8316681: Rewrite URLEncoder.encode to use small reusable buffers

Reviewed-by: dfuchs, rriggs

src/java.base/share/classes/java/net/URLEncoder.java

@@ -26,8 +26,13 @@
 package java.net;
 
 import java.io.UnsupportedEncodingException;
-import java.io.CharArrayWriter;
+import java.nio.ByteBuffer;
+import java.nio.CharBuffer;
+import java.nio.charset.CharacterCodingException;
 import java.nio.charset.Charset;
+import java.nio.charset.CharsetEncoder;
+import java.nio.charset.CoderResult;
+import java.nio.charset.CodingErrorAction;
 import java.nio.charset.IllegalCharsetNameException;
 import java.nio.charset.UnsupportedCharsetException ;
 import java.util.BitSet;
@@ -138,11 +143,6 @@ public class URLEncoder {
         DEFAULT_ENCODING_NAME = StaticProperty.fileEncoding();
     }
 
-    private static void encodeByte(StringBuilder out, byte b) {
-        out.append('%');
-        HexFormat.of().withUpperCase().toHexDigits(out, b);
-    }
-
     /**
      * You can't call the constructor.
      */
@@ -205,6 +205,8 @@ public class URLEncoder {
         }
     }
 
+    private static final int ENCODING_CHUNK_SIZE = 8;
+
     /**
      * Translates a string into {@code application/x-www-form-urlencoded}
      * format using a specific {@linkplain Charset Charset}.
@@ -239,11 +241,16 @@ public class URLEncoder {
         }
 
         StringBuilder out = new StringBuilder(s.length() << 1);
-        CharArrayWriter charArrayWriter = new CharArrayWriter();
         if (i > 0) {
             out.append(s, 0, i);
         }
 
+        CharsetEncoder ce = charset.newEncoder()
+                .onMalformedInput(CodingErrorAction.REPLACE)
+                .onUnmappableCharacter(CodingErrorAction.REPLACE);
+        CharBuffer cb = CharBuffer.allocate(ENCODING_CHUNK_SIZE);
+        ByteBuffer bb = ByteBuffer.allocate((int)(ENCODING_CHUNK_SIZE * ce.maxBytesPerChar()));
+
         while (i < s.length()) {
             char c = s.charAt(i);
             if (DONT_NEED_ENCODING.test(c)) {
@@ -255,7 +262,7 @@ public class URLEncoder {
             } else {
                 // convert to external encoding before hex conversion
                 do {
-                    charArrayWriter.write(c);
+                    cb.put(c);
                     /*
                      * If this character represents the start of a Unicode
                      * surrogate pair, then pass in two characters. It's not
@@ -268,23 +275,63 @@ public class URLEncoder {
                         if ((i + 1) < s.length()) {
                             char d = s.charAt(i + 1);
                             if (Character.isLowSurrogate(d)) {
-                                charArrayWriter.write(d);
+                                cb.put(d);
                                 i++;
                             }
                         }
                     }
+                    // Limit to ENCODING_CHUNK_SIZE - 1 so that we can always fit in
+                    // a surrogate pair on the next iteration
+                    if (cb.position() >= ENCODING_CHUNK_SIZE - 1) {
+                        flushToStringBuilder(out, ce, cb, bb, false);
+                    }
                     i++;
                 } while (i < s.length() && !DONT_NEED_ENCODING.test((c = s.charAt(i))));
-
+                flushToStringBuilder(out, ce, cb, bb, true);
-                String str = charArrayWriter.toString();
-                byte[] ba = str.getBytes(charset);
-                for (byte b : ba) {
-                    encodeByte(out, b);
-                }
-                charArrayWriter.reset();
             }
         }
-
         return out.toString();
     }
+
+    /**
+     * Encodes input chars in {@code cb} and appends the byte values in an escaped
+     * format ({@code "%XX"}) to {@code out}. The temporary byte buffer, {@code bb},
+     * must be able to accept {@code cb.position() * ce.maxBytesPerChar()} bytes.
+     *
+     * @param out the StringBuilder to output encoded and escaped bytes to
+     * @param ce charset encoder. Will be reset if endOfInput is true
+     * @param cb input buffer, will be cleared
+     * @param bb output buffer, will be cleared
+     * @param endOfInput true if this is the last flush for an encoding chunk,
+     *                  to all bytes in ce is flushed to out and reset
+     */
+    private static void flushToStringBuilder(StringBuilder out,
+                                             CharsetEncoder ce,
+                                             CharBuffer cb,
+                                             ByteBuffer bb,
+                                             boolean endOfInput) {
+        cb.flip();
+        try {
+            CoderResult cr = ce.encode(cb, bb, endOfInput);
+            if (!cr.isUnderflow())
+                cr.throwException();
+            if (endOfInput) {
+                cr = ce.flush(bb);
+                if (!cr.isUnderflow())
+                    cr.throwException();
+                ce.reset();
+            }
+        } catch (CharacterCodingException x) {
+            throw new Error(x); // Can't happen
+        }
+        HexFormat hex = HexFormat.of().withUpperCase();
+        byte[] bytes = bb.array();
+        int len = bb.position();
+        for (int i = 0; i < len; i++) {
+            out.append('%');
+            hex.toHexDigits(out, bytes[i]);
+        }
+        cb.clear();
+        bb.clear();
+    }
 }

src/java.base/share/classes/java/util/HexFormat.java

@@ -679,9 +679,8 @@ public final class HexFormat {
      * @throws UncheckedIOException if an I/O exception occurs appending to the output
      */
     public <A extends Appendable> A toHexDigits(A out, byte value) {
-        Objects.requireNonNull(out, "out");
         try {
-            out.append(toHighHexDigit(value));
+            out.append(toHighHexDigit(value)); // implicit null-check
             out.append(toLowHexDigit(value));
             return out;
         } catch (IOException ioe) {

test/jdk/java/net/URLEncoder/SurrogatePairs.java

@@ -58,6 +58,9 @@ public class SurrogatePairs {
                 {"\uDBFF\uDC001", "%F4%8F%B0%801"},
                 {"\uDBFF\uDC00@", "%F4%8F%B0%80%40"},
                 {"\u0101\uDBFF\uDC00", "%C4%81%F4%8F%B0%80"},
+                {"\u0101\u0101\u0101\u0101\u0101\u0101\uDBFF\uDC00\u0101", "%C4%81%C4%81%C4%81%C4%81%C4%81%C4%81%F4%8F%B0%80%C4%81"},
+                {"\u0101\u0101\u0101\u0101\u0101\u0101\u0101\uDBFF\uDC00\u0101", "%C4%81%C4%81%C4%81%C4%81%C4%81%C4%81%C4%81%F4%8F%B0%80%C4%81"},
+                {"\u0101\u0101\u0101\u0101\u0101\u0101\u0101\u0101\uDBFF\uDC00\u0101", "%C4%81%C4%81%C4%81%C4%81%C4%81%C4%81%C4%81%C4%81%F4%8F%B0%80%C4%81"},
                 {"\uDBFF\uDC00\u0101", "%F4%8F%B0%80%C4%81"},
                 {"\uDE0A\uD83D", "%3F%3F"},
                 {"1\uDE0A\uD83D", "1%3F%3F"},

test/micro/org/openjdk/bench/java/net/URLEncodeDecode.java

@@ -190,4 +190,18 @@ public class URLEncodeDecode {
     }
 
 
+    @Benchmark
+    public void testEncodeLatin1(Blackhole bh) throws UnsupportedEncodingException {
+        for (String s : testStringsEncode) {
+            bh.consume(java.net.URLEncoder.encode(s, StandardCharsets.ISO_8859_1));
+        }
+    }
+
+    @Benchmark
+    public void testDecodeLatin1(Blackhole bh) throws UnsupportedEncodingException {
+        for (String s : testStringsDecode) {
+            bh.consume(URLDecoder.decode(s, StandardCharsets.ISO_8859_1));
+        }
+    }
+
 }