use UTF-8 for readString() and write() in net client

benfry · benfry · commit aab8d8fa50ef · 2022-01-10T19:04:38.000-05:00
diff --git a/java/libraries/net/src/processing/net/Client.java b/java/libraries/net/src/processing/net/Client.java
@@ -29,6 +29,7 @@
 import java.io.*;
 import java.lang.reflect.*;
 import java.net.*;
+import java.nio.charset.StandardCharsets;
 
 /**
    * 
@@ -596,11 +597,10 @@ public int readBytesUntil(int interesting, byte byteBuffer[]) {
 
   /**
    * 
-   * Returns the all the data from the buffer as a <b>String</b>. This method 
-   * assumes the incoming characters are ASCII. If you want to transfer 
-   * Unicode data, first convert the String to a byte stream in the 
-   * representation of your choice (i.e. UTF8 or two-byte Unicode data), and 
-   * send it as a byte array.
+   * Returns the all the data from the buffer as a <b>String</b>.
+   *
+   * In 4.0 beta 3, changed to using UTF-8 as the encoding,
+   * otherwise the behavior is platform-dependent.
    * 
    * @webref client
    * @usage application
@@ -609,7 +609,7 @@ public int readBytesUntil(int interesting, byte byteBuffer[]) {
   public String readString() {
     byte b[] = readBytes();
     if (b == null) return null;
-    return new String(b);
+    return new String(b, StandardCharsets.UTF_8);
   }
 
 
@@ -619,10 +619,8 @@ public String readString() {
    * <b>null</b> if it doesn't find what you're looking for.
    * 
    * <h3>Advanced</h3>
-   * <p/>
-   * If you want to move Unicode data, you can first convert the
-   * String to a byte stream in the representation of your choice
-   * (i.e. UTF8 or two-byte Unicode data), and send it as a byte array.
+   * In 4.0 beta 3, changed to using UTF-8 as the encoding,
+   * otherwise the behavior is platform-dependent.
    * 
    * @webref client
    * @usage application
@@ -632,7 +630,7 @@ public String readString() {
   public String readStringUntil(int interesting) {
     byte b[] = readBytesUntil(interesting);
     if (b == null) return null;
-    return new String(b);
+    return new String(b, StandardCharsets.UTF_8);
   }
 
 
@@ -679,20 +677,11 @@ public void write(byte data[]) {
 
 
   /**
-   * <h3>Advanced</h3>
-   * Write a String to the output. Note that this doesn't account
-   * for Unicode (two bytes per char), nor will it send UTF8
-   * characters.. It assumes that you mean to send a byte buffer
-   * (most often the case for networking and serial i/o) and
-   * will only use the bottom 8 bits of each char in the string.
-   * (Meaning that internally it uses String.getBytes)
-   *
-   * If you want to move Unicode data, you can first convert the
-   * String to a byte stream in the representation of your choice
-   * (i.e. UTF8 or two-byte Unicode data), and send it as a byte array.
+   * In 4.0 beta 3, changed to using UTF-8 as the encoding,
+   * otherwise the behavior is platform-dependent.
    */
   public void write(String data) {
-    write(data.getBytes());
+    write(data.getBytes(StandardCharsets.UTF_8));
   }
 
 
diff --git a/todo.txt b/todo.txt
@@ -12,6 +12,9 @@ X   skip fonts starting . and # because they're likely to confuse users
 X get rid of version numbers in the name of the batik.jar file
 X ffmpeg not downloading correctly on M1 machines
 X   https://github.com/processing/processing4/issues/319
+X use UTF-8 for readString() and write() in net client
+X   avoids platform-specific behavior; Java 18 also making UTF-8 the default
+X   https://github.com/processing/processing4/issues/336
 
 sam
 X Error when calling smooth() on PGraphics
@@ -30,8 +33,6 @@ before release
 _ disable Theme Engine from Tools (in build.xml)
 
 
-
-
 before beta 3
 _ Add support for localizing contributions
 _   https://github.com/processing/processing4/pull/237 (updated by Andres)
