Have NATNUMP give t for positive bignums; check limits appropriately. src/ChangeLog addition: 2010-11-20 Aidan Kehoe <kehoea@parhasard.net> * abbrev.c (Fexpand_abbrev): * alloc.c: * alloc.c (Fmake_list): * alloc.c (Fmake_vector): * alloc.c (Fmake_bit_vector): * alloc.c (Fmake_byte_code): * alloc.c (Fmake_string): * alloc.c (vars_of_alloc): * bytecode.c (UNUSED): * bytecode.c (Fbyte_code): * chartab.c (decode_char_table_range): * cmds.c (Fself_insert_command): * data.c (check_integer_range): * data.c (Fnatnump): * data.c (Fnonnegativep): * data.c (Fstring_to_number): * elhash.c (hash_table_size_validate): * elhash.c (decode_hash_table_size): * eval.c (Fbacktrace_frame): * event-stream.c (lisp_number_to_milliseconds): * event-stream.c (Faccept_process_output): * event-stream.c (Frecent_keys): * event-stream.c (Fdispatch_event): * events.c (Fmake_event): * events.c (Fevent_timestamp): * events.c (Fevent_timestamp_lessp): * events.h: * events.h (struct command_builder): * file-coding.c (gzip_putprop): * fns.c: * fns.c (check_sequence_range): * fns.c (Frandom): * fns.c (Fnthcdr): * fns.c (Flast): * fns.c (Fnbutlast): * fns.c (Fbutlast): * fns.c (Fmember): * fns.c (Ffill): * fns.c (Freduce): * fns.c (replace_string_range_1): * fns.c (Freplace): * font-mgr.c (Ffc_pattern_get): * frame-msw.c (msprinter_set_frame_properties): * glyphs.c (check_valid_xbm_inline): * indent.c (Fmove_to_column): * intl-win32.c (mswindows_multibyte_to_unicode_putprop): * lisp.h: * lisp.h (ARRAY_DIMENSION_LIMIT): * lread.c (decode_mode_1): * mule-ccl.c (ccl_get_compiled_code): * number.h: * process-unix.c (unix_open_multicast_group): * process.c (Fset_process_window_size): * profile.c (Fstart_profiling): * unicode.c (Funicode_to_char): Change NATNUMP to return 1 for positive bignums; changes uses of it and of CHECK_NATNUM appropriately, usually by checking for an integer in an appropriate range. Add array-dimension-limit and use it in #'make-vector, #'make-string. Add array-total-size-limit, array-rank-limit while we're at it, for the sake of any Common Lisp-oriented code that uses these limits. Rename check_int_range to check_integer_range, have it take Lisp_Objects (and thus bignums) instead. Remove bignum_butlast(), just set int_n to an appropriately large integer if N is a bignum. Accept bignums in check_sequence_range(), change the functions that use check_sequence_range() appropriately. Move the definition of NATNUMP() to number.h; document why it's a reasonable name, contradicting an old comment. tests/ChangeLog addition: 2010-11-20 Aidan Kehoe <kehoea@parhasard.net> * automated/lisp-tests.el: * automated/lisp-tests.el (featurep): * automated/lisp-tests.el (wrong-type-argument): * automated/mule-tests.el (featurep): Check for args-out-of-range errors instead of wrong-type-argument errors in various places when code is handed a large bignum instead of a fixnum. Also check for the wrong-type-argument errors when giving the same code a non-integer value.

This directory contains a number of XEmacs dynamic modules.  These
modules can be loaded directly with the command 'M-x load-module'.
However, the preferred method of loading a module is to issue a
"(require 'module-name)" command to the Lisp interpreter.  This will
store information so that a later "(unload-feature 'module-name)" can
succeed.

To compile one of these modules, simply enter the desired directory,
type 'configure', and then 'make'.  If you are building the module for
an installed XEmacs, then 'make install' will place the module in the
appropriate directory for XEmacs to find it later (assuming you have
permission to write to that directory).  A subsequent 'load-module' or
'require' will then load the module, as described above.

Each of these demonstrates different features and limitations of the
XEmacs module loading technology.  For a complete discussion on XEmacs
dynamic modules, please consult the XEmacs Module Writers Guide, which
can be found in the ../info directory.

For those wanting to get started with module writing, please see the
'sample' directory.  It contains two subdirectories: internal and
external.  The 'internal' subdirectory contains the framework needed to
migrate some core piece of XEmacs functionality into code that can
either be compiled into the core or built as a separate module.  The
'external' subdirectory contains the somewhat simpler framework needed
to build a module separately from XEmacs.  These should be considered
starting places for module writing.