Javadoc view respect javadoc file encoding

[barulheira]

Environment:

• VS Code Version: 1.99
• Extension Version: 0.29.0
• JDK Type & Version: OpenJDK 21
• OS Type & Version: FreeBSD 14.2

Step to Reproduce:
Hover on any Java class name or method which has javadoc attached.
Do this on a project whose file encoding is anything other than UTF-8, and where the Java class or method belongs to another project whose javadoc was generated as UTF-8.

Expected Behavior:
Read the javadoc contents according to the javadoc file encoding, not according to the current project file encoding.
The text is displayed consistently.

Actual Behavior:
The javadoc contents are read with the current project file encoding instead of the javadoc file encoding. If they don't match, the text appears garbled (when containing non-ASCII characters).

Related Issues:

[github-actions]

Hi @barulheira, I'm an AI Support assistant here to help with your issue. While the team reviews your request, I wanted to provide some possible tips and documentation that might help you in the meantime.

• There’s an open jdt.ls issue for exactly this problem – no OOTB support for per-file javadoc encoding. It suggests using ICU4J to sniff the charset before loading the HTML:
  • Encoding issue with jdt.ls javadoc: Encoding issue with jdt.ls javadoc eclipse-jdtls/eclipse.jdt.ls#524 (comment)

• Workaround: configure VS Code’s files.encoding for HTML (javadoc) files. In your settings.json:

  "files.encoding": {
    "*.html": "utf8",     // or whatever encoding the javadoc uses
    "*.java": "${default}"  
  }

  This forces the editor to decode Javadoc HTML with the right charset.

• If you need a permanent fix, please upvote or comment on the above jdt.ls issue to request a dedicated Java extension setting for Javadoc file encoding.

The team will respond to your issue shortly. I hope these suggestions are helpful in the meantime. If this comment helped you, please give it a 👍. If the suggestion was not helpful or incorrect, please give it a 👎. Your feedback helps us improve!

[barulheira]

We don't need to sniff the encoding when we know which it is. When reading HTML files, the encoding is set within the file itself. When reading source files, we do often know the file encoding of the project, i. e. within the VS Code workspace or in the pom.xml.