Merge: Optimized use of escape_to_c

Merge: Optimized use of escape_to_c

All is in the title, @privat noticed me this morning that text::escape_to_c was taking around 4% of the time in nitc, taking the runtime to 15.183 GIr according to Valgrind.

By shortcutting all Unicode-related services since the C-side escaping is done on ASCII-sequences only and UTF-8 cannot contain invalid C sequences, the gain is substantial, one run of nitc is now done in 14.580 GIr, a gain of 600+ MIr (around 4%, what a coincidence !).

Pull-Request: #1685
Reviewed-by: Alexis Laferrière <alexis.laf@xymus.net>
Reviewed-by: Jean Privat <jean@pryen.org>