Mercurial > hg > xemacs-beta
view src/number-gmp.c @ 4604:e0a8715fdb1f
Support new IGNORE-INVALID-SEQUENCESP argument, #'query-coding-region.
lisp/ChangeLog addition:
2009-02-07 Aidan Kehoe <kehoea@parhasard.net>
* coding.el (query-coding-clear-highlights):
Rename the BUFFER argument to BUFFER-OR-STRING, describe it as
possibly being a string in its documentation.
(default-query-coding-region):
Add a new IGNORE-INVALID-SEQUENCESP argument, document that this
function does not support it.
Bind case-fold-search to nil, we don't want this to influence what the
function thinks is encodable or not.
(query-coding-region):
Add a new IGNORE-INVALID-SEQUENCESP argument, document what it
does; reflect this new argument in the associated compiler macro.
(query-coding-string):
Add a new IGNORE-INVALID-SEQUENCESP argument, document what it
does. Support the HIGHLIGHT argument correctly.
* unicode.el (unicode-query-coding-region):
Add a new IGNORE-INVALID-SEQUENCESP argument, document what it
does, implement this. Document a potential problem.
Use #'query-coding-clear-highlights instead of reimplementing it
ourselves.
Remove some debugging messages.
* mule/arabic.el (iso-8859-6):
* mule/cyrillic.el (iso-8859-5):
* mule/greek.el (iso-8859-7):
* mule/hebrew.el (iso-8859-8):
* mule/latin.el (iso-8859-2):
* mule/latin.el (iso-8859-3):
* mule/latin.el (iso-8859-4):
* mule/latin.el (iso-8859-14):
* mule/latin.el (iso-8859-15):
* mule/latin.el (iso-8859-16):
* mule/latin.el (iso-8859-9):
* mule/latin.el (windows-1252):
* mule/mule-coding.el (iso-8859-1):
Avoid the assumption that characters not given an explicit mapping
in these coding systems map to the ISO 8859-1 characters
corresponding to the octets on disk; this makes it much more
reasonable to implement the IGNORE-INVALID-SEQUENCESP argument to
query-coding-region.
* mule/mule-cmds.el (set-language-info):
Correct the docstring.
* mule/mule-cmds.el (finish-set-language-environment):
Treat invalid Unicode sequences produced from
invalid-sequence-coding-system and corresponding to control
characters the same as control characters in redisplay.
* mule/mule-cmds.el:
Document that encode-coding-char is available in coding.el
* mule/mule-coding.el (make-8-bit-generate-helper):
Change to return the both the encode-program generated and the
relevant non-ASCII charset; update the docstring to reflect this.
* mule/mule-coding.el
(make-8-bit-generate-encode-program-and-skip-chars-strings):
Rename this function; have it return skip-chars-strings as well as
the encode program. Have these skip-chars-strings use ranges for
charsets, where possible.
* mule/mule-coding.el (make-8-bit-create-decode-encode-tables):
Revise this to allow people to specify explicitly characters that
should be undefined (= corresponding to keys in
unicode-error-default-translation-table), and treating unspecified
octets above #x7f as undefined by default.
* mule/mule-coding.el (8-bit-fixed-query-coding-region):
Add a new IGNORE-INVALID-SEQUENCESP argument, implement support
for it using the 8-bit-fixed-invalid-sequences-skip-chars coding
system property; remove some debugging messages.
* mule/mule-coding.el (make-8-bit-coding-system):
This function is dumped, autoloading it makes no sense.
Document what happens when characters above #x7f are not
specified, implement this.
* mule/vietnamese.el:
Correct spelling.
tests/ChangeLog addition:
2009-02-07 Aidan Kehoe <kehoea@parhasard.net>
* automated/query-coding-tests.el:
Add FAILING-CASE arguments to the Assert calls, making #'q-c-debug
mostly unnecessary. Remove #'q-c-debug.
Add new tests that use the IGNORE-INVALID-SEQUENCESP argument to
#'query-coding-region; rework the existing ones to respect it.
author | Aidan Kehoe <kehoea@parhasard.net> |
---|---|
date | Sat, 07 Feb 2009 17:13:37 +0000 |
parents | ee35a8fdcfcd |
children | 313c2cc696b9 |
line wrap: on
line source
/* Numeric types for XEmacs using the GNU MP library. Copyright (C) 2004 Jerry James. This file is part of XEmacs. XEmacs is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2, or (at your option) any later version. XEmacs is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with XEmacs; see the file COPYING. If not, write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ /* Synched up with: Not in FSF. */ #include <config.h> #include <limits.h> #include <math.h> #include "lisp.h" #include "sysproc.h" /* For qxe_getpid */ static mpf_t float_print_min, float_print_max; gmp_randstate_t random_state; CIbyte * bigfloat_to_string(mpf_t f, int base) { mp_exp_t expt; CIbyte *str = mpf_get_str (NULL, &expt, base, 0, f); const int sign = mpf_sgn (f); const int neg = (sign < 0) ? 1 : 0; int len = strlen (str) + 1; /* Count the null terminator */ if (sign == 0 || (mpf_cmp (float_print_min, f) <= 0 && mpf_cmp (f, float_print_max) <= 0)) { /* Move digits down to insert a radix point */ if (expt <= 0) { /* We need room for a radix point and leading zeroes */ const int space = -expt + 2; XREALLOC_ARRAY (str, CIbyte, len + space); memmove (&str[space + neg], &str[neg], len - neg); memset (&str[neg], '0', space); str[neg + 1] = '.'; len += space; } else { /* We just need room for a radix point */ XREALLOC_ARRAY (str, CIbyte, len + 1); memmove (&str[expt + neg + 1], &str[expt + neg], len - (expt + neg)); str[expt + neg] = '.'; len++; } } else { /* Computerized scientific notation */ /* We need room for a radix point, format identifier, and exponent */ const int space = (expt < 0) ? (int)(log ((double) (-expt)) / log ((double) base)) + 3 : (int)(log ((double) expt) / log ((double) base)) + 2; XREALLOC_ARRAY (str, CIbyte, len + space); memmove (&str[neg + 2], &str[neg + 1], len - neg); str[len + 1] = 'l'; sprintf (&str[len + 2], "%ld", expt); } return str; } /* We need the next two functions since GNU MP insists on giving us an extra parameter. */ static void *gmp_realloc (void *ptr, size_t UNUSED (old_size), size_t new_size) { return xrealloc (ptr, new_size); } static void gmp_free (void *ptr, size_t UNUSED (size)) { xfree (ptr, void *); } void init_number_gmp () { mp_set_memory_functions ((void *(*) (size_t)) xmalloc, gmp_realloc, gmp_free); /* The smallest number that is printed without exponents */ mpf_init_set_d (float_print_min, 0.001); /* The largest number that is printed without exponents */ mpf_init_set_ui (float_print_max, 10000000UL); /* Prepare the bignum/bigfloat random number generator */ gmp_randinit_default (random_state); gmp_randseed_ui (random_state, qxe_getpid () + time (NULL)); }