jjgarcia
02622043bc
New implementation of SUBTYPEP.
2003-04-10 14:32:02 +00:00
jjgarcia
1111281642
STRUCTURE is not the name of a type; use STRUCTURE-CLASS instead.
2003-04-08 14:35:50 +00:00
jjgarcia
b6d9ee1820
si_gc() takes only one argument!
2003-04-08 14:33:59 +00:00
jjgarcia
823b5ae50f
Simplify how *FEATURES* is built.
2003-04-08 14:33:05 +00:00
jjgarcia
255b967d3b
Important comment.
2003-04-08 14:32:06 +00:00
jjgarcia
7b5fe28c18
Implement bignum operations using GMP and not dealing with the low-level
...
representation of these numbers.
2003-04-08 14:31:05 +00:00
jjgarcia
04c7cce1d4
Do not try to deallocate pointers outside our heap.
2003-04-08 14:28:05 +00:00
jjgarcia
1e9e0c140c
Do not try to deallocate bignum registers.
2003-04-08 14:00:32 +00:00
jjgarcia
8e34a6179a
(COND) is a valid expression and should return NIL.
2003-03-21 15:30:38 +00:00
jjgarcia
6b76d155ee
Create new functions SI::{GET,PUT,REM}-SYSPROP to handle vital information
...
about functions, SETF forms, DEFTYPEs, etc. Property lists are no longer
used for this task.
2003-03-21 14:18:56 +00:00
jjgarcia
d8300559a9
Protect both calls to the library initialization function, because for
...
some cases, also the first call executes code.
2003-03-21 14:12:54 +00:00
jjgarcia
8bb27d3094
Remove public function putprop(); use si_putprop() instead.
...
Removed public functions FEtype_error_plist().
Low level function getf() and get() renamed as ecl_get[f]().
Simplified the loops which deal with property lists.
Assertions ASSERT and CHECK-TYPE moved from conditions.lisp to assert.lisp; old definitions removed.
2003-03-21 09:24:37 +00:00
jjgarcia
aabafc3c6c
Remove public functions remf() and remprop().
2003-03-20 15:47:44 +00:00
jjgarcia
5f57762ca6
FEcondition does no longer exist.
2003-03-20 15:46:52 +00:00
jjgarcia
57e80be0aa
Fix typo.
2003-03-20 14:10:38 +00:00
jjgarcia
723d24eba7
Remove spurious semicolon.
2003-03-18 10:16:20 +00:00
jjgarcia
41c0868469
Bunch of fixes. See CHANGELOG.
2003-03-17 10:39:08 +00:00
jjgarcia
a767dd4939
Symbol GC moved from package COMMON-LISP to SYSTEM (and imported into CL-USER).
2003-03-13 13:10:15 +00:00
jjgarcia
b7ced3d4ba
Change the way Cblocks are created. Now, within a DLL, several Cblocks may
...
be created for the different components, but in the end they are chained together
and point to a main Cblock with the handle of the library.
2003-03-13 11:10:35 +00:00
jjgarcia
5478bbe21b
DLLs and files have now termination procedures, which unload the DLL or close
...
the file, when these objects are garbage collected.
2003-03-12 10:05:42 +00:00
jjgarcia
94f70da47b
Remove debug statement.
2003-03-11 18:22:51 +00:00
jjgarcia
e10ee7808b
When dealing with compiler-macros, ECL may enter an infinite loop.
2003-03-11 14:56:48 +00:00
jjgarcia
c972e21255
Make symbol COMPILER-MACRO built-in.
2003-03-11 14:55:58 +00:00
jjgarcia
ab08b07ac1
ECL now detects wrong function calls, such as ("foo"), (setq a ("foo")), etc
2003-03-11 12:47:58 +00:00
jjgarcia
a627819a71
Typo fixed.
2003-03-10 14:02:44 +00:00
jjgarcia
5f6f65b407
Bunch of fixes inspired by the ANSI test suite of GCL.
2003-03-10 09:10:13 +00:00
jjgarcia
2be5baa3e4
A stupid character got in the file.
2003-03-05 09:33:19 +00:00
jjgarcia
75b7207b6d
Fixed typo.
2003-03-04 17:25:43 +00:00
jjgarcia
9771e91572
+ New implementation of DESTRUCTURING-BIND
...
+ The bytecodes compiler now alloes a lambda list with &key and no keywords
+ WITH-HASHTABLE-ITERATOR implemented
+ QUIT is now silent
+ In MAKE-PATHNAME, values supplied with a value of NIL would be overwritten
by the content of :DEFAULTS.
2003-03-04 16:35:03 +00:00
jjgarcia
c935faf44a
In cl_clear_input(), do not complain when fseek() fails.
2002-12-17 10:53:29 +00:00
jjgarcia
b0e60245a4
c::*funs* may contain symbols (like CB marking closure boundary),
...
function structures (marking function definitions), or lists with the
format (fname MACRO interpreted-function), to denote MACROLET
expansions.
2002-12-17 10:45:36 +00:00
jjgarcia
8209fc021e
Write FTYPE declarations correctly.
2002-12-17 09:53:52 +00:00
jjgarcia
d21a1fd389
Previous fixed removed. Hopefully this time I got FTYPE declarations
...
correct: they are restricted to (FTYPE (FUNCTION (arg-types*) ret-type*) name*)
2002-12-17 09:53:04 +00:00
jjgarcia
8852166529
FTYPE should accept two types of declarations: the ones with a formal type
...
(FTYPE (FUNCTION (T T) T) MY-FUN1 MY-FUN2)
and the abbreviated ones
(FTYPE MY-FUN1 (T T) T)
2002-12-17 09:15:37 +00:00
jjgarcia
22280e7f6c
Fix macroexpander for define-condition.
2002-12-14 16:57:11 +00:00
jjgarcia
c9fb52f320
In APPLY, FUNCALL, etc, check whether the input argument is OBJNULL. This
...
way, when, in low-safety mode, we funcall a symbol which is unbound, the
right error is signaled.
2002-12-14 16:03:44 +00:00
jjgarcia
73447f1b3e
Set safe compilation as default.
2002-12-14 15:44:47 +00:00
jjgarcia
912a7fb03c
The declaration specifier for functions is FTYPE, not FUNCTION.
2002-12-14 15:37:29 +00:00
jjgarcia
aaa68d02a8
LOGICAL-PATHNAME implemented.
2002-12-14 15:03:07 +00:00
jjgarcia
0c0e58d308
Use @'' to access the symbols instead of searching them.
2002-12-14 14:09:07 +00:00
jjgarcia
443a2895f0
update configure to the latest changes in aclocal.m4
2002-12-09 14:37:32 +00:00
jjgarcia
7828d15172
"conftestval" has to be truncated to zero before writing the output of
...
our test program.
2002-12-09 14:36:18 +00:00
jjgarcia
9210270227
In ECL_FIXNUM_TYPE, the C code outputs one less bit than there really
...
are. Furthermore, it fails to produce the right limits for the integers.
2002-12-09 13:08:11 +00:00
jjgarcia
491f242fa4
Remove harmless type warning.
2002-12-09 12:53:44 +00:00
jjgarcia
43099ff052
More fixes about integer types.
2002-12-09 12:51:40 +00:00
jjgarcia
1b64032b99
We need string.h for strerror().
2002-12-09 12:49:49 +00:00
jjgarcia
495d5d0172
We need string.h for memcpy().
2002-12-09 12:47:44 +00:00
jjgarcia
31791eb3d7
More integer type fixes.
2002-12-09 12:45:43 +00:00
jjgarcia
d8ffa12fda
We need string.h for memcpy().
2002-12-09 12:44:34 +00:00
jjgarcia
81fd297e7e
Fix prototypes for functions which changed from int -> cl_fixnum.
2002-12-09 12:39:26 +00:00
