+2012-08-16 Paul Eggert <eggert@cs.ucla.edu>
+
+ Use ASCII tests for character types.
+ * category.c, dispnew.c, doprnt.c, editfns.c, syntax.c, term.c:
+ * xfns.c, xterm.c:
+ Don't include <ctype.h>; was not needed.
+ * charset.c, doc.c, fileio.c, font.c, frame.c, gtkutil.c, image.c:
+ * sysdep.c, xfaces.c:
+ Include <c-ctype.h> instead of <ctype.h>.
+ * nsterm.m: Include <c-ctype.h>.
+ * charset.c (read_hex):
+ * doc.c (Fsnarf_documentation):
+ * fileio.c (IS_DRIVE) [WINDOWSNT]:
+ (DRIVE_LETTER) [DOS_NT]:
+ (Ffile_name_directory, Fexpand_file_name)
+ (Fsubstitute_in_file_name):
+ * font.c (font_parse_xlfd, font_parse_fcname):
+ * frame.c (x_set_font_backend):
+ * gtkutil.c (xg_get_font):
+ * image.c (xbm_scan, xpm_scan, pbm_scan_number):
+ * nsimage.m (hexchar):
+ * nsterm.m (ns_xlfd_to_fontname):
+ * sysdep.c (system_process_attributes):
+ * xfaces.c (hash_string_case_insensitive):
+ Use C-locale tests instead of locale-specific tests for character
+ types, since we want the ASCII interpretation here, not the
+ interpretation suitable for whatever happens to be the current locale.
+