Strange unicode issue in JavaDoc for parameters

[michael-browne]

In the JavaDoc for a controller the text descriptions for the parameters are encoded in the \\uxxxxx format.

Here's a sample from the json:

"comment":"テスト","parameters":{"from":"\\u30c6\\u30b9\\u30c8"}

and the original JavaDoc:

/**
 * テスト
 * @param from テスト
 * @return テスト
 */

As you can see the comment is encoded correctly, just the parameters descriptions are affected.

The final HTML files ends up looking like this:

Does anyone know what is causing this?
Thanks

[fbenz]

Which Java version are you using? The Doclet API changed with Java 9 and we thus have two different implementations.

For Java 8, we are not doing any transformation and I just ran the test with the characters and it works:

...
  "methods": {
    "initiate": {
      "comment": "Initiates request テスト",
      "parameters": {
        "force": "true if force",
        "when": "when to initiate テスト"
      },
...

[michael-browne]

I'm using Java 11. Do I need to do something different?

[michael-browne]

Sorry for the late reply, but I just checked this again and I can confirm I'm using the JDK 9+ JSON Doclet with Java 11.

Both 2.0.5 and 2.0.6 produce the same results.