* doc/ref/api-debug.texi (VM Hooks): Update documentation.
* libguile/vm.c (vm_dispatch_hook):
* libguile/vm-engine.c: Rework the hook machinery so that they can
receive an arbitrary number of arguments. The return and abort
hooks will pass the values that they return to their continuations.
(vm_engine): Adapt to ABORT_CONTINUATION_HOOK change.
* libguile/vm-i-system.c (return, return/values): Adapt to
POP_CONTINUATION_HOOK change.
* module/system/vm/frame.scm (frame-return-values): Remove. The
pop-continuation-hook will pass the values directly.
* module/system/vm/trace.scm (print-return):
(trace-calls-to-procedure):
(trace-calls-in-procedure): Update to receive return values
directly.
* module/system/vm/traps.scm (trap-in-procedure)
(trap-in-dynamic-extent): Ignore return values.
(trap-frame-finish, trap-calls-in-dynamic-extent)
(trap-calls-to-procedure): Pass return values to the handlers.
* libguile/vm-engine.c: Remove the register assignments inherited from
the 1990s. GCC does seem to allocate reasonably on systems with
enough registers (e.g. x86-64), and on system with too few (ia32) we
disabled manual allocation. Anyway this code was never tested, so
it's better to leave the compiler to do its own thing, until proven
otherwise. Also in the RTL VM we don't need to allocate a register to
the SP, because it isn't accessed as much.
* libguile/vm-engine.c: Remove the ability for the VM to check object
access, free variable access, and the ip. They were off by default.
Since they will be different in the RTL VM, their presence is just
making things confusing.
* libguile/vm.c: Remove corresponding error helpers.
* libguile/objcodes.c (register_elf, scm_find_mapped_elf_image): New
interfaces that keep a list of all ELF mappings. Exported from the
(system vm objcode) module.
* module/system/vm/objcode.scm: Export find-mapped-elf-image.
* libguile/objcodes.c (sniff_elf_alignment, alloc_aligned)
(copy_and_align_elf_data): New helpers for portably re-aligning ELF
data from read(2) or from a bytevector.
(load_thunk_from_memory): Simplify! Now there is only one procedure
that loads ELF, and it does less: it simply receives the whole image
in one array, hopefully from mmap.
(scm_load_thunk_from_file): Use new map_file_contents helper, and go
through load_thunk_from_memory.
(scm_load_thunk_from_memory): Pass load_thunk_from_memory a piece of
memory that it owns, and that is appropriately aligned.
* libguile/numbers.h (scm_to_intptr_t, scm_from_intptr_t)
(scm_to_uintptr_t, scm_from_uintptr_t): New defines.
* libguile/foreign.c: Remove definitions here; adapt callers to use new
names with _t suffixes.
* libguile/stime.c: Remove #ifdef HAVE_SYS_TIMES_H, HAVE_SYS_TYPES_H,
and HAVE_STRING_H since those headers are always provided by Gnulib
when the corresponding macro is undefined. Fixes a regression on
MinGW introduced by commit 40e2a31.
* libguile/stime.c (scm_init_stime): Fix previous patch. At this point
in the source, we definitely have a run-time function, but don't
necessarily have a real-time function.
* configure.ac: Remove checks for pipe, stftime, times, and select, as
our Gnulib modules already ensure they are provided.
* libguile/stime.c: Rely on gnulib providing `times'.
* libguile/filesys.c (scm_sendfile): In Linux-style sendfile(2) code, if
EINTR or EAGAIN occurs, set result to 1 (not 0) so that we actually
keep going. In non-sendfile(2) code, deal gracefully with short reads
due to EOF.
* test-suite/tests/filesys.test ("sendfile"): Use 'let*' to guarantee
the needed order of operations: write (test-file) and then read it.
Add code to check the written data (not just the returned length) in
all tests, including the cases that hit EOF prematurely.
* libguile/filesys.c (scm_sendfile)[HAVE_SYS_SENDFILE_H &&
HAVE_SENDFILE]: Compare RESULT with C_COUNT. Loop until C_COUNT bytes
have been sent.
* doc/ref/posix.texi (File System): Update the description. Explain the
new semantics.
* test-suite/tests/filesys.test ("sendfile"): Rewrite using
`pass-if-equal'. Check the return value for all the tests.
["file with offset past the end", "file with offset near the end"]:
New tests.
* libguile/fports.c (scm_open_file_with_encoding): New API function,
containing the code previously found in 'scm_open_file', but modified
to accept the new 'guess_encoding' and 'encoding' arguments.
(scm_open_file): Now just a simple wrapper that calls
'scm_open_file_with_encoding'.
(scm_i_open_file): New implementation of 'open-file' that accepts
keyword arguments '#:guess-encoding' and '#:encoding', and calls
'scm_open_file_with_encoding'.
(scm_init_fports_keywords): New initialization function that gets
called after keywords are initialized.
* libguile/fports.h (scm_open_file_with_encoding,
scm_init_fports_keywords): Add prototypes.
* libguile/init.c (scm_i_init_guile): Call 'scm_init_fports_keywords'.
* module/ice-9/boot-9.scm: Add enhanced versions of 'open-input-file',
'open-output-file', 'call-with-input-file', 'call-with-output-file',
'with-input-from-file', 'with-output-to-file', and
'with-error-to-file', that accept keyword arguments '#:binary',
'#:encoding', and (for input port constructors) '#:guess-encoding'.
* doc/ref/api-io.texi (File Ports): Update documentation.
* test-suite/tests/ports.test ("keyword arguments for file openers"):
Add tests.
* libguile/ports.c (scm_i_unget_bytes): New static function.
(scm_unget_bytes): New API function.
(scm_unget_byte): Rewrite to simply call 'scm_i_unget_bytes'.
(scm_ungetc, scm_peek_char, looking_at_bytes): Use 'scm_i_unget_bytes'.
* libguile/ports.h: Add prototype for 'scm_unget_bytes'.
* libguile/fports.c (scm_setvbuf): Use 'scm_unget_bytes'.
* libguile/r6rs-ports.c (scm_unget_bytevector): New procedure.
* module/ice-9/binary-ports.scm (unget-bytevector): New export.
* doc/ref/api-io.texi (R6RS Binary Input): Add documentation.
(R6RS I/O Ports): Update brief description of (ice-9 binary-ports) to
reflect the new reality: it is no longer a subset of (rnrs io ports).
* test-suite/tests/ports.test ("unget-bytevector"): Add test.
* libguile/fports.c (scm_open_file): Do not scan for coding
declarations. Replace 'use_encoding' local variable with
'binary'. Update documentation string.
* module/ice-9/psyntax.scm (include): Add the same file-encoding
logic that's used in compile-file and scm_primitive_load.
* module/ice-9/psyntax-pp.scm: Regenerate.
* doc/ref/api-io.texi (File Ports): Update docs.
* test-suite/tests/ports.test: Change "open-file HONORS file coding
declarations" test to "open-file IGNORES file coding declaration".
* test-suite/tests/coding.test (scan-coding): Use 'file-encoding' to
scan for the encoding, since 'open-input-file' no longer does so.
* libguile/keywords.c (scm_keyword_argument_error): New variable.
(scm_c_bind_keyword_arguments): New API function.
* libguile/keywords.h (enum scm_keyword_arguments_flags): New enum.
(scm_t_keyword_arguments_flags): New typedef.
(scm_c_bind_keyword_arguments): New prototype.
* doc/ref/api-data.texi (Coding With Keywords, Keyword Procedures): Add
documentation.
* test-suite/standalone/test-scm-c-bind-keyword-arguments.c: New file.
* test-suite/standalone/Makefile.am: Add
test-scm-c-bind-keyword-arguments test.
* libguile/array-map.c: new function rafill, like scm_array_fill_int,
but factors GVSET out of the loop. Use it in scm_array_fill_x instead of
scm_array_fill_int.
* test-suite/tests/arrays.test: add test for array-fill! with stride != 1.
Signed-off-by: Ludovic Courtès <ludo@gnu.org>
* libguile/ports-internal.h (struct scm_port_internal): Add new members
'at_stream_start_for_bom_read' and 'at_stream_start_for_bom_write'.
(SCM_UNICODE_BOM): New macro.
(scm_i_port_iconv_descriptors): Add 'mode' parameter to prototype.
* libguile/ports.c (scm_new_port_table_entry): Initialize
'at_stream_start_for_bom_read' and 'at_stream_start_for_bom_write'.
(get_iconv_codepoint): Pass new 'mode' parameter to
'scm_i_port_iconv_descriptors'.
(get_codepoint): After reading a codepoint at stream start, record
that we're no longer at stream start, and consume a BOM where
appropriate.
(scm_seek): Set the stream start flags according to the new position.
(looking_at_bytes): New static function.
(scm_utf8_bom, scm_utf16be_bom, scm_utf16le_bom, scm_utf32be_bom,
scm_utf32le_bom): New static const arrays.
(decide_utf16_encoding, decide_utf32_encoding): New static functions.
(scm_i_port_iconv_descriptors): Add new 'mode' parameter. If the
specified encoding is UTF-16 or UTF-32, make that precise by deciding
what byte order to use, and construct iconv descriptors based on the
precise encoding.
(scm_i_set_port_encoding_x): Record that we are now at stream start.
Do not open the new iconv descriptors immediately; let them be
initialized lazily.
* libguile/print.c (display_string_using_iconv): Record that we're no
longer at stream start. Write a BOM if appropriate.
* doc/ref/api-io.texi (BOM Handling): New node.
* test-suite/tests/ports.test ("set-port-encoding!, wrong encoding"):
Adapt test to cope with the fact that 'set-port-encoding!' does not
immediately open the iconv descriptors.
(bv-read-test): New procedure.
("unicode byte-order marks (BOMs)"): New test prefix.
* libguile/ports.c (scm_i_set_port_encoding_x): Always copy the
user-provided port encoding string, so that its case will be preserved
and returned exactly by subsequent calls to 'port-encoding'.
* libguile/array-map.c: (racp): factor scm_generalized_vector_ref,
scm_generalized_vector_set_x out of the rank-1 loop.
Signed-off-by: Ludovic Courtès <ludo@gnu.org>
* libguile/ports.c (scm_i_get_byte_or_eof, scm_i_peek_byte_or_eof):
Rename to 'scm_slow_get_byte_or_eof' and 'scm_slow_peek_byte_or_eof',
respectively.
* libguile/ports.h (scm_i_get_byte_or_eof, scm_i_peek_byte_or_eof):
Rename to 'scm_slow_get_byte_or_eof' and 'scm_slow_peek_byte_or_eof',
respectively, and mark them as SCM_API.
* libguile/inline.h (scm_get_byte_or_eof, scm_peek_byte_or_eof): Adjust
to use the new names.
* libguile/ports.c (get_iconv_codepoint): Rewrite to fix a bug and
improve efficiency and clarity. Previously, it incorrectly assumed
that iconv would never consume input without producing output, which
led to a buffer overrun and subsequent assertion failure. This
happens when a byte-order mark is consumed by iconv at the beginning
of the stream when using the UTF-16 or UTF-32 encodings.
* test-suite/tests/ports.test (unicode byte-order marks (BOMs)):
Add tests.
Suggested by Andy Wingo.
* libguile/inline.h (scm_get_byte_or_eof, scm_peek_byte_or_eof): Keep
only the fast path here, with fallback to 'scm_i_get_byte_or_eof' and
'scm_i_peek_byte_or_eof'.
* libguile/ports.c (scm_i_get_byte_or_eof, scm_i_peek_byte_or_eof):
New internal functions.
* libguile/ports.h (scm_i_get_byte_or_eof, scm_i_peek_byte_or_eof): Add
prototypes.
* libguile/ports.c (scm_i_fill_input): New static function, containing
the code that was previously in 'scm_fill_input'.
(scm_fill_input): Simply call 'scm_i_fill_input'.
(scm_c_read): Use 'scm_i_fill_input'.
* libguile/ports-internal.h (struct scm_port_internal): Add 'alist'
member.
* libguile/ports.c (scm_i_port_alist, scm_i_set_port_alist_x): New
internal functions.
(scm_i_port_weak_hash): Update comment: the hash table is no longer
used to store the port's alist.
(scm_new_port_table_entry): Initialize 'alist'. Store SCM_BOOL_F in
the port weak hash, not SCM_EOL.
* libguile/ports.h (scm_i_port_alist, scm_i_set_port_alist_x): Add
protoypes.
* libguile/read.c (set_port_read_option, init_read_options): Access the
port's alist via 'scm_i_port_alist' and 'scm_i_set_port_alist_x'.
* libguile/r6rs-ports.c (scm_get_bytevector_some): Rewrite to
efficiently take the contents of the read/putback buffers. In the
docstring, clarify that it might not return all available bytes.
* doc/ref/api-io.texi (R6RS Binary Input): Clarify that
'get-bytevector-some' might not return all available bytes.
* test-suite/tests/r6rs-ports.test ("get-bytevector-some [only-some]"):
Remove bogus test, which requires more than the R6RS requires.
Based on 6c98257f2e by Andy Wingo.
* libguile/ports-internal.h (struct scm_port_internal): Add a flag
for the port encoding mode: UTF8 or iconv. The iconv descriptors
are now in a separate structure so that we can avoid attaching
finalizers to the ports themselves in the future.
(enum scm_port_encoding_mode): New enum.
(struct scm_iconv_descriptors): New struct.
(scm_i_port_iconv_descriptors): Add prototype.
* libguile/ports.c (finalize_port): Don't close iconv descriptors here.
(scm_new_port_table_entry): Adapt to the iconv descriptors being
moved. Initialize 'encoding_mode'.
(scm_i_remove_port): Adapt to call 'close_iconv_descriptors'.
(close_iconv_descriptors): New static function.
(get_iconv_codepoint): Use 'scm_i_port_iconv_descriptors'.
(get_codepoint): Check the port 'encoding_mode'.
(finalize_iconv_descriptors, open_iconv_descriptors,
close_iconv_descriptors, scm_i_port_iconv_descriptors): New static
functions.
(scm_i_set_port_encoding_x): Adapt to iconv descriptors being moved
to separate structure, to set the 'encoding_mode' flag, and to use
'open_iconv_descriptors' and 'close_iconv_descriptors'.
* libguile/print.c (display_string_using_iconv): Use
'scm_i_port_iconv_descriptors'.
(display_string): Use 'encoding_mode' flag.
* libguile/ports-internal.h: New file.
* libguile/Makefile.am (noinst_HEADERS): Add ports-internal.h.
* libguile/ports.h (scm_t_port): Add a comment mentioning that the
'input_cd' and 'output_cd' fields of the public structure are no
longer what they seem to be.
* libguile/ports.c: Include ports-internal.h.
(finalize_port, scm_i_remove_port, get_iconv_codepoint, get_codepoint,
scm_i_set_port_encoding_x): Access 'input_cd' and 'output_cd' via the
new internal port structure.
(scm_new_port_table_entry): Allocate and initialize the internal port
structure.
* libguile/print.c: Include ports-internal.h.
(display_string_using_iconv, display_string): Access 'input_cd' and
'output_cd' via 'internal' pointer.
* libguile/strings.c (scm_string_append): Check for numerical overflow
while computing the length of the result. Double-check that we don't
overflow the result string, and that it is the correct length in the
end (in case another thread changed the list). When copying a narrow
string to a wide result, avoid calling 'scm_i_string_length' and
'scm_i_string_chars' on each character.
