
+import static org.simdjson.SimdJsonPaddingUtil.padWithSpaces;
+
+@State(Scope.Benchmark)
+@BenchmarkMode(Mode.Throughput)
+@OutputTimeUnit(TimeUnit.SECONDS)
+public class NumberParserBenchmark {
+
+    private final Tape tape = new Tape(100);
+    private final NumberParser numberParser = new NumberParser(tape);
+
+    @Param({
+            "2.2250738585072013e-308", // fast path
+            "1.00000000000000188558920870223463870174566020691753515394643550663070558368373221972569761144603605635692374830246134201063722058e-309" // slow path
+    })
+    String number;
+    byte[] numberUtf8Bytes;
+
+    @Setup(Level.Trial)
+    public void setup() throws IOException {
+        numberUtf8Bytes = padWithSpaces(number);
+    }
+
+    @Benchmark
+    public double baseline() {
+        return Double.parseDouble(number);
+    }
+
+    @Benchmark
+    public double simdjson() {
+        tape.reset();
+        numberParser.parseNumber(numberUtf8Bytes, 0);
+        return tape.getDouble(0);
+    }
+}

src/jmh/java/org/simdjson/SimdJsonPaddingUtil.java

@@ -1,10 +1,22 @@
 package org.simdjson;
 
+import java.util.Arrays;
+
+import static java.nio.charset.StandardCharsets.UTF_8;
+
 class SimdJsonPaddingUtil {
 
     static byte[] padded(byte[] src) {
         byte[] bufferPadded = new byte[src.length + 64];
         System.arraycopy(src, 0, bufferPadded, 0, src.length);
         return bufferPadded;
     }
+
+    static byte[] padWithSpaces(String str) {
+        byte[] strBytes = str.getBytes(UTF_8);
+        byte[] padded = new byte[strBytes.length + 64];
+        Arrays.fill(padded, (byte) ' ');
+        System.arraycopy(strBytes, 0, padded, 0, strBytes.length);
+        return padded;
+    }
 }

src/main/java/org/simdjson/JsonValue.java

@@ -112,6 +112,9 @@ public String toString() {
             case INT64 -> {
                 return String.valueOf(asLong());
             }
+            case DOUBLE -> {
+                return String.valueOf(asDouble());
+            }
             case TRUE_VALUE, FALSE_VALUE -> {
                 return String.valueOf(asBoolean());
             }