Patch for i18n of gcc/java/ programs

[Andrew Haley <aph at redhat dot com>]

Joseph S. Myers writes:
 > Various of the programs in gcc/java fail to support i18n (despite a
 > FIXME comment in one of them about code bad for i18n).  This patch
 > fixes this (but not the FIXME).  I didn't do anything about
 > translation of the jcf-dump output file as that may need further
 > consideration about whether any languages might require further
 > reordering of the output.
 > 
 > Bootstrapped with no regressions on i686-pc-linux-gnu (and checked
 > that new messages appear when gcc.pot is regenerated).  OK to commit
 > to mainline?
 > 
 > 2004-02-08  Joseph S. Myers  <jsm@polyomino.org.uk>
 > 
 > 	* gjavah.c: Include "intl.h".
 > 	(error): New function.
 > 	(main): Call gcc_init_libintl.
 > 	(get_field_name, throwable_p, print_c_decl, print_full_cxx_name,
 > 	print_stub_or_jni, process_file, main): Use error rather than
 > 	fprintf.
 > 	(print_method_info, usage, help, version, main): Mark strings for
 > 	translation with _.  Avoid splitting up sentences.  Send
 > 	information messages to stdout.
 > 	* jcf-dump.c: Include "intl.h".
 > 	(main): Call gcc_init_libintl.
 > 	(process_class, usage, help, version, main, CHECK_PC_IN_RANGE):
 > 	Mark error, usage and version messages for translation with _.
 > 	Avoid splitting up sentences.
 > 	* jv-scan.c: Include "intl.h".
 > 	(fatal_error, warning): Change parameter s to msgid.  Translate
 > 	messages.
 > 	(main): Call gcc_init_libintl.
 > 	(usage, help, version): Mark error, usage and version messages for
 > 	translation with _.  Avoid splitting up sentences.
 > 	* jvgenmain.c: Include "intl.h".
 > 	(main): Call gcc_init_libintl.
 > 	(usage, main): Mark error messages for translation with _.
 > 	* Make-lang.in (GCJH_OBJS, JVSCAN_OBJS, JCFDUMP_OBJS,
 > 	JVGENMAIN_OBJS): Add intl.o.
 > 	(java/jcf-dump.o, java/gjavah.o, java/jv-scan.o,
 > 	java/jvgenmain.o): Update dependencies.

OK.

Andrew.