bpt/guile.git
25 years ago*** empty log message ***
Jim Blandy [Wed, 14 Oct 1998 08:22:07 +0000 (08:22 +0000)]
*** empty log message ***

25 years ago* configure.in: Allow tabs and whitespace between `void' and
Jim Blandy [Wed, 14 Oct 1998 08:22:00 +0000 (08:22 +0000)]
* configure.in: Allow tabs and whitespace between `void' and
`usleep'.  (Thanks to Harvey J. Stein.)
* configure: Regenerated.

25 years ago*** empty log message ***
Jim Blandy [Wed, 14 Oct 1998 07:32:28 +0000 (07:32 +0000)]
*** empty log message ***

25 years ago* (__scm.h, alist.c, eval.c, feature.c, gsubr.c, numbers.c,
Jim Blandy [Wed, 14 Oct 1998 07:32:13 +0000 (07:32 +0000)]
* (__scm.h, alist.c, eval.c, feature.c, gsubr.c, numbers.c,
options.c): Rename RECKLESS -> SCM_RECKLESS, CAUTIOUS ->
SCM_CAUTIOUS; this way, 1) there's only one version of each flag
to define (we used to have both RECKLESS and SCM_RECKLESS), and 2)
if we want to use them in a header file some day, we can.  (Thanks
to Michael Livshin.)

25 years ago*** empty log message ***
Jim Blandy [Tue, 13 Oct 1998 23:56:29 +0000 (23:56 +0000)]
*** empty log message ***

25 years ago* stime.c (scm_get_internal_real_time): Do the arithmetic using
Jim Blandy [Tue, 13 Oct 1998 23:56:08 +0000 (23:56 +0000)]
* stime.c (scm_get_internal_real_time): Do the arithmetic using
SCM numbers, so we won't have rollover problems; the range of a
signed long in milliseconds is about 25 days.  (Thanks to Karl
Hegbloom.)

25 years ago*** empty log message ***
Jim Blandy [Tue, 13 Oct 1998 23:19:04 +0000 (23:19 +0000)]
*** empty log message ***

25 years ago* threads.h (scm_thread_usleep, scm_thread_sleep): New declarations.
Jim Blandy [Tue, 13 Oct 1998 23:17:34 +0000 (23:17 +0000)]
* threads.h (scm_thread_usleep, scm_thread_sleep): New declarations.
* scmsigs.c (usleep): Clean up oddities declaring usleep; since
we're just using it, not redefining it, we can use a K&R style
declaration here.
(sleep): Declare this, too, if the system hasn't.
(scm_sleep, scm_usleep): Use scm_thread_sleep and
scm_uthread_sleep if they're available; otherwise, just call the
system functions.
* scmconfig.h.in: Regenerated.

25 years agoDon't redefine sleep and usleep.
Jim Blandy [Tue, 13 Oct 1998 23:17:09 +0000 (23:17 +0000)]
Don't redefine sleep and usleep.
* coop.c (sleep, usleep): Remove declarations; we don't use or
redefine these any more.
(scm_thread_usleep, scm_thread_sleep): New functions which do the
job of usleep and sleep in a thread-friendly way.  We can use
these in the rest of Guile.  Define versions for systems both with
and without iselect.
* coop.c (scm_thread_sleep): Make `slept' an unsigned long.
* coop.c (coop_sleephelp): Remove ANSI #ifdef hair.

25 years agoDon't redefine sleep/usleep.
Jim Blandy [Tue, 13 Oct 1998 23:16:59 +0000 (23:16 +0000)]
Don't redefine sleep/usleep.
* configure.in: Remove tests for usleep's argument type; we only
need that if we're going to replace it.
* acconfig.h (USLEEP_ARG_TYPE): Delete.  All the other SLEEP
garbage is needed just to use usleep and sleep without compiler
warnings.
* configure: Regenerated.

25 years ago*** empty log message ***
Jim Blandy [Mon, 12 Oct 1998 21:30:41 +0000 (21:30 +0000)]
*** empty log message ***

25 years ago*** empty log message ***
Jim Blandy [Mon, 12 Oct 1998 21:08:56 +0000 (21:08 +0000)]
*** empty log message ***

25 years ago* threads.c: Doc fix.
Jim Blandy [Mon, 12 Oct 1998 21:08:47 +0000 (21:08 +0000)]
* threads.c: Doc fix.

25 years agoThe argument type of usleep varies from system to system,
Jim Blandy [Mon, 12 Oct 1998 21:08:38 +0000 (21:08 +0000)]
The argument type of usleep varies from system to system,
as does the return type.  We really shouldn't be redefining usleep
at all, but I don't have time to clean that up before the 1.3.
release.  It's on the schedule for afterwards.  (Thanks to Julian
Satchell.)
* coop.c (usleep): Use USLEEP_ARG_TYPE in prototype and
definition.
* scmsigs.c (usleep): Use USLEEP_ARG_TYPE in prototype.
* scmconfig.h: Regenerated.

25 years ago* configure.in (GUILE_FUNC_DECLARED): Name the cache variables
Jim Blandy [Mon, 12 Oct 1998 21:08:17 +0000 (21:08 +0000)]
* configure.in (GUILE_FUNC_DECLARED): Name the cache variables
starting with guile_cv_; ac_cv_ is autoconf's namespace.
The type of the argument to usleep varies from system to system,
as does the return type.  We really shouldn't be redefining usleep
at all, but I don't have time to clean that up before the 1.3
release.  It's on the schedule for afterwards.
* configure.in: Cache results from usleep return value test.
Test for the type of the usleep argument, and cache that too.
* acconfig.h (USLEEP_ARG_TYPE): New macro.

25 years ago* configure: Regenerated.
Jim Blandy [Mon, 12 Oct 1998 21:08:03 +0000 (21:08 +0000)]
* configure: Regenerated.

25 years agoThe type of the argument to usleep varies from system to system,
Jim Blandy [Mon, 12 Oct 1998 21:07:43 +0000 (21:07 +0000)]
The type of the argument to usleep varies from system to system,
as does the return type.  We really shouldn't be redefining usleep
at all, but I don't have time to clean that up before the 1.3
release.  It's on the schedule for afterwards.
* configure.in: Cache results from usleep return value test.
Test for the type of the usleep argument, and cache that too.
* acconfig.h (USLEEP_ARG_TYPE): New macro.

25 years ago*** empty log message ***
Jim Blandy [Mon, 12 Oct 1998 20:15:14 +0000 (20:15 +0000)]
*** empty log message ***

25 years ago*** empty log message ***
Jim Blandy [Mon, 12 Oct 1998 20:13:27 +0000 (20:13 +0000)]
*** empty log message ***

25 years ago* simpos.c (scm_software_type): Procedure deleted. This isn't the
Jim Blandy [Mon, 12 Oct 1998 20:13:11 +0000 (20:13 +0000)]
* simpos.c (scm_software_type): Procedure deleted.  This isn't the
right way to handle system variation.  Autoconf's approach is the
way and the light.
* simpos.h (scm_software_type): Declaration deleted.

25 years ago* script.c (scm_find_executable): Don't test if unix is #defined
Jim Blandy [Mon, 12 Oct 1998 20:11:38 +0000 (20:11 +0000)]
* script.c (scm_find_executable): Don't test if unix is #defined
here; first, NetBSD doesn't #define it, and second, it's the wrong
way to go about these things.
(dld_find_executable): Delete this MSDOS support code.  This isn't
the way we want to support this; it needs to be rethunk at a
larger scale.

25 years ago* genio.c (scm_do_read_line): Don't just politely check that the
Jim Blandy [Mon, 12 Oct 1998 20:09:34 +0000 (20:09 +0000)]
* genio.c (scm_do_read_line): Don't just politely check that the
line was either non-empty or EOF.  Abort if it's empty and not
EOF.

25 years ago*** empty log message ***
Jim Blandy [Mon, 12 Oct 1998 20:08:44 +0000 (20:08 +0000)]
*** empty log message ***

25 years ago* r4rs.scm (OPEN_READ, OPEN_WRITE, OPEN_BOTH): Don't bother
Jim Blandy [Mon, 12 Oct 1998 20:08:31 +0000 (20:08 +0000)]
* r4rs.scm (OPEN_READ, OPEN_WRITE, OPEN_BOTH): Don't bother
testing software-type here.  That's the least of our Windows
porting issues, and it's done wrong anyway.

25 years ago*** empty log message ***
Jim Blandy [Sun, 11 Oct 1998 21:56:45 +0000 (21:56 +0000)]
*** empty log message ***

25 years ago* scmconfig.h.in: Regenerated.
Jim Blandy [Sun, 11 Oct 1998 21:56:18 +0000 (21:56 +0000)]
* scmconfig.h.in: Regenerated.

25 years ago* acconfig.h (HAVE_RL_GETC_FUNCTION): Fix this entry.
Jim Blandy [Sun, 11 Oct 1998 21:56:08 +0000 (21:56 +0000)]
* acconfig.h (HAVE_RL_GETC_FUNCTION): Fix this entry.

25 years ago*** empty log message ***
Jim Blandy [Sun, 11 Oct 1998 20:35:28 +0000 (20:35 +0000)]
*** empty log message ***

25 years ago* libguile.h: Don't omit the dynamic linking functions. (Thanks
Jim Blandy [Sun, 11 Oct 1998 20:35:07 +0000 (20:35 +0000)]
* libguile.h: Don't omit the dynamic linking functions.  (Thanks
to Greg Badros.)

25 years ago*** empty log message ***
Jim Blandy [Sun, 11 Oct 1998 18:42:32 +0000 (18:42 +0000)]
*** empty log message ***

25 years ago* genio.c (scm_do_read_line): Count lines correctly when the file
Jim Blandy [Sun, 11 Oct 1998 18:42:17 +0000 (18:42 +0000)]
* genio.c (scm_do_read_line): Count lines correctly when the file
doesn't end in a newline.

25 years ago*** empty log message ***
Jim Blandy [Sat, 10 Oct 1998 21:29:37 +0000 (21:29 +0000)]
*** empty log message ***

25 years ago* genio.c (scm_do_read_line): Maintain the line count correctly.
Jim Blandy [Sat, 10 Oct 1998 21:27:34 +0000 (21:27 +0000)]
* genio.c (scm_do_read_line): Maintain the line count correctly.

25 years ago* GUILE-VERSION: bump to 1.2.91, since we're doing snapshots again.
Jim Blandy [Sat, 10 Oct 1998 21:00:57 +0000 (21:00 +0000)]
* GUILE-VERSION: bump to 1.2.91, since we're doing snapshots again.

25 years ago*** empty log message ***
Jim Blandy [Sat, 10 Oct 1998 21:00:45 +0000 (21:00 +0000)]
*** empty log message ***

25 years ago*** empty log message ***
Jim Blandy [Sat, 10 Oct 1998 14:27:14 +0000 (14:27 +0000)]
*** empty log message ***

25 years ago*** empty log message ***
Jim Blandy [Sat, 10 Oct 1998 12:59:50 +0000 (12:59 +0000)]
*** empty log message ***

25 years ago* numbers.c (big2str): Protect t from garbage collection until
Jim Blandy [Sat, 10 Oct 1998 12:59:37 +0000 (12:59 +0000)]
* numbers.c (big2str): Protect t from garbage collection until
we're done.  (Thanks to Gary Houston.)

25 years ago* gc.c (scm_return_first): Remove #ifdef __STDC__ garbage; Guile
Jim Blandy [Sat, 10 Oct 1998 12:59:29 +0000 (12:59 +0000)]
* gc.c (scm_return_first): Remove #ifdef __STDC__ garbage; Guile
requires ANSI now.

25 years ago* Guile 1.2.90 released --- beta.
Jim Blandy [Sat, 10 Oct 1998 12:58:00 +0000 (12:58 +0000)]
* Guile 1.2.90 released --- beta.
* GUILE-VERSION: Set to 1.2.90.  This would appear to be a
regression from 1.3a, but everyone knows that the next release is
1.3, I want to switch to a more coherent version numbering system,
and now is the time.

25 years ago*** empty log message ***
Jim Blandy [Sat, 10 Oct 1998 12:39:22 +0000 (12:39 +0000)]
*** empty log message ***

25 years ago*** empty log message ***
Jim Blandy [Sat, 10 Oct 1998 12:37:34 +0000 (12:37 +0000)]
*** empty log message ***

25 years ago*** empty log message ***
Jim Blandy [Fri, 9 Oct 1998 15:25:08 +0000 (15:25 +0000)]
*** empty log message ***

25 years ago* Makefile.am (libguile_la_LDFLAGS): Increment shared library
Jim Blandy [Fri, 9 Oct 1998 15:24:40 +0000 (15:24 +0000)]
* Makefile.am (libguile_la_LDFLAGS): Increment shared library
version number.
* Makefile.in: Regenerated.

25 years agoGo to 1.3b, for the first 1.3 beta release.
Jim Blandy [Fri, 9 Oct 1998 15:23:33 +0000 (15:23 +0000)]
Go to 1.3b, for the first 1.3 beta release.

25 years agoRemove uses of the #/ syntax.
Jim Blandy [Fri, 9 Oct 1998 15:13:12 +0000 (15:13 +0000)]
Remove uses of the #/ syntax.

25 years ago*** empty log message ***
Jim Blandy [Fri, 9 Oct 1998 15:13:11 +0000 (15:13 +0000)]
*** empty log message ***

25 years ago* boot-9.scm (read-path-list-notation-warning): New function:
Jim Blandy [Fri, 9 Oct 1998 15:12:51 +0000 (15:12 +0000)]
* boot-9.scm (read-path-list-notation-warning): New function:
print a warning the first time we see `#/' notation.

25 years ago* guile-config.in: Add copyright notice.
Jim Blandy [Fri, 9 Oct 1998 15:12:40 +0000 (15:12 +0000)]
* guile-config.in: Add copyright notice.

25 years agoMention that the #/ syntax now causes warnings.
Jim Blandy [Fri, 9 Oct 1998 15:12:26 +0000 (15:12 +0000)]
Mention that the #/ syntax now causes warnings.

25 years ago*** empty log message ***
Jim Blandy [Fri, 9 Oct 1998 14:25:04 +0000 (14:25 +0000)]
*** empty log message ***

25 years ago*** empty log message ***
Jim Blandy [Fri, 9 Oct 1998 13:56:27 +0000 (13:56 +0000)]
*** empty log message ***

25 years ago* q.scm (sync-q!, q?, q-remove!, q-push!, enq!): Lots of bugs, and
Jim Blandy [Fri, 9 Oct 1998 13:56:16 +0000 (13:56 +0000)]
* q.scm (sync-q!, q?, q-remove!, q-push!, enq!): Lots of bugs, and
(eq?  #f '()) assumptions.  Make functions that aren't documented
to return anything else return the queue itself.  (Bug report from
Michael Livshin --- thanks!)

25 years ago*** empty log message ***
Jim Blandy [Fri, 9 Oct 1998 12:47:01 +0000 (12:47 +0000)]
*** empty log message ***

25 years agoCentralize the creation of port objects based on stdio FILE * in
Jim Blandy [Fri, 9 Oct 1998 12:46:31 +0000 (12:46 +0000)]
Centralize the creation of port objects based on stdio FILE * in
fports.c; don't just throw them together anywhere.
* fports.c (scm_stdio_to_port): Make NAME a SCM value, which is
what the rest of Guile wants.  Don't set the revealed count;
that's only appropriate for stdin, stdout, stderr.
(scm_standard_stream_to_port): This function does set the revealed
count.
* init.c (scm_init_standard_ports): Use scm_standard_stream_to_port,
not scm_stdio_to_port.
* filesys.c (scm_open): Call scm_stdio_to_port; don't write it out.
* fports.c (scm_open_file): Same.
* posix.c (scm_pipe): Same.
* socket.c (scm_sock_fd_to_port): Same.
* ioext.c (scm_fdopen): Same.
(scm_freopen): Moved from here to...
* fports.c (scm_freopen): ... here.  This is really something that
munges the internals of an fport, so it should go here.
* fports.h (scm_stdio_to_port): Adjust prototype.
(scm_standard_stream_to_port, scm_freopen): New protoypes.
* ioext.h (scm_freopen): Prototype removed.

25 years ago* fports.h (scm_setbuf0, scm_setvbuf, scm_setfileno,
Jim Blandy [Fri, 9 Oct 1998 12:46:13 +0000 (12:46 +0000)]
* fports.h (scm_setbuf0, scm_setvbuf, scm_setfileno,
scm_evict_ports, scm_open_file, scm_stdio_to_port): Get rid of
SCM_P macro.
Do magic to mix reads and writes on stdio FILE-based ports.
* fports.c (FPORT_READ_SAFE, FPORT_WRITE_SAFE, FPORT_ALL_OKAY):
New macros.
(pre_read, pre_write): New functions.
(local_fgetc, local_fgets, local_ffwrite, local_fputc,
local_fputs): Call them.
(local_fflush): Mark the port as ready for reading and writing.
(scm_stdio_to_port): Set the FPORT_READ_SAFE, FPORT_WRITE_SAFE
flags on new port objects.  This might not be accurate --- who
knows what state the FILE * is in when we get it --- but it won't
do extraneous calls to fflush or fseek, so it's no worse than the
behavior before this change.
* ports.h: Add comment.
Centralize the creation of port objects based on stdio FILE * in
fports.c; don't just throw them together anywhere.
* fports.c (scm_stdio_to_port): Make NAME a SCM value, which is
what the rest of Guile wants.  Don't set the revealed count;
that's only appropriate for stdin, stdout, stderr.
(scm_standard_stream_to_port): This function does set the revealed
count.
* init.c (scm_init_standard_ports): Use scm_standard_stream_to_port,
not scm_stdio_to_port.
* filesys.c (scm_open): Call scm_stdio_to_port; don't write it out.
* fports.c (scm_open_file): Same.
* posix.c (scm_pipe): Same.
* socket.c (scm_sock_fd_to_port): Same.
* ioext.c (scm_fdopen): Same.
(scm_freopen): Moved from here to...
* fports.c (scm_freopen): ... here.  This is really something that
munges the internals of an fport, so it should go here.
* fports.h (scm_stdio_to_port): Adjust prototype.
(scm_standard_stream_to_port, scm_freopen): New protoypes.
* ioext.h (scm_freopen): Prototype removed.

25 years agoDo magic to mix reads and writes on stdio FILE-based ports.
Jim Blandy [Fri, 9 Oct 1998 12:45:20 +0000 (12:45 +0000)]
Do magic to mix reads and writes on stdio FILE-based ports.
* fports.c (FPORT_READ_SAFE, FPORT_WRITE_SAFE, FPORT_ALL_OKAY):
New macros.
(pre_read, pre_write): New functions.
(local_fgetc, local_fgets, local_ffwrite, local_fputc,
local_fputs): Call them.
(local_fflush): Mark the port as ready for reading and writing.
(scm_stdio_to_port): Set the FPORT_READ_SAFE, FPORT_WRITE_SAFE
flags on new port objects.  This might not be accurate --- who
knows what state the FILE * is in when we get it --- but it won't
do extraneous calls to fflush or fseek, so it's no worse than the
behavior before this change.
* ports.h: Add comment.
Centralize the creation of port objects based on stdio FILE * in
fports.c; don't just throw them together anywhere.
* fports.c (scm_stdio_to_port): Make NAME a SCM value, which is
what the rest of Guile wants.  Don't set the revealed count;
that's only appropriate for stdin, stdout, stderr.
(scm_standard_stream_to_port): This function does set the revealed
count.
* init.c (scm_init_standard_ports): Use scm_standard_stream_to_port,
not scm_stdio_to_port.
* filesys.c (scm_open): Call scm_stdio_to_port; don't write it out.
* fports.c (scm_open_file): Same.
* posix.c (scm_pipe): Same.
* socket.c (scm_sock_fd_to_port): Same.
* ioext.c (scm_fdopen): Same.
(scm_freopen): Moved from here to...
* fports.c (scm_freopen): ... here.  This is really something that
munges the internals of an fport, so it should go here.
* fports.h (scm_stdio_to_port): Adjust prototype.
(scm_standard_stream_to_port, scm_freopen): New protoypes.
* ioext.h (scm_freopen): Prototype removed.

25 years agoCentralize the creation of port objects based on stdio FILE * in
Jim Blandy [Fri, 9 Oct 1998 12:44:37 +0000 (12:44 +0000)]
Centralize the creation of port objects based on stdio FILE * in
fports.c; don't just throw them together anywhere.
* fports.c (scm_stdio_to_port): Make NAME a SCM value, which is
what the rest of Guile wants.  Don't set the revealed count;
that's only appropriate for stdin, stdout, stderr.
(scm_standard_stream_to_port): This function does set the revealed
count.
* init.c (scm_init_standard_ports): Use scm_standard_stream_to_port,
not scm_stdio_to_port.
* filesys.c (scm_open): Call scm_stdio_to_port; don't write it out.
* fports.c (scm_open_file): Same.
* posix.c (scm_pipe): Same.
* socket.c (scm_sock_fd_to_port): Same.
* ioext.c (scm_fdopen): Same.
(scm_freopen): Moved from here to...
* fports.c (scm_freopen): ... here.  This is really something that
munges the internals of an fport, so it should go here.
* fports.h (scm_stdio_to_port): Adjust prototype.
(scm_standard_stream_to_port, scm_freopen): New protoypes.
* ioext.h (scm_freopen): Prototype removed.
* filesys.c (set_element, get_element): This can work on both pipe
and file ports, so use SCM_FPORTP to typecheck, instead of testing
for scm_tc16_fport.

25 years ago* scmconfig.h.in: Regenerated.
Jim Blandy [Fri, 9 Oct 1998 12:44:06 +0000 (12:44 +0000)]
* scmconfig.h.in: Regenerated.

25 years ago* fports.h (scm_setbuf0, scm_setvbuf, scm_setfileno,
Jim Blandy [Fri, 9 Oct 1998 12:43:51 +0000 (12:43 +0000)]
* fports.h (scm_setbuf0, scm_setvbuf, scm_setfileno,
scm_evict_ports, scm_open_file, scm_stdio_to_port): Get rid of
SCM_P macro.
Centralize the creation of port objects based on stdio FILE * in
fports.c; don't just throw them together anywhere.
* fports.c (scm_stdio_to_port): Make NAME a SCM value, which is
what the rest of Guile wants.  Don't set the revealed count;
that's only appropriate for stdin, stdout, stderr.
(scm_standard_stream_to_port): This function does set the revealed
count.
* init.c (scm_init_standard_ports): Use scm_standard_stream_to_port,
not scm_stdio_to_port.
* filesys.c (scm_open): Call scm_stdio_to_port; don't write it out.
* fports.c (scm_open_file): Same.
* posix.c (scm_pipe): Same.
* socket.c (scm_sock_fd_to_port): Same.
* ioext.c (scm_fdopen): Same.
(scm_freopen): Moved from here to...
* fports.c (scm_freopen): ... here.  This is really something that
munges the internals of an fport, so it should go here.
* fports.h (scm_stdio_to_port): Adjust prototype.
(scm_standard_stream_to_port, scm_freopen): New protoypes.
* ioext.h (scm_freopen): Prototype removed.

25 years ago*** empty log message ***
Jim Blandy [Fri, 9 Oct 1998 12:12:33 +0000 (12:12 +0000)]
*** empty log message ***

25 years ago* configure.in: Call AC_C_INLINE, so we can use inline happily in
Jim Blandy [Fri, 9 Oct 1998 12:12:26 +0000 (12:12 +0000)]
* configure.in: Call AC_C_INLINE, so we can use inline happily in
libguile.
* configure: Regenerated.

25 years ago*** empty log message ***
Jim Blandy [Fri, 9 Oct 1998 11:50:21 +0000 (11:50 +0000)]
*** empty log message ***

25 years agoChange the definition of the functions in scm_ptobfuns so that
Jim Blandy [Fri, 9 Oct 1998 10:02:41 +0000 (10:02 +0000)]
Change the definition of the functions in scm_ptobfuns so that
they get passed the port object, not the port's stream.
* ports.h (scm_ptobfuns): Rename all `stream' arguments to `port'.
* gc.c (scm_gc_sweep): Pass the port itself to the free function.
* genio.c (scm_putc, scm_puts, scm_lfwrite, scm_fflush, scm_getc):
Pass the port itself to the scm_ptobs function.
* ports.c (scm_close_port, scm_force_output, scm_flush_all_ports,
scm_generic_fgets): Same.
(putc_void_port, puts_void_port, write_void_port, flush_void_port,
getc_void_port, fgets_void_port, close_void_port): Just change the
argument names; these functions don't really do anything.
* fports.c (local_fgetc, local_fgets, local_fclose, local_fflush,
local_fputc, local_fputs, local_ffwrite, local_pclose): Take the
port as an argument, and use SCM_STREAM to get the stdio FILE *.
Also, use prototyped definitions, and get rid of the extra
declarations.
(scm_fptob, scm_pipob): We don't need casts here any more.
* strports.c (prinstpt): Use prototype declarations.
(stputc, stwrite, stputs, stgetc): Take the port as an argument,
and use SCM_STREAM to get the string info.  Also, use prototyped
definitions, and get rid of the extra declarations.
* vports.c (sfputc, sfwrite, sfputs, sfflush, sfgetc, sfclose,
noop0): Same.

25 years agoChange the definition of the functions in scm_ptobfuns so that
Jim Blandy [Fri, 9 Oct 1998 10:01:12 +0000 (10:01 +0000)]
Change the definition of the functions in scm_ptobfuns so that
they get passed the port object, not the port's stream.
* ports.h (scm_ptobfuns): Rename all `stream' arguments to `port'.
* gc.c (scm_gc_sweep): Pass the port itself to the free function.
* genio.c (scm_putc, scm_puts, scm_lfwrite, scm_fflush, scm_getc):
Pass the port itself to the scm_ptobs function.
* ports.c (scm_close_port, scm_force_output, scm_flush_all_ports,
scm_generic_fgets): Same.
(putc_void_port, puts_void_port, write_void_port, flush_void_port,
getc_void_port, fgets_void_port, close_void_port): Just change the
argument names; these functions don't really do anything.
* fports.c (local_fgetc, local_fgets, local_fclose, local_fflush,
local_fputc, local_fputs, local_ffwrite, local_pclose): Take the
port as an argument, and use SCM_STREAM to get the stdio FILE *.
Also, use prototyped definitions, and get rid of the extra
declarations.
(scm_fptob, scm_pipob): We don't need casts here any more.
* strports.c (prinstpt): Use prototype declarations.
(stputc, stwrite, stputs, stgetc): Take the port as an argument,
and use SCM_STREAM to get the string info.  Also, use prototyped
definitions, and get rid of the extra declarations.
* vports.c (sfputc, sfwrite, sfputs, sfflush, sfgetc, sfclose,
noop0): Same.
* fports.c (local_fgetc, local_fgets): Renamed from scm_fgetc and
scm_fgets, for consistency.
(scm_fptop, scm_pipob): References updated.

25 years agoChange the definition of the functions in scm_ptobfuns so that
Jim Blandy [Fri, 9 Oct 1998 10:00:55 +0000 (10:00 +0000)]
Change the definition of the functions in scm_ptobfuns so that
they get passed the port object, not the port's stream.
* ports.h (scm_ptobfuns): Rename all `stream' arguments to `port'.
* gc.c (scm_gc_sweep): Pass the port itself to the free function.
* genio.c (scm_putc, scm_puts, scm_lfwrite, scm_fflush, scm_getc):
Pass the port itself to the scm_ptobs function.
* ports.c (scm_close_port, scm_force_output, scm_flush_all_ports,
scm_generic_fgets): Same.
(putc_void_port, puts_void_port, write_void_port, flush_void_port,
getc_void_port, fgets_void_port, close_void_port): Just change the
argument names; these functions don't really do anything.
* fports.c (local_fgetc, local_fgets, local_fclose, local_fflush,
local_fputc, local_fputs, local_ffwrite, local_pclose): Take the
port as an argument, and use SCM_STREAM to get the stdio FILE *.
Also, use prototyped definitions, and get rid of the extra
declarations.
(scm_fptob, scm_pipob): We don't need casts here any more.
* strports.c (prinstpt): Use prototype declarations.
(stputc, stwrite, stputs, stgetc): Take the port as an argument,
and use SCM_STREAM to get the string info.  Also, use prototyped
definitions, and get rid of the extra declarations.
* vports.c (sfputc, sfwrite, sfputs, sfflush, sfgetc, sfclose,
noop0): Same.
* ports.h (scm_ptobfuns): Replace uses of SCM_P with a straight
prototype; it's okay (preferred, even!) to use ANSI C in Guile.

25 years ago*** empty log message ***
Jim Blandy [Thu, 8 Oct 1998 20:17:46 +0000 (20:17 +0000)]
*** empty log message ***

25 years ago* .cvsignore: New file, containing data-rep.info. I'm not sure
Jim Blandy [Thu, 8 Oct 1998 20:17:38 +0000 (20:17 +0000)]
* .cvsignore: New file, containing data-rep.info.  I'm not sure
whether we want to check this file into CVS, because it's
generated; if you find compelling reasons it should be, let me
know.

25 years agoAdjust formatting.
Jim Blandy [Thu, 8 Oct 1998 20:12:53 +0000 (20:12 +0000)]
Adjust formatting.

25 years ago*** empty log message ***
Jim Blandy [Thu, 8 Oct 1998 20:12:22 +0000 (20:12 +0000)]
*** empty log message ***

25 years agoInclude the source location in error messages for scripts.
Jim Blandy [Thu, 8 Oct 1998 20:11:59 +0000 (20:11 +0000)]
Include the source location in error messages for scripts.
* init.c (scm_boot_guile_1): Use scm_internal_lazy_catch, so the
stack is still there when we catch the error.
* throw.c (handler_message): If we are handling an error with a
message, then put together the proper arguments and call
scm_display_error, instead of scm_display_error_message.  That
displays source location, if it can find it.

25 years ago*** empty log message ***
Jim Blandy [Wed, 7 Oct 1998 20:06:29 +0000 (20:06 +0000)]
*** empty log message ***

25 years ago* gc.c (scm_unprotect_object): Change this so that calls to
Jim Blandy [Wed, 7 Oct 1998 20:05:56 +0000 (20:05 +0000)]
* gc.c (scm_unprotect_object): Change this so that calls to
scm_protect_object and scm_unprotect_object nest properly.
(scm_protect_object): Doc fixes.

25 years ago*** empty log message ***
Jim Blandy [Wed, 7 Oct 1998 20:00:11 +0000 (20:00 +0000)]
*** empty log message ***

25 years ago*** empty log message ***
Jim Blandy [Wed, 7 Oct 1998 19:35:56 +0000 (19:35 +0000)]
*** empty log message ***

25 years ago*** empty log message ***
Jim Blandy [Wed, 7 Oct 1998 15:52:46 +0000 (15:52 +0000)]
*** empty log message ***

25 years ago* strings.c (scm_string_set_x): Require the argument to be a
Jim Blandy [Wed, 7 Oct 1998 15:52:31 +0000 (15:52 +0000)]
* strings.c (scm_string_set_x): Require the argument to be a
writable string, not a substring or a symbol.
* strings.h (SCM_RWSTRINGP, SCM_NRWSTRINGP): New predicates.

25 years ago*** empty log message ***
Jim Blandy [Wed, 7 Oct 1998 10:16:28 +0000 (10:16 +0000)]
*** empty log message ***

25 years ago* configure.in: Don't forget to #define HAVE_RL_GETC_FUNCTION if
Jim Blandy [Wed, 7 Oct 1998 10:15:57 +0000 (10:15 +0000)]
* configure.in: Don't forget to #define HAVE_RL_GETC_FUNCTION if
we do find the rl_getc_function variable in the readline library;
AC_CHECK_FUNCS used to do this for us, but we're not using it any
more.
* acconfig.h: Add an entry for HAVE_RL_GETC_FUNCTION.

25 years ago* scmconfig.h.in: Regenerated; ../acconfig.h has changed.
Jim Blandy [Wed, 7 Oct 1998 10:15:44 +0000 (10:15 +0000)]
* scmconfig.h.in: Regenerated; ../acconfig.h has changed.

25 years ago*** empty log message ***
Jim Blandy [Wed, 7 Oct 1998 10:03:07 +0000 (10:03 +0000)]
*** empty log message ***

25 years ago* eval.c (safe_setjmp): Remove this misunderstanding.
Jim Blandy [Wed, 7 Oct 1998 10:02:34 +0000 (10:02 +0000)]
* eval.c (safe_setjmp): Remove this misunderstanding.
(ENTER_APPLY, SCM_CEVAL, SCM_APPLY): Replace with references to
ordinary setjmp.

25 years ago* configure.in: Don't forget to #define HAVE_RL_GETC_FUNCTION if
Jim Blandy [Wed, 7 Oct 1998 09:57:16 +0000 (09:57 +0000)]
* configure.in: Don't forget to #define HAVE_RL_GETC_FUNCTION if
we do find the rl_getc_function variable in the readline library;
AC_CHECK_FUNCS used to do this for us, but we're not using it any
more.

25 years ago*** empty log message ***
Jim Blandy [Wed, 7 Oct 1998 09:56:43 +0000 (09:56 +0000)]
*** empty log message ***

25 years ago* configure.in: Properly test for the presence of rl_getc_function;
Jim Blandy [Wed, 7 Oct 1998 09:09:47 +0000 (09:09 +0000)]
* configure.in: Properly test for the presence of rl_getc_function;
it's a variable, not a function.
* configure: Regenerated.

25 years ago* configure.in: Properly test for the presence of rl_getc_function;
Jim Blandy [Wed, 7 Oct 1998 09:02:39 +0000 (09:02 +0000)]
* configure.in: Properly test for the presence of rl_getc_function;
it's a variable, not a function.
* configure: Regenerated.

25 years agoRegenerated, after adding ChangeLog.
Jim Blandy [Wed, 7 Oct 1998 09:02:25 +0000 (09:02 +0000)]
Regenerated, after adding ChangeLog.

25 years ago*** empty log message ***
Jim Blandy [Wed, 7 Oct 1998 07:37:40 +0000 (07:37 +0000)]
*** empty log message ***

25 years ago* New directory for documentation.
Jim Blandy [Wed, 7 Oct 1998 07:37:17 +0000 (07:37 +0000)]
* New directory for documentation.
* README: New file.
* data-rep.texi: It's not a real manual, but it's better than
nothing.
* Makefile.am, Makefile.in, data-rep.info, data-rep.texi,
mdate-sh, stamp-vti, texinfo.tex, version.texi: The usual support
files.

25 years ago* doc: New subdirectory.
Jim Blandy [Wed, 7 Oct 1998 07:36:55 +0000 (07:36 +0000)]
* doc: New subdirectory.
* Makefile.am (SUBDIRS): List it.
* configure.in (AC_OUTPUT): Build its Makefile.
* configure, Makefile.in: Regenerated.

25 years ago*** empty log message ***
Jim Blandy [Wed, 7 Oct 1998 07:17:15 +0000 (07:17 +0000)]
*** empty log message ***

25 years ago*** empty log message ***
Jim Blandy [Tue, 6 Oct 1998 22:07:55 +0000 (22:07 +0000)]
*** empty log message ***

25 years ago* guile-config.in (build-link): It isn't. Revert the change.
Jim Blandy [Tue, 6 Oct 1998 22:07:49 +0000 (22:07 +0000)]
* guile-config.in (build-link): It isn't.  Revert the change.

25 years ago*** empty log message ***
Jim Blandy [Tue, 6 Oct 1998 22:00:25 +0000 (22:00 +0000)]
*** empty log message ***

25 years ago* guile-config.in (build-link): Include a -R flag in the output
Jim Blandy [Tue, 6 Oct 1998 22:00:13 +0000 (22:00 +0000)]
* guile-config.in (build-link): Include a -R flag in the output
from link.  Not sure if this is the right thing to do.

25 years ago* guile.m4 (GUILE_FLAGS): New macro.
Jim Blandy [Tue, 6 Oct 1998 21:59:54 +0000 (21:59 +0000)]
* guile.m4 (GUILE_FLAGS): New macro.
* guile.m4 (AM_INIT_GUILE_MODULE): Deleted; it doesn't do anything
terribly helpful any more, nobody's using it, and this is not
really the way I want to handle modules anyway.

25 years agoFormatting tweaks.
Jim Blandy [Tue, 6 Oct 1998 20:57:10 +0000 (20:57 +0000)]
Formatting tweaks.

25 years ago*** empty log message ***
Jim Blandy [Tue, 6 Oct 1998 15:47:07 +0000 (15:47 +0000)]
*** empty log message ***

25 years ago* libguile.h: Mark these as C declarations, for compilation by C++
Jim Blandy [Tue, 6 Oct 1998 15:46:06 +0000 (15:46 +0000)]
* libguile.h: Mark these as C declarations, for compilation by C++
compilers.

25 years ago* snarf.h (SCM_PROC, SCM_PROC1): Remove very odd code in #ifdef
Jim Blandy [Tue, 6 Oct 1998 13:01:46 +0000 (13:01 +0000)]
* snarf.h (SCM_PROC, SCM_PROC1): Remove very odd code in #ifdef
__cplusplus clause.  I seriously doubt this ever worked the way
the author seems to have intended.