Mercurial > hg > xemacs-beta
diff src/data.c @ 5374:d967d96ca043
Conditionalise the old-* functions and byte codes at compile time.
src/ChangeLog addition:
2011-03-15 Aidan Kehoe <kehoea@parhasard.net>
* config.h.in (SUPPORT_CONFOUNDING_FUNCTIONS): New #define,
equivalent NEED_TO_HANDLE_21_4_CODE by default, describing whether
this XEmacs should support the old-eq, old-equal and related
functions and byte codes.
* bytecode.c (UNUSED):
Only interpret old-eq, old-equal, old-memq if
SUPPORT_CONFOUNDING_FUNCTIONS is defined.
* data.c:
Move Fold_eq to fns.c with the rest of the Fold_* functions.
* fns.c:
* fns.c (Fmemq):
* fns.c (memq_no_quit):
* fns.c (assoc_no_quit):
* fns.c (Frassq):
* fns.c (Fequal):
* fns.c (Fold_equal):
* fns.c (syms_of_fns):
Group old-eq, old-equal, old-memq etc together, surround them with
#ifdef SUPPORT_CONFOUNDING_FUNCTIONS.
lisp/ChangeLog addition:
2011-03-15 Aidan Kehoe <kehoea@parhasard.net>
* bytecomp.el:
Don't generate the old-eq, old-memq, old-equal bytecodes any more,
but keep the information about them around for the sake of the
disassembler.
man/ChangeLog addition:
2011-03-15 Aidan Kehoe <kehoea@parhasard.net>
* lispref/objects.texi (Character Type):
* lispref/objects.texi (Equality Predicates):
No longer document `old-eq', `old-equal', they haven't been used
in years.
tests/ChangeLog addition:
2011-03-17 Aidan Kehoe <kehoea@parhasard.net>
* automated/lisp-tests.el:
Only test the various old-* function if old-eq is bound and a
subr.
author | Aidan Kehoe <kehoea@parhasard.net> |
---|---|
date | Thu, 17 Mar 2011 20:13:00 +0000 |
parents | 47298dcf2e8f |
children | 4f39e57a82b4 ac37a5f7e5be |
line wrap: on
line diff
--- a/src/data.c Mon Mar 14 21:04:45 2011 +0000 +++ b/src/data.c Thu Mar 17 20:13:00 2011 +0000 @@ -183,24 +183,6 @@ return EQ_WITH_EBOLA_NOTICE (object1, object2) ? Qt : Qnil; } -DEFUN ("old-eq", Fold_eq, 2, 2, 0, /* -Return t if the two args are (in most cases) the same Lisp object. - -Special kludge: A character is considered `old-eq' to its equivalent integer -even though they are not the same object and are in fact of different -types. This is ABSOLUTELY AND UTTERLY HORRENDOUS but is necessary to -preserve byte-code compatibility with v19. This kludge is known as the -\"char-int confoundance disease\" and appears in a number of other -functions with `old-foo' equivalents. - -Do not use this function! -*/ - (object1, object2)) -{ - /* #### blasphemy */ - return HACKEQ_UNSAFE (object1, object2) ? Qt : Qnil; -} - DEFUN ("null", Fnull, 1, 1, 0, /* Return t if OBJECT is nil. */ @@ -3568,7 +3550,6 @@ DEFSUBR (Fdiv); #endif DEFSUBR (Feq); - DEFSUBR (Fold_eq); DEFSUBR (Fnull); Ffset (intern ("not"), intern ("null")); DEFSUBR (Flistp);