[⚠️HELP] Problem displaying French Accentuated or Special Characters (Mojibake encoding mismatch?)

[NSA]

Hello,

I am using Syside Modeler: SysML v2 Professional

sensmetry.syside-modeler Version 0.10.3

I have some text in French language including accentuated characters (e.g. “décrire”).

I tried different encoding (UTF-8, BOM, 1252)

the correct character is displayed in the textual notation, but not the visualizer of SysIDE which I assume based, behind the hood, on tomsawyer sysml-v2-webview.

Please help fix the problem. Attached are some snapshots as well as mini-sysml example including letter “é”.

NB: This is applicable to any file with french characters in its documentation.

ViewTest.sysml (718 Bytes)

According to Gemini Flash 3.6:

The Encoding Conflict

The SysIDE extension consists of two engines that do not use the same encoding:

The parser (SysIDE Modeler): It validates the file’s SysML v2 syntax. It strictly requires UTF-8. If it encounters a Windows-1252 file, it crashes immediately with the decoding error you received.
The graphics engine (Tom Sawyer): It renders the diagram. It ignores UTF-8 and reads the file using your French Windows system’s default encoding (which is Windows-1252). That is why it displays characters like “é” instead of accented characters.
This creates a technical impasse: if you provide UTF-8, the graphics engine becomes corrupted; if you provide Windows-1252, the parser crashes.

Workarounds that did NOT work :

• Use Unicode Escapes in String Literals: Replace accented characters with standard 4-digit Hex Unicode escape sequences directly inside doc comments/strings.

  • Example: Replace décrire with d\u00E9crire or A D\u00E9crire.

• HTML/XML Entity Encoding: If SysIDE passes documentation blocks to an SVG/HTML rendering node inside Tom Sawyer Software’s backend, numeric character references often bypass the encoding pipeline cleanly.

  • Example: A Décrire or A Décrire.

shot1.sysml (185 Bytes)

[Simas]

Hi nsalhab,

This seems to be an issue with the Tom Sawyer visualization; I’ve forwarded it to the dev team.

As a workaround, you can use the Modeler CLI, which uses a new visualization engine and does not have this issue.

[NSA]

Hello,

Thank you very much for reply.

To help the community, I managed to solve the problem using the following workaround forcing the visualizer to use UTF-8 encoding

The Sensmetry SysIDE extension has a specific setting to pass arguments to its backend language server.

1. Open your IDE Settings (usually Ctrl + ,).

2. Search for SysIDE.

3. Look for a setting related to JVM Arguments, Backend Arguments, or Encoding.

4. If there is an arguments field, add "-Dfile.encoding=UTF-8" to it and restart the extension.