LilyPond warning: no glyph for character U+EFFFFFFF in font (Eclipse 2019-09 R)

[Viktor Velthuijs]

Dear David,

I am getting a strange error trying to parse lyrics containing special characters like the umlaut (e.g. 'ö').

The following errors get printed:

Warning: Invalid 4-byte UTF8 found in input of graph compositionModel -
treated as Latin-1. Perhaps "-Gcharset=latin1" is needed?

and

Processing `AutoExpansion_Score.ly'

Parsing...

AutoExpansion_Score.ly:37:27: warning: non-UTF-8 input

           Freu -- de sch

                         ö -- ner Göt -- ter -- fun -- ken, Toch -- ter aus E -- ly -- si -- um

AutoExpansion_Score.ly:37:37: warning: non-UTF-8 input

           Freu -- de schö -- ner G

                                   öt -- ter -- fun -- ken, Toch -- ter aus E -- ly -- si -- um

Interpreting music...

Interpreting music...[8]

Preprocessing graphical objects...

(lilypond.exe:14248): Pango-WARNING **: Invalid UTF-8 string passed to pango_layout_set_text()

warning: no glyph for character U+EFFFFFFF in font `C:/Program

Files (x86)/LilyPond/usr/share/lilypond/current/fonts/otf/CenturySchL-Roma.otf'

(lilypond.exe:14248): Pango-WARNING **: Invalid UTF-8 string passed to pango_layout_set_text()

warning: no glyph for character U+EFFFFFFF in font `C:/Program
Files (x86)/LilyPond/usr/share/lilypond/current/fonts/otf/CenturySchL-Roma.otf'

MIDI output to `AutoExpansion_Score.midi'...

Finding the ideal number of pages...

Fitting music on 1 page...
Drawing systems...

Layout output to `AutoExpansion_Score.ps'...

Converting to `./AutoExpansion_Score.pdf'...

Success: compilation successfully completed

LilyPond terminated in 25.181 seconds.

In the .pdf that is produced the letters with an umlaut are missing (see attached file).

Sincerely,

Viktor

[Dave Hofmann]

Hi Viktor,

I could reproduce the issue on my Windows Machine. The problem is that the file should be UTF-8 encoded, but it is apparently written in another encoding on Windows. You can check if the file is marked as UTF-8 in Eclipse by right-clicking it and then navigating to Properties -> Resource -> Text File Encoding.

I assume that MPS relies on the default system encoding for writing the LilyPond file, which is UTF-8 on Linux and Mac, but not on Windows apparently. I will fix this bug in the next release by enforcing UTF-8 encoding on all operating systems.

As a workaround you can edit the LilyPond file directly and replace the wrong characters. If you need to compile often, I would propose to remove the characters from the .mcl source file and add them in the end when you're finished with the rest of the composition.

Best

Dave

[Dave Hofmann]

Hi Viktor,

this should be fixed in MPS 1.6.0. Please let me know if the problem persists.

Best

Dave