optimize isASCII

alibaba · Jan 22, 2025 · 0451d5c · 0451d5c
1 parent e4454ef
commit 0451d5c
Show file tree

Hide file tree

Showing 3 changed files with 22 additions and 25 deletions.
diff --git a/benchmark/src/main/java/com/alibaba/fastjson2/benchmark/BytesAsciiCheck.java b/benchmark/src/main/java/com/alibaba/fastjson2/benchmark/BytesAsciiCheck.java
@@ -70,7 +70,7 @@ public void direct(Blackhole bh) throws Throwable {
         bh.consume(hasNegatives(bytes, 0, bytes.length));
     }
 
-    @Benchmark
+//    @Benchmark
     public void isASCII(Blackhole bh) throws Throwable {
         bh.consume(com.alibaba.fastjson2.util.IOUtils.isASCII(bytes, 0, bytes.length));
     }
@@ -80,27 +80,27 @@ public void isLatin1(Blackhole bh) throws Throwable {
         bh.consume(com.alibaba.fastjson2.util.IOUtils.isLatin1(chars, 0, chars.length));
     }
 
-    @Benchmark
+//    @Benchmark
     public void isASCIIJDK(Blackhole bh) throws Throwable {
         bh.consume(com.alibaba.fastjson2.util.JDKUtils.PREDICATE_IS_ASCII.test(bytes));
     }
 
-    @Benchmark
+//    @Benchmark
     public void indexOfSlash(Blackhole bh) throws Throwable {
         bh.consume(com.alibaba.fastjson2.util.IOUtils.indexOfSlash(bytes, 0, bytes.length));
     }
 
-    @Benchmark
+//    @Benchmark
     public void indexOfSlashV(Blackhole bh) throws Throwable {
         bh.consume(com.alibaba.fastjson2.util.IOUtils.indexOfSlashV(bytes, 0, bytes.length));
     }
 
-    @Benchmark
+//    @Benchmark
     public void indexOfChar(Blackhole bh) throws Throwable {
         bh.consume(indexOfChar(bytes, '\'', 0, bytes.length));
     }
 
-    @Benchmark
+//    @Benchmark
     public void indexOfString(Blackhole bh) throws Throwable {
         bh.consume(str.indexOf('\\'));
     }

diff --git a/benchmark/src/test/java/com/alibaba/fastjson2/benchmark/BytesAsciiCheckTest.java b/benchmark/src/test/java/com/alibaba/fastjson2/benchmark/BytesAsciiCheckTest.java
@@ -60,11 +60,11 @@ public static void direct() throws Throwable {
         }
     }
 
-    public static void isASCII_chars() throws Throwable {
+    public static void isLatin1() throws Throwable {
         for (int j = 0; j < 5; j++) {
             long start = System.currentTimeMillis();
             for (int i = 0; i < LOOP_COUNT; ++i) {
-                benchmark.isASCII_chars(BH);
+                benchmark.isLatin1(BH);
             }
             long millis = System.currentTimeMillis() - start;
             System.out.println("BytesAsciiCheck-isASCII_chars : " + millis);
@@ -88,7 +88,7 @@ public static void main(String[] args) throws Throwable {
 //        handler();
 //        lambda();
 //        direct();
-        isASCII_chars();
+        isLatin1();
 //        isASCII();
     }
 }
diff --git a/core/src/main/java/com/alibaba/fastjson2/util/IOUtils.java b/core/src/main/java/com/alibaba/fastjson2/util/IOUtils.java
@@ -1872,36 +1872,33 @@ static short convEndian(boolean big, short n) {
     public static boolean isLatin1(char[] chars, int off, int len) {
         int upperBound = off + (len & ~7);
         int end = off + len;
-        long address = ARRAY_BYTE_BASE_OFFSET + ((long) off << 1);
-        while (off < upperBound
-                && (convEndian(false, UNSAFE.getLong(chars, address) | UNSAFE.getLong(chars, address + 8)) & 0xFF00FF00FF00FF00L) == 0
-        ) {
+        long address = ARRAY_CHAR_BASE_OFFSET + off;
+        long value = 0;
+        while (off < upperBound) {
+            value |= UNSAFE.getLong(chars, address) | UNSAFE.getLong(chars, address + 8);
             address += 16;
             off += 8;
         }
-
-        while (off < end) {
-            if (chars[off++] > 0xFF) {
-                return false;
-            }
+        while (off++ < end) {
+            value |= UNSAFE.getShort(chars, address);
+            address += 2;
         }
-        return true;
+        return (convEndian(false, value) & 0xFF00FF00FF00FF00L) == 0;
     }
 
     public static boolean isASCII(byte[] bytes, int off, int len) {
         int upperBound = off + (len & ~7);
         int end = off + len;
         long address = ARRAY_BYTE_BASE_OFFSET + off;
-        while (off < upperBound && (UNSAFE.getLong(bytes, address) & 0x8080808080808080L) == 0) {
+        long value = 0;
+        while (off < upperBound) {
+            value |= UNSAFE.getLong(bytes, address);
             address += 8;
             off += 8;
         }
-
         while (off < end) {
-            if ((bytes[off++] & 0x80) != 0) {
-                return false;
-            }
+            value |= bytes[off++];
         }
-        return true;
+        return (value & 0x8080808080808080L) == 0;
     }
 }