[RFA] Make word-break-characters language dependent (3/3)

[Joel Brobecker <brobecker at gnat dot com>]

[Grrr, I just noticed the typo in the subject. I corrected it for this message]

This is the last patch of the series. My many thanks to Eli who
discussed this issue with me.

This time, I replaced all usage of gdb_completer_word_break_characters
and get_gdb_completer_word_break_characters mostly by a call to
the current_language la_word_break_characters function.

There is one instance in top.c where I used default_word_break_characters
instead of the language dependent one, because I wasn't completely sure
if the current language would be already set or not, and whether the
current language one was appropriate. I prefered to leave it the way it
is now, and deal with a potential latent bug later, rather than
introduce one...

Since the word_break_characters of all language without exception is
set to the default_word_break_characters, the is no behavior change.
Only when a language starts defining its own word_break_characters
(we will do that for Ada) can the completer code behavior be slightly
tuned.

2003-10-01  J. Brobecker  <brobecker@gnat.com>

        * completer.h (get_gdb_completer_word_break_characters): Delete.
        * completer.c: include language.h.
        (gdb_completer_word_break_characters): Delete.
        (get_gdb_completer_word_break_characters): Delete.
        (location_completer): Use the word break characters of the current
        language.
        (complete_line): Likewise.
        (line_completion_function): Likewise.
        (skip_quoted_chars): Likewise.
        * Makefile.in (completer.o): Add dependency on language.h.
        * top.c (init_main): Set the readline word break characters
        to GDB's default word break characters.

Tested on x86-linux, no regressions.

Ok to apply?

Thanks,
-- 
Joel

Attachment: patch3.diff
Description: Text document