upstream-mirrors/gnulib
1
0
Fork 0
mirror of https://https.git.savannah.gnu.org/git/gnulib.git synced 2026-04-28 06:33:36 +00:00
Code Activity

Let linebreak rely on uniwidth/width.

Browse Source
This commit is contained in:
Bruno Haible
2007-09-01 15:52:06 +00:00
parent 9da9f6021c
commit f582cd8fd6
5 changed files with 25 additions and 511 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
ChangeLog
View File

@@ -1,3 +1,18 @@
2007-09-01 Bruno Haible <bruno@clisp.org>
Fix collision between uniwidth/* and linebreak modules.
* lib/linebreak.h (locale_charset, uc_width, u8_width, u16_width,
u32_width): Remove declarations.
* lib/linebreak.c: Include uniwidth.h, uniwidth/cjk.h, streq.h.
(u32_mbtouc_unsafe, streq9, streq8, streq7, streq6, streq5, streq4,
streq3, streq2, streq1, streq0): Remove functions.
(STREQ): Remove macro.
(is_cjk_encoding): Remove function.
(nonspacing_table_data, nonspacing_table_ind): Remove constants.
(uc_width, u8_width, u16_width, u32_width): Remove functions.
* modules/linebreak (Depends-on): Add streq, uniwidth/width.
* NEWS: Document the change.
2007-08-28 Jim Meyering <jim@meyering.net>
Rename mreadlink_with_size to areadlink_with_size.

4
NEWS
View File

@@ -6,6 +6,10 @@ User visible incompatible changes
Date Modules Changes
2007-09-01 linebreak "linebreak.h" no longer declares the functions
locale_charset, uc_width, u{8,16,32}_width. Use
"uniwidth.h" to get these functions declared.
2007-08-28 areadlink-with-size
Renamed from mreadlink-with-size.
Function renamed: mreadlink_with_size ->

495
lib/linebreak.c
View File

@@ -26,182 +26,10 @@ Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
#include "c-ctype.h"
#include "xsize.h"
#include "unistr.h"
#include "uniwidth.h"
#include "uniwidth/cjk.h"
#include "streq.h"
#ifdef unused
static inline int
u32_mbtouc_unsafe (unsigned int *puc, const unsigned int *s, size_t n)
{
*puc = *s;
return 1;
}
#endif
/* Help GCC to generate good code for string comparisons with
immediate strings. */
#if defined (__GNUC__) && defined (__OPTIMIZE__)
static inline int
streq9 (const char *s1, const char *s2)
{
return strcmp (s1 + 9, s2 + 9) == 0;
}
static inline int
streq8 (const char *s1, const char *s2, char s28)
{
if (s1[8] == s28)
{
if (s28 == 0)
return 1;
else
return streq9 (s1, s2);
}
else
return 0;
}
static inline int
streq7 (const char *s1, const char *s2, char s27, char s28)
{
if (s1[7] == s27)
{
if (s27 == 0)
return 1;
else
return streq8 (s1, s2, s28);
}
else
return 0;
}
static inline int
streq6 (const char *s1, const char *s2, char s26, char s27, char s28)
{
if (s1[6] == s26)
{
if (s26 == 0)
return 1;
else
return streq7 (s1, s2, s27, s28);
}
else
return 0;
}
static inline int
streq5 (const char *s1, const char *s2, char s25, char s26, char s27, char s28)
{
if (s1[5] == s25)
{
if (s25 == 0)
return 1;
else
return streq6 (s1, s2, s26, s27, s28);
}
else
return 0;
}
static inline int
streq4 (const char *s1, const char *s2, char s24, char s25, char s26, char s27, char s28)
{
if (s1[4] == s24)
{
if (s24 == 0)
return 1;
else
return streq5 (s1, s2, s25, s26, s27, s28);
}
else
return 0;
}
static inline int
streq3 (const char *s1, const char *s2, char s23, char s24, char s25, char s26, char s27, char s28)
{
if (s1[3] == s23)
{
if (s23 == 0)
return 1;
else
return streq4 (s1, s2, s24, s25, s26, s27, s28);
}
else
return 0;
}
static inline int
streq2 (const char *s1, const char *s2, char s22, char s23, char s24, char s25, char s26, char s27, char s28)
{
if (s1[2] == s22)
{
if (s22 == 0)
return 1;
else
return streq3 (s1, s2, s23, s24, s25, s26, s27, s28);
}
else
return 0;
}
static inline int
streq1 (const char *s1, const char *s2, char s21, char s22, char s23, char s24, char s25, char s26, char s27, char s28)
{
if (s1[1] == s21)
{
if (s21 == 0)
return 1;
else
return streq2 (s1, s2, s22, s23, s24, s25, s26, s27, s28);
}
else
return 0;
}
static inline int
streq0 (const char *s1, const char *s2, char s20, char s21, char s22, char s23, char s24, char s25, char s26, char s27, char s28)
{
if (s1[0] == s20)
{
if (s20 == 0)
return 1;
else
return streq1 (s1, s2, s21, s22, s23, s24, s25, s26, s27, s28);
}
else
return 0;
}
#define STREQ(s1,s2,s20,s21,s22,s23,s24,s25,s26,s27,s28) \
streq0 (s1, s2, s20, s21, s22, s23, s24, s25, s26, s27, s28)
#else
#define STREQ(s1,s2,s20,s21,s22,s23,s24,s25,s26,s27,s28) \
(strcmp (s1, s2) == 0)
#endif
static int
is_cjk_encoding (const char *encoding)
{
if (0
/* Legacy Japanese encodings */
|| STREQ (encoding, "EUC-JP", 'E', 'U', 'C', '-', 'J', 'P', 0, 0, 0)
/* Legacy Chinese encodings */
|| STREQ (encoding, "GB2312", 'G', 'B', '2', '3', '1', '2', 0, 0, 0)
|| STREQ (encoding, "GBK", 'G', 'B', 'K', 0, 0, 0, 0, 0, 0)
|| STREQ (encoding, "EUC-TW", 'E', 'U', 'C', '-', 'T', 'W', 0, 0, 0)
|| STREQ (encoding, "BIG5", 'B', 'I', 'G', '5', 0, 0, 0, 0, 0)
/* Legacy Korean encodings */
|| STREQ (encoding, "EUC-KR", 'E', 'U', 'C', '-', 'K', 'R', 0, 0, 0)
|| STREQ (encoding, "CP949", 'C', 'P', '9', '4', '9', 0, 0, 0, 0)
|| STREQ (encoding, "JOHAB", 'J', 'O', 'H', 'A', 'B', 0, 0, 0, 0))
return 1;
return 0;
}
static int
is_utf8_encoding (const char *encoding)
@@ -212,323 +40,6 @@ is_utf8_encoding (const char *encoding)
}
/* Determine number of column positions required for UC. */
int uc_width (unsigned int uc, const char *encoding);
/*
* Non-spacing attribute table.
* Consists of:
* - Non-spacing characters; generated from PropList.txt or
* "grep '^[^;]*;[^;]*;[^;]*;[^;]*;NSM;' UnicodeData.txt"
* - Format control characters; generated from
* "grep '^[^;]*;[^;]*;Cf;' UnicodeData.txt"
* - Zero width characters; generated from
* "grep '^[^;]*;ZERO WIDTH ' UnicodeData.txt"
*/
static const unsigned char nonspacing_table_data[16*64] = {
/* 0x0000-0x01ff */
0xff, 0xff, 0xff, 0xff, 0x00, 0x00, 0x00, 0x00, /* 0x0000-0x003f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x80, /* 0x0040-0x007f */
0xff, 0xff, 0xff, 0xff, 0x00, 0x20, 0x00, 0x00, /* 0x0080-0x00bf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x00c0-0x00ff */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0100-0x013f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0140-0x017f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0180-0x01bf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x01c0-0x01ff */
/* 0x0200-0x03ff */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0200-0x023f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0240-0x027f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0280-0x02bf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x02c0-0x02ff */
0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, /* 0x0300-0x033f */
0xff, 0xff, 0xff, 0xe0, 0xff, 0xff, 0x00, 0x00, /* 0x0340-0x037f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0380-0x03bf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x03c0-0x03ff */
/* 0x0400-0x05ff */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0400-0x043f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0440-0x047f */
0x78, 0x03, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0480-0x04bf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x04c0-0x04ff */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0500-0x053f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0540-0x057f */
0x00, 0x00, 0xfe, 0xff, 0xfb, 0xff, 0xff, 0xbb, /* 0x0580-0x05bf */
0x16, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x05c0-0x05ff */
/* 0x0600-0x07ff */
0x0f, 0x00, 0x3f, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0600-0x063f */
0x00, 0xf8, 0xff, 0x01, 0x00, 0x00, 0x01, 0x00, /* 0x0640-0x067f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0680-0x06bf */
0x00, 0x00, 0xc0, 0xff, 0x9f, 0x3d, 0x00, 0x00, /* 0x06c0-0x06ff */
0x00, 0x80, 0x02, 0x00, 0x00, 0x00, 0xff, 0xff, /* 0x0700-0x073f */
0xff, 0x07, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0740-0x077f */
0x00, 0x00, 0x00, 0x00, 0xc0, 0xff, 0x01, 0x00, /* 0x0780-0x07bf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x07c0-0x07ff */
/* 0x0800-0x09ff */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0800-0x083f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0840-0x087f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0880-0x08bf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x08c0-0x08ff */
0x06, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x10, /* 0x0900-0x093f */
0xfe, 0x21, 0x1e, 0x00, 0x0c, 0x00, 0x00, 0x00, /* 0x0940-0x097f */
0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x10, /* 0x0980-0x09bf */
0x1e, 0x20, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00, /* 0x09c0-0x09ff */
/* 0x0a00-0x0bff */
0x06, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x10, /* 0x0a00-0x0a3f */
0x86, 0x39, 0x00, 0x00, 0x00, 0x00, 0x03, 0x00, /* 0x0a40-0x0a7f */
0x06, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x10, /* 0x0a80-0x0abf */
0xbe, 0x21, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00, /* 0x0ac0-0x0aff */
0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x90, /* 0x0b00-0x0b3f */
0x0e, 0x20, 0x40, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0b40-0x0b7f */
0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0b80-0x0bbf */
0x01, 0x20, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0bc0-0x0bff */
/* 0x0c00-0x0dff */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xc0, /* 0x0c00-0x0c3f */
0xc1, 0x3d, 0x60, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0c40-0x0c7f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x10, /* 0x0c80-0x0cbf */
0x00, 0x30, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0cc0-0x0cff */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0d00-0x0d3f */
0x0e, 0x20, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0d40-0x0d7f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0d80-0x0dbf */
0x00, 0x04, 0x5c, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0dc0-0x0dff */
/* 0x0e00-0x0fff */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xf2, 0x07, /* 0x0e00-0x0e3f */
0x80, 0x7f, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0e40-0x0e7f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xf2, 0x1b, /* 0x0e80-0x0ebf */
0x00, 0x3f, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0ec0-0x0eff */
0x00, 0x00, 0x00, 0x03, 0x00, 0x00, 0xa0, 0x02, /* 0x0f00-0x0f3f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xfe, 0x7f, /* 0x0f40-0x0f7f */
0xdf, 0x00, 0xff, 0xfe, 0xff, 0xff, 0xff, 0x1f, /* 0x0f80-0x0fbf */
0x40, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0fc0-0x0fff */
/* 0x1000-0x11ff */
0x00, 0x00, 0x00, 0x00, 0x00, 0xe0, 0xc5, 0x02, /* 0x1000-0x103f */
0x00, 0x00, 0x00, 0x03, 0x00, 0x00, 0x00, 0x00, /* 0x1040-0x107f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x1080-0x10bf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x10c0-0x10ff */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x1100-0x113f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x1140-0x117f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x1180-0x11bf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x11c0-0x11ff */
/* 0x1600-0x17ff */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x1600-0x163f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x1640-0x167f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x1680-0x16bf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x16c0-0x16ff */
0x00, 0x00, 0x1c, 0x00, 0x00, 0x00, 0x1c, 0x00, /* 0x1700-0x173f */
0x00, 0x00, 0x0c, 0x00, 0x00, 0x00, 0x0c, 0x00, /* 0x1740-0x177f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xb0, 0x3f, /* 0x1780-0x17bf */
0x40, 0xfe, 0x0f, 0x20, 0x00, 0x00, 0x00, 0x00, /* 0x17c0-0x17ff */
/* 0x1800-0x19ff */
0x00, 0x38, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x1800-0x183f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x1840-0x187f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, /* 0x1880-0x18bf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x18c0-0x18ff */
0x00, 0x00, 0x00, 0x00, 0x87, 0x0f, 0x04, 0x0e, /* 0x1900-0x193f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x1940-0x197f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x1980-0x19bf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x19c0-0x19ff */
/* 0x2000-0x21ff */
0x00, 0xf8, 0x00, 0x00, 0x00, 0x7c, 0x00, 0x00, /* 0x2000-0x203f */
0x00, 0x00, 0x00, 0x00, 0x0f, 0xfc, 0x00, 0x00, /* 0x2040-0x207f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x2080-0x20bf */
0x00, 0x00, 0xff, 0xff, 0xff, 0x07, 0x00, 0x00, /* 0x20c0-0x20ff */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x2100-0x213f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x2140-0x217f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x2180-0x21bf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x21c0-0x21ff */
/* 0x3000-0x31ff */
0x00, 0x00, 0x00, 0x00, 0x00, 0xfc, 0x00, 0x00, /* 0x3000-0x303f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x3040-0x307f */
0x00, 0x00, 0x00, 0x06, 0x00, 0x00, 0x00, 0x00, /* 0x3080-0x30bf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x30c0-0x30ff */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x3100-0x313f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x3140-0x317f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x3180-0x31bf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x31c0-0x31ff */
/* 0xfa00-0xfbff */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xfa00-0xfa3f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xfa40-0xfa7f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xfa80-0xfabf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xfac0-0xfaff */
0x00, 0x00, 0x00, 0x40, 0x00, 0x00, 0x00, 0x00, /* 0xfb00-0xfb3f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xfb40-0xfb7f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xfb80-0xfbbf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xfbc0-0xfbff */
/* 0xfe00-0xffff */
0xff, 0xff, 0x00, 0x00, 0x0f, 0x00, 0x00, 0x00, /* 0xfe00-0xfe3f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xfe40-0xfe7f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xfe80-0xfebf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x80, /* 0xfec0-0xfeff */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xff00-0xff3f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xff40-0xff7f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xff80-0xffbf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0e, /* 0xffc0-0xffff */
/* 0x1d000-0x1d1ff */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x1d000-0x1d03f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x1d040-0x1d07f */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x1d080-0x1d0bf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x1d0c0-0x1d0ff */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x1d100-0x1d13f */
0x00, 0x00, 0x00, 0x00, 0x80, 0x03, 0x00, 0xf8, /* 0x1d140-0x1d17f */
0xe7, 0x0f, 0x00, 0x00, 0x00, 0x3c, 0x00, 0x00, /* 0x1d180-0x1d1bf */
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 /* 0x1d1c0-0x1d1ff */
};
static const signed char nonspacing_table_ind[240] = {
0, 1, 2, 3, 4, 5, 6, 7, /* 0x0000-0x0fff */
8, -1, -1, 9, 10, -1, -1, -1, /* 0x1000-0x1fff */
11, -1, -1, -1, -1, -1, -1, -1, /* 0x2000-0x2fff */
12, -1, -1, -1, -1, -1, -1, -1, /* 0x3000-0x3fff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x4000-0x4fff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x5000-0x5fff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x6000-0x6fff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x7000-0x7fff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x8000-0x8fff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x9000-0x9fff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0xa000-0xafff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0xb000-0xbfff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0xc000-0xcfff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0xd000-0xdfff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0xe000-0xefff */
-1, -1, -1, -1, -1, 13, -1, 14, /* 0xf000-0xffff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x10000-0x10fff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x11000-0x11fff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x12000-0x12fff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x13000-0x13fff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x14000-0x14fff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x15000-0x15fff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x16000-0x16fff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x17000-0x17fff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x18000-0x18fff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x19000-0x19fff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x1a000-0x1afff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x1b000-0x1bfff */
-1, -1, -1, -1, -1, -1, -1, -1, /* 0x1c000-0x1cfff */
15, -1, -1, -1, -1, -1, -1, -1 /* 0x1d000-0x1dfff */
};
/* Determine number of column positions required for UC. */
int
uc_width (unsigned int uc, const char *encoding)
{
/* Test for non-spacing or control character. */
if ((uc >> 9) < 240)
{
int ind = nonspacing_table_ind[uc >> 9];
if (ind >= 0)
if ((nonspacing_table_data[64*ind + ((uc >> 3) & 63)] >> (uc & 7)) & 1)
{
if (uc > 0 && uc < 0xa0)
return -1;
else
return 0;
}
}
else if ((uc >> 9) == (0xe0000 >> 9))
{
if (uc < 0xe0100
? (uc >= 0xe0020 ? uc <= 0xe007f : uc == 0xe0001)
: (uc <= 0xe01ef))
return 0;
}
/* Test for double-width character.
* Generated from "grep '^....;[WF]' EastAsianWidth.txt"
* and "grep '^....;[^WF]' EastAsianWidth.txt"
*/
if (uc >= 0x1100
&& ((uc < 0x1160) /* Hangul Jamo */
|| (uc >= 0x2e80 && uc < 0x4dc0 /* CJK */
&& !(uc == 0x303f))
|| (uc >= 0x4e00 && uc < 0xa4d0) /* CJK ... Yi */
|| (uc >= 0xac00 && uc < 0xd7a4) /* Hangul Syllables */
|| (uc >= 0xf900 && uc < 0xfb00) /* CJK Compatibility Ideographs */
|| (uc >= 0xfe30 && uc < 0xfe70) /* CJK Compatibility Forms */
|| (uc >= 0xff00 && uc < 0xff61) /* Fullwidth Forms */
|| (uc >= 0xffe0 && uc < 0xffe7)
|| (uc >= 0x20000 && uc <= 0x2fffd) /* CJK, CJK Compatibility Ideographs */
|| (uc >= 0x30000 && uc <= 0x3fffd)
) )
return 2;
/* In ancient CJK encodings, Cyrillic and most other characters are
double-width as well. */
if (uc >= 0x00A1 && uc < 0xFF61 && uc != 0x20A9
&& is_cjk_encoding (encoding))
return 2;
return 1;
}
/* Determine number of column positions required for first N units
(or fewer if S ends before this) in S. */
int
u8_width (const unsigned char *s, size_t n, const char *encoding)
{
const unsigned char *s_end = s + n;
int width = 0;
while (s < s_end)
{
unsigned int uc;
int w;
s += u8_mbtouc_unsafe (&uc, s, s_end - s);
if (uc == 0)
break; /* end of string reached */
w = uc_width (uc, encoding);
if (w >= 0) /* ignore control characters in the string */
width += w;
}
return width;
}
int
u16_width (const unsigned short *s, size_t n, const char *encoding)
{
const unsigned short *s_end = s + n;
int width = 0;
while (s < s_end)
{
unsigned int uc;
int w;
s += u16_mbtouc_unsafe (&uc, s, s_end - s);
if (uc == 0)
break; /* end of string reached */
w = uc_width (uc, encoding);
if (w >= 0) /* ignore control characters in the string */
width += w;
}
return width;
}
int
u32_width (const unsigned int *s, size_t n, const char *encoding)
{
const unsigned int *s_end = s + n;
int width = 0;
while (s < s_end)
{
unsigned int uc = *s++;
int w;
if (uc == 0)
break; /* end of string reached */
w = uc_width (uc, encoding);
if (w >= 0) /* ignore control characters in the string */
width += w;
}
return width;
}
/* Determine the line break points in S, and store the result at p[0..n-1]. */
/* We don't support line breaking of complex-context dependent characters
(Thai, Lao, Myanmar, Khmer) yet, because it requires dictionary lookup. */

20
lib/linebreak.h
View File

@@ -1,5 +1,5 @@
/* linebreak.h - line breaking of Unicode strings
Copyright (C) 2001-2003, 2006 Free Software Foundation, Inc.
Copyright (C) 2001-2003, 2006-2007 Free Software Foundation, Inc.
Written by Bruno Haible <haible@clisp.cons.org>, 2001.
This program is free software; you can redistribute it and/or modify
@@ -28,24 +28,6 @@ extern "C" {
#endif
/* Display width. */
/* These functions are locale dependent. The encoding argument identifies
the encoding (e.g. "ISO-8859-2" for Polish). */
/* Return the encoding of the current locale. */
extern const char * locale_charset (void);
/* Determine number of column positions required for UC. */
extern int uc_width (unsigned int uc, const char *encoding);
/* Determine number of column positions required for first N units
(or fewer if S ends before this) in S. */
extern int u8_width (const unsigned char *s, size_t n, const char *encoding);
extern int u16_width (const unsigned short *s, size_t n, const char *encoding);
extern int u32_width (const unsigned int *s, size_t n, const char *encoding);
/* Line breaking. */
enum {

2
modules/linebreak
View File

@@ -14,6 +14,8 @@ iconv
iconv_open
utf8-ucs4-unsafe
utf16-ucs4-unsafe
streq
uniwidth/width
xsize
configure.ac: