TETRAEDGE: Encode text before draw #4924
Conversation
| @@ -145,9 160,9 @@ void TeFont3::draw(TeImage &destImage, const Common::String &str, int fontSize, | |||
|
|
|||
| uint32 uintcol = ((uint32)col.a() << fmt.aShift) | ((uint32)(col.r()) << fmt.rShift) | |||
| | ((uint32)(col.g()) << fmt.gShift) | ((uint32)(col.b()) << fmt.bShift); | |||
| Common::String line(str); | |||
| Common::U32String line = str.decode(_codePage); | |||
| if (g_engine->getCore()->language() == "he") | |||
There was a problem hiding this comment.
This check (that the language is "he") actually became redundant with the changes in this PR, but can be left in for the sake of making sure it cannot affect other languages
|
Nice, thanks for that - I'd been meaning to work out why the RU translation wasn't working. There is an ugly |
engines/tetraedge/te/te_font3.h
Outdated
| @@ -86,7 86,9 @@ class TeFont3 : public TeResource { | |||
| private: | |||
| void init(); | |||
| Graphics::Font *getAtSize(uint size); | |||
| Common::CodePage codePage(); | |||
There was a problem hiding this comment.
as a nit, I guess this should be static or at least const
There was a problem hiding this comment.
Thank you, I think it can't be static because it uses g_engine.
const on the return type is a warning (Common::CodePage is enum)
added const that marks that the instance does not change, but it might be relevant to more methods here.
please note that this method is currently used once in the init.
There was a problem hiding this comment.
I could have been more clear - yep, setting the function const is what I meant, which means it won't change the contents of the object, not return const.
I think it could also be static which in c has a bit of a different meaning to Java in that it just means "does not depend on which instance of this class it's called on".
Since g_engine is a global rather than a class member it's also be fine to make this function static, but I haven't been super consistent about that myself so const is also fine as it's not performance-critical code.
Thanks again for the fix!
This fixes an issue in the text drawing where the glyphs to use are implicitly casted to uint32 and does not get the current glyph range in the font.
I understand that implicit conversion from 8-bit encoding as-is is similiar to Latin-1 encoding, please correct me if it's not the case... I didn't notice problem with these languages but I'm not a native speaker.
Specifically, it fixes russian translation and allow better integration in the hebrew fan translation...
before:
after:
(Also not a native speaker, but it seems to match the voice over)
Thank you