不幸的是，Gson似乎不支持它。所有JSON输入/输出分别集中在Gson（从2.8.0开始）
JsonReader和
JsonWriter。
JsonReader可以使用其私有
readEscapeCharacter方法读取Unipre转义。但是，与有所不同的是
JsonReader，
JsonWriter只需将字符串写入后备
Writer实例，除了
u2028和以外，不对127以上的字符进行字符校正
u2029。可能唯一可以做的是编写自定义转义符，
Writer以便发出Unipre转义。
final class EscapedWriter extends Writer { private static final char[] hex = { '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 'a', 'b', 'c', 'd', 'e', 'f' }; private final Writer writer; // I/O components are usually implemented in not thread-safe manner // so we can save some time on constructing a single UTF-16 escape private final char[] escape = { '\', 'u', 0, 0, 0, 0 }; EscapedWriter(final Writer writer) { this.writer = writer; } // This implementation is not very efficient and is open for enhancements: // * constructing a single "normalized" buffer character array so that it could be passed to the downstream writer // rather than writing characters one by one // * etc... @Override public void write(final char[] buffer, final int offset, final int length) throws IOException { for ( int i = offset; i < length; i++ ) { final int ch = buffer[i]; if ( ch < 128 ) { writer.write(ch); } else { escape[2] = hex[(ch & 0xF000) >> 12]; escape[3] = hex[(ch & 0x0F00) >> 8]; escape[4] = hex[(ch & 0x00F0) >> 4]; escape[5] = hex[ch & 0x000F]; writer.write(escape); } } } @Override public void flush() throws IOException { writer.flush(); } @Override public void close() throws IOException { writer.close(); } // Some java.io.Writer subclasses may use java.lang.Object.toString() to materialize their accumulated state by design // so it has to be overridden and forwarded as well @Override public String toString() { return writer.toString(); }}该作家未经充分测试，不尊重
u2028和
u2029。然后只需在调用
toJson方法时配置输出目标即可：
final String input = "{"description":"Tikrovi\u0161kai para\u0161ytas k\u016brinys"}";final Book book = gson.fromJson(input, Book.class);final Writer output = new EscapedWriter(new StringWriter());gson.toJson(book, output);System.out.println(input);System.out.println(output);输出：
{“描述”：“ Tikrovi u0161ytas k u016brinys”}
这是一个有趣的问题，您可能还会在google /
gson上提出一个问题，以添加字符串编写配置选项-
或至少从开发团队那里获得一些意见。我确实相信他们非常了解这种行为，并通过设计使它像这样工作，但是他们也可以对此有所了解（我现在唯一想到的是，他们目前有更多的性能，无法再增加在编写字符串之前进行转换，但这是一个微不足道的猜测）。
