Debug printing of Unicode chars is inconsistent

[erkyrath]

The --trace verbs option calls print_verbs_by_number(), which dumps from English_verbs_text, which is cloned directly from the token_text and therefore uses whichever characters or escape sequence the author used.

But for the prepositions, --trace verbs calls print_dict_word(), which calls show_uchar(), which will print fancy characters using UTF-8 output if the -Cu switch is set. It does not look at how the author wrote the word.

The --trace dict and -r (gametext.txt output) options go through dictword_to_text(), which calls zscii_to_text(), which prints some (8-bit) Unicode characters according to the -C switch, but falls back to @-escapes for everything else. It does not try UTF-8 even in -Cu mode.

I don't think anybody cares about this, but it's a lot of slightly confusing differences.

(See also #301 .)

[erkyrath]

Gah, not all of the above is correct. This is such a mess internally.

[heasm66]

Your efforts are appreciated...

[erkyrath]

(Post #354):

One problem is that the only UTF-8 output code is in show_uchar(), which is part of the (still) rather clunky dict_show_buf system. There's nothing that just prints Unicode characters to stdout.

Arguably zscii_to_text() should gain a UTF-8 option.