Is it possible to have a string constant (output datatype: string) generated in UTF-8?

"Simulink strings support 256 characters of the ISO/IEC 8859-1 character set" according to Simulink Strings in the Help Center.

I've created a "string constant" block called "String ãºç" in a simulink model where I've inserted the value "exampleàñçº".

The block's name is exported correctly in UTF-8 ("String ãºç"), but what I get in the exported file for the value is "example\340\361\347\272" (octal values for extended characters in ISO/IEC 8859-1).

I was wondering if there was a way to export the string value in UTF-8.

Answer 1

[Walter Roberson]

No, you cannot do this. Strings are exported as plain C / C++ character vectors, and plain C / C++ character vectors are 8 bit characters. There is no way in C / C++ to mark strings as being UTF-8. The closest you can come with constant char vectors is to have a sequence of bytes that happens to encode UTF-8

Unfortunately std::wstring uses 16 bit code points on Windows. For portability you need to use std:u32string

https://stackoverflow.com/questions/50403342/how-do-i-properly-use-stdstring-on-utf-8-in-c

https://en.cppreference.com/w/cpp/locale/codecvt_utf8

[Eduardo]

Thank you for the answer! In the end, we ended up encoding the generated character vector ourselves.

Answer 2

[Mathy]

Hi Eduardo,

When saving Simulink models, you can ensure that the string data is stored in UTF-8 format. Here’s how:

Save Models in SLX Format: Starting from R2023a release Simulink saves models in the SLX format by default. This format uses Unicode UTF-8 in XML and other international formats. It typically reduces file size compared to MDL files and enables incremental loading and saving.

You can learn more about this by clicking on the link below: https://www.mathworks.com/help/simulink/ug/save-models.html

Hope this helps!

[Eduardo]

Thank you for the answer! However, it unfortunately doesn't solve our issue...

Sorry for the misleading question and poor explanation of our issue.

Our issue is with code generation for C/C++.