The text is defined with the pipe as line separator, but we need
to replace it with a space to form a proper sentence before we
send the text to the TTS, otherwise we get unnatural pauses and
some incorrect text.
This was a regression from 54b0b4ac4 that prevents from adding
null characters to a Common::String. For those screens the game
passes multiple strings as a single strings containing a '\0'
to separate the different parts. But that '\0' was no longer
added. I am now using '\1' and updated the parsing in the text
display to handle this.
Add constant for inverting sections (number 128)
Move some initialization in initState from derived classes to
base class.
Simpilfy AxacussCorridor4::onEntrance
I am merging the second engine to the first one. Both should be
fully functional as before right now
Current merge file status:
console: should be done
graphics: should be done
detection: 0% merged
imageid: appears to not be used anywhere, so it may be removed
resman: partialy merged
rooms: totaly different, 0% merged
screen: should be done
screenstatic: done
sound: partialy done
state: a lot different, just started to merge
supernova: mostly done
Although renderImage for ImageId doesn't use _currentImage, it needs to
be set for compatibility with the section based renderImage function.
The functionality doesn't change with this commit, except that the
palette brightness gets updated if the image is in a different file.
Although SupernovaEngine forwards render calls to screen, it also
exposed members publicly like _brightness. Therefore, a few changes were
necessary in rooms and state.
