50668cf626
2005-03-14 Geoffrey Keating <geoffk@apple.com> * doc/cppopts.texi (-fexec-charset): Add concept index entry. (-fwide-exec-charset): Likewise. (-finput-charset): Likewise. * doc/invoke.texi (Warning Options): Document -Wnormalized=. * c-opts.c (c_common_handle_option): Handle -Wnormalized=. * c.opt (Wnormalized): New. Index: libcpp/ChangeLog 2005-03-14 Geoffrey Keating <geoffk@apple.com> * init.c (cpp_create_reader): Default warn_normalize to normalized_C. * charset.c: Update for new format of ucnid.h. (ucn_valid_in_identifier): Update for new format of ucnid.h. Add NST parameter, and update it; update callers. (cpp_valid_ucn): Add NST parameter, update callers. Replace abort with cpp_error. (convert_ucn): Pass normalize_state to cpp_valid_ucn. * internal.h (struct normalize_state): New. (INITIAL_NORMALIZE_STATE): New. (NORMALIZE_STATE_RESULT): New. (NORMALIZE_STATE_UPDATE_IDNUM): New. (_cpp_valid_ucn): New. * lex.c (warn_about_normalization): New. (forms_identifier_p): Add normalize_state parameter, update callers. (lex_identifier): Add normalize_state parameter, update callers. Keep the state current. (lex_number): Likewise. (_cpp_lex_direct): Pass normalize_state to subroutines. Check it with warn_about_normalization. * makeucnid.c: New. * ucnid.h: Replace. * ucnid.pl: Remove. * ucnid.tab: Make appropriate for input to makeucnid.c. Remove comments about obsolete version of C++. * include/cpplib.h (enum cpp_normalize_level): New. (struct cpp_options): Add warn_normalize field. Index: gcc/testsuite/ChangeLog 2005-03-14 Geoffrey Keating <geoffk@apple.com> * gcc.dg/cpp/normalize-1.c: New. * gcc.dg/cpp/normalize-2.c: New. * gcc.dg/cpp/normalize-3.c: New. * gcc.dg/cpp/normalize-4.c: New. * gcc.dg/cpp/ucnid-4.c: New. * gcc.dg/cpp/ucnid-5.c: New. * g++.dg/cpp/normalize-1.C: New. * g++.dg/cpp/ucnid-1.C: New. From-SVN: r96459
343 lines
10 KiB
C
343 lines
10 KiB
C
/* Make ucnid.h from various sources.
|
|
Copyright (C) 2005 Free Software Foundation, Inc.
|
|
|
|
This program 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.
|
|
|
|
This program 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 this program; if not, write to the Free Software
|
|
Foundation, 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */
|
|
|
|
/* Run this program as
|
|
./makeucnid ucnid.tab UnicodeData.txt DerivedNormalizationProps.txt \
|
|
> ucnid.h
|
|
*/
|
|
|
|
#include <stdio.h>
|
|
#include <string.h>
|
|
#include <ctype.h>
|
|
#include <stdbool.h>
|
|
#include <stdlib.h>
|
|
|
|
enum {
|
|
C99 = 1,
|
|
CXX = 2,
|
|
digit = 4,
|
|
not_NFC = 8,
|
|
not_NFKC = 16,
|
|
maybe_not_NFC = 32
|
|
};
|
|
|
|
static unsigned flags[65536];
|
|
static unsigned short decomp[65536][2];
|
|
static unsigned char combining_value[65536];
|
|
|
|
/* Die! */
|
|
|
|
static void
|
|
fail (const char *s)
|
|
{
|
|
fprintf (stderr, "%s\n", s);
|
|
exit (1);
|
|
}
|
|
|
|
/* Read ucnid.tab and set the C99 and CXX flags in header[]. */
|
|
|
|
static void
|
|
read_ucnid (const char *fname)
|
|
{
|
|
FILE *f = fopen (fname, "r");
|
|
unsigned fl = 0;
|
|
|
|
if (!f)
|
|
fail ("opening ucnid.tab");
|
|
for (;;)
|
|
{
|
|
char line[256];
|
|
|
|
if (!fgets (line, sizeof (line), f))
|
|
break;
|
|
if (strcmp (line, "[C99]\n") == 0)
|
|
fl = C99;
|
|
else if (strcmp (line, "[CXX]\n") == 0)
|
|
fl = CXX;
|
|
else if (isxdigit (line[0]))
|
|
{
|
|
char *l = line;
|
|
while (*l)
|
|
{
|
|
unsigned long start, end;
|
|
char *endptr;
|
|
start = strtoul (l, &endptr, 16);
|
|
if (endptr == l || (*endptr != '-' && ! isspace (*endptr)))
|
|
fail ("parsing ucnid.tab [1]");
|
|
l = endptr;
|
|
if (*l != '-')
|
|
end = start;
|
|
else
|
|
{
|
|
end = strtoul (l + 1, &endptr, 16);
|
|
if (end < start)
|
|
fail ("parsing ucnid.tab, end before start");
|
|
l = endptr;
|
|
if (! isspace (*l))
|
|
fail ("parsing ucnid.tab, junk after range");
|
|
}
|
|
while (isspace (*l))
|
|
l++;
|
|
if (end > 0xFFFF)
|
|
fail ("parsing ucnid.tab, end too large");
|
|
while (start <= end)
|
|
flags[start++] |= fl;
|
|
}
|
|
}
|
|
}
|
|
if (ferror (f))
|
|
fail ("reading ucnid.tab");
|
|
fclose (f);
|
|
}
|
|
|
|
/* Read UnicodeData.txt and set the 'digit' flag, and
|
|
also fill in the 'decomp' table to be the decompositions of
|
|
characters for which both the character decomposed and all the code
|
|
points in the decomposition are either C99 or CXX. */
|
|
|
|
static void
|
|
read_table (char *fname)
|
|
{
|
|
FILE * f = fopen (fname, "r");
|
|
|
|
if (!f)
|
|
fail ("opening UnicodeData.txt");
|
|
for (;;)
|
|
{
|
|
char line[256];
|
|
unsigned long codepoint, this_decomp[4];
|
|
char *l;
|
|
int i;
|
|
int decomp_useful;
|
|
|
|
if (!fgets (line, sizeof (line), f))
|
|
break;
|
|
codepoint = strtoul (line, &l, 16);
|
|
if (l == line || *l != ';')
|
|
fail ("parsing UnicodeData.txt, reading code point");
|
|
if (codepoint > 0xffff || ! (flags[codepoint] & (C99 | CXX)))
|
|
continue;
|
|
|
|
do {
|
|
l++;
|
|
} while (*l != ';');
|
|
/* Category value; things starting with 'N' are numbers of some
|
|
kind. */
|
|
if (*++l == 'N')
|
|
flags[codepoint] |= digit;
|
|
|
|
do {
|
|
l++;
|
|
} while (*l != ';');
|
|
/* Canonical combining class; in NFC/NFKC, they must be increasing
|
|
(or zero). */
|
|
if (! isdigit (*++l))
|
|
fail ("parsing UnicodeData.txt, combining class not number");
|
|
combining_value[codepoint] = strtoul (l, &l, 10);
|
|
if (*l++ != ';')
|
|
fail ("parsing UnicodeData.txt, junk after combining class");
|
|
|
|
/* Skip over bidi value. */
|
|
do {
|
|
l++;
|
|
} while (*l != ';');
|
|
|
|
/* Decomposition mapping. */
|
|
decomp_useful = flags[codepoint];
|
|
if (*++l == '<') /* Compatibility mapping. */
|
|
continue;
|
|
for (i = 0; i < 4; i++)
|
|
{
|
|
if (*l == ';')
|
|
break;
|
|
if (!isxdigit (*l))
|
|
fail ("parsing UnicodeData.txt, decomposition format");
|
|
this_decomp[i] = strtoul (l, &l, 16);
|
|
decomp_useful &= flags[this_decomp[i]];
|
|
while (isspace (*l))
|
|
l++;
|
|
}
|
|
if (i > 2) /* Decomposition too long. */
|
|
fail ("parsing UnicodeData.txt, decomposition too long");
|
|
if (decomp_useful)
|
|
while (--i >= 0)
|
|
decomp[codepoint][i] = this_decomp[i];
|
|
}
|
|
if (ferror (f))
|
|
fail ("reading UnicodeData.txt");
|
|
fclose (f);
|
|
}
|
|
|
|
/* Read DerivedNormalizationProps.txt and set the flags that say whether
|
|
a character is in NFC, NFKC, or is context-dependent. */
|
|
|
|
static void
|
|
read_derived (const char *fname)
|
|
{
|
|
FILE * f = fopen (fname, "r");
|
|
|
|
if (!f)
|
|
fail ("opening DerivedNormalizationProps.txt");
|
|
for (;;)
|
|
{
|
|
char line[256];
|
|
unsigned long start, end;
|
|
char *l;
|
|
bool not_NFC_p, not_NFKC_p, maybe_not_NFC_p;
|
|
|
|
if (!fgets (line, sizeof (line), f))
|
|
break;
|
|
not_NFC_p = (strstr (line, "; NFC_QC; N") != NULL);
|
|
not_NFKC_p = (strstr (line, "; NFKC_QC; N") != NULL);
|
|
maybe_not_NFC_p = (strstr (line, "; NFC_QC; M") != NULL);
|
|
if (! not_NFC_p && ! not_NFKC_p && ! maybe_not_NFC_p)
|
|
continue;
|
|
|
|
start = strtoul (line, &l, 16);
|
|
if (l == line)
|
|
fail ("parsing DerivedNormalizationProps.txt, reading start");
|
|
if (start > 0xffff)
|
|
continue;
|
|
if (*l == '.' && l[1] == '.')
|
|
end = strtoul (l + 2, &l, 16);
|
|
else
|
|
end = start;
|
|
|
|
while (start <= end)
|
|
flags[start++] |= ((not_NFC_p ? not_NFC : 0)
|
|
| (not_NFKC_p ? not_NFKC : 0)
|
|
| (maybe_not_NFC_p ? maybe_not_NFC : 0)
|
|
);
|
|
}
|
|
if (ferror (f))
|
|
fail ("reading DerivedNormalizationProps.txt");
|
|
fclose (f);
|
|
}
|
|
|
|
/* Write out the table.
|
|
The table consists of two words per entry. The first word is the flags
|
|
for the unicode code points up to and including the second word. */
|
|
|
|
static void
|
|
write_table (void)
|
|
{
|
|
unsigned i;
|
|
unsigned last_flag = flags[0];
|
|
bool really_safe = decomp[0][0] == 0;
|
|
unsigned char last_combine = combining_value[0];
|
|
|
|
for (i = 1; i <= 65536; i++)
|
|
if (i == 65536
|
|
|| (flags[i] != last_flag && ((flags[i] | last_flag) & (C99 | CXX)))
|
|
|| really_safe != (decomp[i][0] == 0)
|
|
|| combining_value[i] != last_combine)
|
|
{
|
|
printf ("{ %s|%s|%s|%s|%s|%s|%s, %3d, %#06x },\n",
|
|
last_flag & C99 ? "C99" : " 0",
|
|
last_flag & digit ? "DIG" : " 0",
|
|
last_flag & CXX ? "CXX" : " 0",
|
|
really_safe ? "CID" : " 0",
|
|
last_flag & not_NFC ? " 0" : "NFC",
|
|
last_flag & not_NFKC ? " 0" : "NKC",
|
|
last_flag & maybe_not_NFC ? "CTX" : " 0",
|
|
combining_value[i - 1],
|
|
i - 1);
|
|
last_flag = flags[i];
|
|
last_combine = combining_value[0];
|
|
really_safe = decomp[i][0] == 0;
|
|
}
|
|
}
|
|
|
|
/* Print out the huge copyright notice. */
|
|
|
|
static void
|
|
write_copyright (void)
|
|
{
|
|
static const char copyright[] = "\
|
|
/* Unicode characters and various properties.\n\
|
|
Copyright (C) 2003, 2005 Free Software Foundation, Inc.\n\
|
|
\n\
|
|
This program is free software; you can redistribute it and/or modify it\n\
|
|
under the terms of the GNU General Public License as published by the\n\
|
|
Free Software Foundation; either version 2, or (at your option) any\n\
|
|
later version.\n\
|
|
\n\
|
|
This program is distributed in the hope that it will be useful,\n\
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of\n\
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the\n\
|
|
GNU General Public License for more details.\n\
|
|
\n\
|
|
You should have received a copy of the GNU General Public License\n\
|
|
along with this program; if not, write to the Free Software\n\
|
|
Foundation, 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.\n\
|
|
\n\
|
|
\n\
|
|
Copyright (C) 1991-2005 Unicode, Inc. All rights reserved.\n\
|
|
Distributed under the Terms of Use in\n\
|
|
http://www.unicode.org/copyright.html.\n\
|
|
\n\
|
|
Permission is hereby granted, free of charge, to any person\n\
|
|
obtaining a copy of the Unicode data files and any associated\n\
|
|
documentation (the \"Data Files\") or Unicode software and any\n\
|
|
associated documentation (the \"Software\") to deal in the Data Files\n\
|
|
or Software without restriction, including without limitation the\n\
|
|
rights to use, copy, modify, merge, publish, distribute, and/or\n\
|
|
sell copies of the Data Files or Software, and to permit persons to\n\
|
|
whom the Data Files or Software are furnished to do so, provided\n\
|
|
that (a) the above copyright notice(s) and this permission notice\n\
|
|
appear with all copies of the Data Files or Software, (b) both the\n\
|
|
above copyright notice(s) and this permission notice appear in\n\
|
|
associated documentation, and (c) there is clear notice in each\n\
|
|
modified Data File or in the Software as well as in the\n\
|
|
documentation associated with the Data File(s) or Software that the\n\
|
|
data or software has been modified.\n\
|
|
\n\
|
|
THE DATA FILES AND SOFTWARE ARE PROVIDED \"AS IS\", WITHOUT WARRANTY\n\
|
|
OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE\n\
|
|
WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND\n\
|
|
NONINFRINGEMENT OF THIRD PARTY RIGHTS. IN NO EVENT SHALL THE\n\
|
|
COPYRIGHT HOLDER OR HOLDERS INCLUDED IN THIS NOTICE BE LIABLE FOR\n\
|
|
ANY CLAIM, OR ANY SPECIAL INDIRECT OR CONSEQUENTIAL DAMAGES, OR ANY\n\
|
|
DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS,\n\
|
|
WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS\n\
|
|
ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE\n\
|
|
OF THE DATA FILES OR SOFTWARE.\n\
|
|
\n\
|
|
Except as contained in this notice, the name of a copyright holder\n\
|
|
shall not be used in advertising or otherwise to promote the sale,\n\
|
|
use or other dealings in these Data Files or Software without prior\n\
|
|
written authorization of the copyright holder. */\n";
|
|
|
|
puts (copyright);
|
|
}
|
|
|
|
/* Main program. */
|
|
|
|
int
|
|
main(int argc, char ** argv)
|
|
{
|
|
if (argc != 4)
|
|
fail ("too few arguments to makeucn");
|
|
read_ucnid (argv[1]);
|
|
read_table (argv[2]);
|
|
read_derived (argv[3]);
|
|
|
|
write_copyright ();
|
|
write_table ();
|
|
return 0;
|
|
}
|