search

k-omura / truetype_Arduino

display truetype font for Arduino
MIT License
40 stars 14 forks source link

utf8 input #5

Closed k-omura closed 3 years ago

k-omura commented 3 years ago

Suggestion from Greg in #4 

I also have an old algo for translating utf8 to unicode that I used on your old code. I can drop it in here if you like.
utf8 is especially good for European languages and data transfer. Shorter strings and a consistent single byte stream to read.
or you could leave it as it is and as I did
the user can translate the string to a wchar string encoding the string themselves.
I'll hack out a simple example Arduino example once I have time to start working with the code.
That is probably the best way I can help you.
My cpp is very limited I work and think mostly in C.
Yardie- commented 3 years ago

I will test it but having looked at your code as you said it's already there in what you have done maybe this will just work

void truetypeClass::string(uint16_t _x, uint16_t _y, const uint8_t _character[]){
  this->string(_x, _y, (wchar_t*)_character);
}

or just calling the function like truetype.string(10, 10, L"L'année 1866 fut marquée par un événement bizarre,");

I'll try that.

Yardie- commented 3 years ago

Cool the code you sent works nicely with alex brush. 

String test = "èàùëïü";
    //Write a string to the framebuffer
    truetype.string(10, 0, test);
    truetype.string(10, 10, L"èàùëïü");
    truetype.string(10, 20, L"L'année");

All works but another font https://www.1001fonts.com/kg-les-bouquinistes-de-paris-font.html blows out the heap. with

3360
/FONTFILE.ttf
68400
CORRUPT HEAP: Bad head at 0x3ffb2d24. Expected 0xabba1234 got 0x3ffb6380
abort() was called at PC 0x400832f1 on core 1

ELF file SHA256: 0000000000000000

Backtrace: 0x40085004:0x3ffb1a70 0x40085279:0x3ffb1a90 0x400832f1:0x3ffb1ab0 0x4008341d:0x3ffb1ae0 0x400da493:0x3ffb1b00 0x400d618d:0x3ffb1dc0 0x400d60ac:0x3ffb1e10 0x400893e1:0x3ffb1e40 0x40081546:0x3ffb1e60 0x400831e9:0x3ffb1e80 0x4000bec7:0x3ffb1ea0 0x400d11f9:0x3ffb1ec0 0x400d2366:0x3ffb1ee0 0x400d1121:0x3ffb1f30 0x400d3cfa:0x3ffb1fb0 0x40086289:0x3ffb1fd0

But I can work with this thanks heaps. I'll let you enjoy the rest of your holiday in peace.

Yardie- commented 3 years ago

but https://www.1001fonts.com/chopin-script-font.html works fine I guess you can't be expected to support all the fonts in the world :)

k-omura commented 3 years ago

That's right I have confirmed that some font files have bugs. It's one of the to-do lists.

Thanks for letting me know what the buggy font files have in common! If the font file isn't available for some reason, I think I should give feedback to the user.

I can only touch my favorite code on holidays, so it's rather welcome! I want to stay home while fixing the code

Yardie- commented 3 years ago

I like to try different fonts out so I will let you know what ones do what things as I find them. Fonts are notoriously variable. That is why I like that site. There are so many different typesa. It helps me test lots of different things.

It would be good to at least reject a font rather than crash the code. It seems that is quite tricky.

k-omura commented 3 years ago

https://www.1001fonts.com/kg-les-bouquinistes-de-paris-font.html blows out the heap. with

I checked the above font. It turns out that the cause is that it does not currently support characters that are a combination of multiple glyphs called "Compound glyphs". https://developer.apple.com/fonts/TrueType-Reference-Manual/RM01/Chap1.html#compound

Supporting "Compound glyphs" is one of the future issues, but at the moment, in the case of "Compound glyphs", it is left blank to prevent it from crashing.

I recognize that "Compound glyphs" are especially important in European languages.

Yardie- commented 3 years ago

Well done that is fantastic and the perfect response to the problem. Yep ok that's a problem stop the code crashing in a nice clean way not doing that part and moving on. The API can clearly now say one day these types will be supported but not now. I am sure many of the people who require compound glyphs understand that it is not trivial. You are getting a deep understanding of how ttf works now. Well done again.

That is ground many fear to tread.

Yardie- commented 3 years ago

You are welcome to close this if you like.

Yardie- commented 3 years ago

Actually it maybe worth another look.

http://utf8everywhere.org/ Although (wchar_t*)_character seems to work so far it might be better to somehow use the algorithms in stringToWchar

k-omura commented 3 years ago

This issue is closed because it supports single-byte characters. Certainly it may be the "right way" to explicitly convert to utf16...

The "Compound glyphs" mentioned in this issue will be supported in #13.