The commit "don't take string-write mutex in read.c:read_token", from
8b0d7b9d94, had a number of bugs. Not sure
how I missed these before.
* libguile/read.c (read_token): Remove a couple of bogus
scm_i_string_stop_writing () calls, now that we no longer take the
string-write mutex.
(read_complete_token): read_token really needs a fresh buffer, which
was not the case when we are reading long tokens and thus hit the
overflow case. Fixes fractions.test.
When converting a character into a character name, an incorrect
character name could theoretically have been returned.
* libguile/chars.c (scm_i_charname): used wrong index into
scm_alt_charnames
* configure.ac: Check availability of getsid() C library call.
* libguile/posix.c (scm_getsid): New primitive.
* libguile/posix.h: Declaration for new primitive.
* libguile/chars.c, libguile/chars.h (scm_char_general_category): New function.
* test-suite/tests/chars.test: Unit tests for `char-general-category'.
* doc/ref/api-data.texi (Characters): Documentation for
`char-general-category'.
* libguile/vm-engine.c (VM_NAME): Keyword arg errors are now thrown to
'keyword-argument-error.
* libguile/vm.c: Define sym_keyword_argument_error, and statically
allocate some other symbols.
* module/ice-9/optargs.scm (parse-lambda-case): Throw to
'keyword-argument-error in kwarg error cases.
* module/ice-9/psyntax.scm (build-lambda-case): Remove a couple
workarounds for the old memoizer. Throw to 'wrong-number-of-args if
the lambda-case fails to parse.
* module/ice-9/psyntax-pp.scm: Regenerated.
* test-suite/tests/optargs.test: Update expected exceptions.
* doc/ref/api-data.texi (Characters): Documentation for `char-titlecase'.
* doc/ref/api-i18n.texi (Character Case Mapping): Documentation for
`char-locale-titlecase' and `string-locale-titlecase'.
* libguile/chars.c, libguile/chars.h (scm_char_titlecase, scm_c_titlecase): New
functions.
* libguile/i18n.c, libguile/i18n.h (chr_to_case, scm_char_locale_titlecase,
str_to_case, scm_string_locale_titlecase): New functions.
* libguile/i18n.c (scm_char_locale_downcase, scm_char_locale_upcase,
scm_string_locale_downcase, scm_string_locale_upcase): Refactor to share code
via chr_to_case and str_to_case, as appropriate.
* module/ice-9/i18n.scm (char-locale-title-case, string-locale-titlecase): New
functions.
* libguile/srfi-13.c (string_titlecase_x): Use uc_totitle instead of uc_toupper.
* test-suite/tests/chars.test: Tests for `char-titlecase'.
* test-suite/tests/i18n.test: Tests for `char-locale-titlecase' and
`string-locale-titlecase'.
* test-suite/tests/srfi-13.test: Tests for `string-titlecase'.
* libguile/read.c (read_token): Don't take the string-write mutex when
reading a token into a buffer, because it's assumed that the buffer is
fresh (not seen by other threads), and a soft port can call a
procedure that needs the string-write mutex.
* libguile/vm-engine.c (VM_NAME): Engines take the VM itself (not the
vp), so they can pass the VM to hooks. No more hook args, we dispatch
without them.
* libguile/vm-engine.h (RUN_HOOK): Dispatch the hook if the trace level
is positive (instead of if the hook is there). Don't cache registers
on return from the dispatch.
* libguile/vm.h:
* libguile/vm.c (vm_dispatch_hook): Don't bother with a dynwind; instead
decrement the trace level when going into a hook, and if we have a
nonlocal exit, the trace level never gets incremented again. Worse is
better.
(make_vm, scm_vm_trace_level, scm_set_vm_trace_level_x): New concept,
trace level. If positive, we run the hooks, otherwise we don't. Should
work. Removed scm_vm_trace_frame, I don't think that was the right way
to do it.
* module/system/vm/vm.scm: Replace vm-trace-frame with vm-trace-level
and set-vm-trace-level!; the hooks actually get the frame as an
argument now.
* libguile/eval.h:
* libguile/eval.c (scm_call_n): New function, applies a function to an
array of args.
* libguile/hooks.h:
* libguile/hooks.c (scm_c_run_hookn): New function, runs a hook with an
array of args.
* libguile/goops.h:
* libguile/goops.c (scm_add_method): Remove this function. It didn't
work, and even if we fixed it to work it wouldn't be as good as using
Scheme's `add-method!'.
* libguile/load.c (scm_init_load_path): Search first in Guile's
version-specific path (the library path), *then* in the site dir. This
should allow us to import some of Guile-lib's modules into Guile,
perhaps customizing them with Guile-version-specific foo, without
accidentally falling back to system-installed guile-lib in the
sitedir. Also should reduce the number of stats by half, in the common
case.
They made Sun C 5.8 emit a warning such as:
line 71: warning: dead part of constant expression is nonconstant
* libguile/print.c (scm_print_opts): Don't use `SCM_UNPACK ()' here.
* libguile/read.c (scm_read_opts): Likewise.
* libguile/frames.c, libguile/objcodes.c, libguile/programs.c,
libguile/vm-engine.c, libguile/vm-i-system.c, libguile/vm.c: Use
`SCM_C_OBJCODE_BASE ()' instead of accessing the `base' field of
`struct scm_objcode'.
* libguile/objcodes.h (struct scm_objcode)[base]: Remove.
* libguile/memoize.c (MAKMEMO_CALL): Memoize in the number of arguments
at the call site.
(memoize, scm_m_cond, memoize_named_let, unmemoize):
* libguile/eval.c (eval): Adapt to changes in call memoization.
* module/ice-9/eval.scm (primitive-eval): For calls, pass the first N
arguments directly on the stack, and only the rest as a consed
argument list to apply. Currently N is 4.
* libguile/vm.h (struct scm_vm): Remove "time" and "clock" members. The
time was bogusly measured, and the "clock" measured instructions
retired, which is not a very useful measurement, and it was causing
lots of memory accesses. Not that I have done a proper profile,
though...
(scm_vm_stats): Remove this procedure, which provided access to "time"
and "clock".
* libguile/vm.c:
* libguile/vm-engine.h:
* libguile/vm-engine.c:
* libguile/vm-i-system.c: Adapt to scm_vm changes and scm_vm_stats
removal.
* module/system/repl/command.scm:
* module/system/vm/vm.scm: Adapt to vm-stats removal by removing
vm-stats from <repl>.
* libguile/__scm.h (SCM_ASYNC_TICK): Add some branch prediction.
(SCM_ASYNC_TICK_WITH_CODE): New helper for when BUILDING_LIBGUILE,
runs code only if we're going to call async_click().
* libguile/vm-engine.h (VM_HANDLE_INTERRUPTS): New helper, uses
SCM_ASYNC_TICK_WITH_CODE to only save regs if we'll handle an
interrupt.
* libguile/vm-i-system.c (call, goto/args, return): use
VM_HANDLE_INTERRUPTS.
* module/language/elisp/compile-tree-il.scm: Update for changes to
tree-il (lambda-case, mainly).
* module/language/elisp/spec.scm: Update GPL version to 3. Update reader
for new taking a port and environment argument.
* libguile/_scm.h: Bump objcode version.
* libguile/vm-i-system.c: Fix conflicts.
* module/Makefile.am: Fix conflicts, and add elisp modules to the build.
* libguile/load.c (scm_init_eval_in_scheme): If we have no eval.go, null
out the load-compiled path so that we load no compiled code. Prevents
partially compiled Guile from turning tail-calls into stack-munching
calls.
* libguile/smob.h (SCM_SMOB_TYPE_BITS, SCM_SMOB_TYPE_MASK): New macros,
for when you want to identify a SMOB by type and mask.
* libguile/arrays.c:
* libguile/bitvectors.c:
* libguile/srfi-4.c: Use the new macros.
* libguile/struct.h (SCM_VTABLE_INSTANCE_FINALIZER): Use DATA_REF, not
SLOT_REF.
(SCM_SET_VTABLE_INSTANCE_FINALIZER, SCM_SET_VTABLE_INSTANCE_PRINTER):
New accessor macros.
* libguile/struct.h (scm_standard_vtable_vtable)
(scm_applicable_struct_vtable_vtable)
(scm_applicable_struct_with_setter_vtable_vtable)
* libguile/struct.c: Make these stock meta-tables public to C.
* libguile/modules.c (SCM_F_EVAL_CLOSURE_INTERFACE)
(SCM_EVAL_CLOSURE_INTERFACE_P, scm_standard_interface_eval_closure):
Fix to not use SCM_CELL macros.
