Patch 7.4.088
Problem: When spell checking is enabled Asian characters are always marked
as error.
Solution: When 'spelllang' contains "cjk" do not mark Asian characters as
error. (Ken Takata)
Files: runtime/doc/options.txt, runtime/doc/spell.txt, src/mbyte.c,
src/option.c, src/spell.c, src/structs.h
*** ../vim-7.4.087/runtime/doc/options.txt 2013-11-06 05:26:08.000000000
+0100
--- runtime/doc/options.txt 2013-11-12 04:00:51.000000000 +0100
***************
*** 6555,6560 ****
--- 6555,6563 ----
region by listing them: "en_us,en_ca" supports both US and Canadian
English, but not words specific for Australia, New Zealand or Great
Britain.
+ If the name "cjk" is included East Asian characters are excluded from
+ spell checking. This is useful when editing text that also has Asian
+ words.
*E757*
As a special case the name of a .spl file can be given as-is. The
first "_xx" in the name is removed and used as the region name
*** ../vim-7.4.087/runtime/doc/spell.txt 2013-08-10 13:25:01.000000000
+0200
--- runtime/doc/spell.txt 2013-11-12 04:02:27.000000000 +0100
***************
*** 269,274 ****
--- 269,281 ----
latin1 yi transliterated Yiddish
utf-8 yi-tr transliterated Yiddish
+ *spell-cjk*
+ Chinese, Japanese and other East Asian characters are normally marked as
+ errors, because spell checking of these characters is not supported. If
+ 'spelllang' includes "cjk", these characters are not marked as errors. This
+ is useful when editing text with spell checking while some Asian words are
+ present.
+
SPELL FILES *spell-load*
*** ../vim-7.4.087/src/mbyte.c 2013-07-05 20:07:21.000000000 +0200
--- src/mbyte.c 2013-11-12 03:55:50.000000000 +0100
***************
*** 947,954 ****
{
case 0x2121: /* ZENKAKU space */
return 0;
! case 0x2122: /* KU-TEN (Japanese comma) */
! case 0x2123: /* TOU-TEN (Japanese period) */
case 0x2124: /* ZENKAKU comma */
case 0x2125: /* ZENKAKU period */
return 1;
--- 947,954 ----
{
case 0x2121: /* ZENKAKU space */
return 0;
! case 0x2122: /* TOU-TEN (Japanese comma) */
! case 0x2123: /* KU-TEN (Japanese period) */
case 0x2124: /* ZENKAKU comma */
case 0x2125: /* ZENKAKU period */
return 1;
***************
*** 2477,2485 ****
/* sorted list of non-overlapping intervals */
static struct clinterval
{
! unsigned short first;
! unsigned short last;
! unsigned short class;
} classes[] =
{
{0x037e, 0x037e, 1}, /* Greek question mark */
--- 2477,2485 ----
/* sorted list of non-overlapping intervals */
static struct clinterval
{
! unsigned int first;
! unsigned int last;
! unsigned int class;
} classes[] =
{
{0x037e, 0x037e, 1}, /* Greek question mark */
***************
*** 2544,2549 ****
--- 2544,2553 ----
{0xff1a, 0xff20, 1}, /* half/fullwidth ASCII */
{0xff3b, 0xff40, 1}, /* half/fullwidth ASCII */
{0xff5b, 0xff65, 1}, /* half/fullwidth ASCII */
+ {0x20000, 0x2a6df, 0x4e00}, /* CJK Ideographs */
+ {0x2a700, 0x2b73f, 0x4e00}, /* CJK Ideographs */
+ {0x2b740, 0x2b81f, 0x4e00}, /* CJK Ideographs */
+ {0x2f800, 0x2fa1f, 0x4e00}, /* CJK Ideographs */
};
int bot = 0;
int top = sizeof(classes) / sizeof(struct clinterval) - 1;
***************
*** 2563,2571 ****
while (top >= bot)
{
mid = (bot + top) / 2;
! if (classes[mid].last < c)
bot = mid + 1;
! else if (classes[mid].first > c)
top = mid - 1;
else
return (int)classes[mid].class;
--- 2567,2575 ----
while (top >= bot)
{
mid = (bot + top) / 2;
! if (classes[mid].last < (unsigned int)c)
bot = mid + 1;
! else if (classes[mid].first > (unsigned int)c)
top = mid - 1;
else
return (int)classes[mid].class;
*** ../vim-7.4.087/src/option.c 2013-11-08 04:30:06.000000000 +0100
--- src/option.c 2013-11-12 04:34:46.000000000 +0100
***************
*** 7122,7127 ****
--- 7122,7132 ----
if (varp == &(curwin->w_s->b_p_spl))
{
char_u fname[200];
+ char_u *q = curwin->w_s->b_p_spl;
+
+ /* Skip the first name if it is "cjk". */
+ if (STRNCMP(q, "cjk,", 4) == 0)
+ q += 4;
/*
* Source the spell/LANG.vim in 'runtimepath'.
***************
*** 7129,7139 ****
* Use the first name in 'spelllang' up to '_region' or
* '.encoding'.
*/
! for (p = curwin->w_s->b_p_spl; *p != NUL; ++p)
if (vim_strchr((char_u *)"_.,", *p) != NULL)
break;
! vim_snprintf((char *)fname, 200, "spell/%.*s.vim",
! (int)(p - curwin->w_s->b_p_spl),
curwin->w_s->b_p_spl);
source_runtime(fname, TRUE);
}
#endif
--- 7134,7143 ----
* Use the first name in 'spelllang' up to '_region' or
* '.encoding'.
*/
! for (p = q; *p != NUL; ++p)
if (vim_strchr((char_u *)"_.,", *p) != NULL)
break;
! vim_snprintf((char *)fname, 200, "spell/%.*s.vim", (int)(p - q), q);
source_runtime(fname, TRUE);
}
#endif
*** ../vim-7.4.087/src/spell.c 2013-09-29 13:38:25.000000000 +0200
--- src/spell.c 2013-11-12 04:37:33.000000000 +0100
***************
*** 754,762 ****
static void clear_spell_chartab __ARGS((spelltab_T *sp));
static int set_spell_finish __ARGS((spelltab_T *new_st));
static int spell_iswordp __ARGS((char_u *p, win_T *wp));
! static int spell_iswordp_nmw __ARGS((char_u *p));
#ifdef FEAT_MBYTE
! static int spell_mb_isword_class __ARGS((int cl));
static int spell_iswordp_w __ARGS((int *p, win_T *wp));
#endif
static int write_spell_prefcond __ARGS((FILE *fd, garray_T *gap));
--- 754,762 ----
static void clear_spell_chartab __ARGS((spelltab_T *sp));
static int set_spell_finish __ARGS((spelltab_T *new_st));
static int spell_iswordp __ARGS((char_u *p, win_T *wp));
! static int spell_iswordp_nmw __ARGS((char_u *p, win_T *wp));
#ifdef FEAT_MBYTE
! static int spell_mb_isword_class __ARGS((int cl, win_T *wp));
static int spell_iswordp_w __ARGS((int *p, win_T *wp));
#endif
static int write_spell_prefcond __ARGS((FILE *fd, garray_T *gap));
***************
*** 1149,1155 ****
/* When we are at a non-word character there is no error, just
* skip over the character (try looking for a word after it). */
! else if (!spell_iswordp_nmw(ptr))
{
if (capcol != NULL && wp->w_s->b_cap_prog != NULL)
{
--- 1149,1155 ----
/* When we are at a non-word character there is no error, just
* skip over the character (try looking for a word after it). */
! else if (!spell_iswordp_nmw(ptr, wp))
{
if (capcol != NULL && wp->w_s->b_cap_prog != NULL)
{
***************
*** 1561,1567 ****
* accept a no-caps word, even when the dictionary
* word specifies ONECAP. */
mb_ptr_back(mip->mi_word, p);
! if (spell_iswordp_nmw(p)
? capflags == WF_ONECAP
: (flags & WF_ONECAP) != 0
&& capflags != WF_ONECAP)
--- 1561,1567 ----
* accept a no-caps word, even when the dictionary
* word specifies ONECAP. */
mb_ptr_back(mip->mi_word, p);
! if (spell_iswordp_nmw(p, mip->mi_win)
? capflags == WF_ONECAP
: (flags & WF_ONECAP) != 0
&& capflags != WF_ONECAP)
***************
*** 4234,4240 ****
if (spl_copy == NULL)
goto theend;
! /* loop over comma separated language names. */
for (splp = spl_copy; *splp != NUL; )
{
/* Get one language name. */
--- 4234,4242 ----
if (spl_copy == NULL)
goto theend;
! wp->w_s->b_cjk = 0;
!
! /* Loop over comma separated language names. */
for (splp = spl_copy; *splp != NUL; )
{
/* Get one language name. */
***************
*** 4242,4247 ****
--- 4244,4255 ----
region = NULL;
len = (int)STRLEN(lang);
+ if (STRCMP(lang, "cjk") == 0)
+ {
+ wp->w_s->b_cjk = 1;
+ continue;
+ }
+
/* If the name ends in ".spl" use it as the name of the spell file.
* If there is a region name let "region" point to it and remove it
* from the name. */
***************
*** 4601,4607 ****
int past_second = FALSE; /* past second word char */
/* find first letter */
! for (p = word; !spell_iswordp_nmw(p); mb_ptr_adv(p))
if (end == NULL ? *p == NUL : p >= end)
return 0; /* only non-word characters, illegal word */
#ifdef FEAT_MBYTE
--- 4609,4615 ----
int past_second = FALSE; /* past second word char */
/* find first letter */
! for (p = word; !spell_iswordp_nmw(p, curwin); mb_ptr_adv(p))
if (end == NULL ? *p == NUL : p >= end)
return 0; /* only non-word characters, illegal word */
#ifdef FEAT_MBYTE
***************
*** 4617,4623 ****
* But a word with an upper char only at start is a ONECAP.
*/
for ( ; end == NULL ? *p != NUL : p < end; mb_ptr_adv(p))
! if (spell_iswordp_nmw(p))
{
c = PTR2CHAR(p);
if (!SPELL_ISUPPER(c))
--- 4625,4631 ----
* But a word with an upper char only at start is a ONECAP.
*/
for ( ; end == NULL ? *p != NUL : p < end; mb_ptr_adv(p))
! if (spell_iswordp_nmw(p, curwin))
{
c = PTR2CHAR(p);
if (!SPELL_ISUPPER(c))
***************
*** 9907,9913 ****
c = mb_ptr2char(s);
if (c > 255)
! return spell_mb_isword_class(mb_get_class(s));
return spelltab.st_isw[c];
}
#endif
--- 9915,9921 ----
c = mb_ptr2char(s);
if (c > 255)
! return spell_mb_isword_class(mb_get_class(s), wp);
return spelltab.st_isw[c];
}
#endif
***************
*** 9920,9927 ****
* Unlike spell_iswordp() this doesn't check for "midword" characters.
*/
static int
! spell_iswordp_nmw(p)
char_u *p;
{
#ifdef FEAT_MBYTE
int c;
--- 9928,9936 ----
* Unlike spell_iswordp() this doesn't check for "midword" characters.
*/
static int
! spell_iswordp_nmw(p, wp)
char_u *p;
+ win_T *wp;
{
#ifdef FEAT_MBYTE
int c;
***************
*** 9930,9936 ****
{
c = mb_ptr2char(p);
if (c > 255)
! return spell_mb_isword_class(mb_get_class(p));
return spelltab.st_isw[c];
}
#endif
--- 9939,9945 ----
{
c = mb_ptr2char(p);
if (c > 255)
! return spell_mb_isword_class(mb_get_class(p), wp);
return spelltab.st_isw[c];
}
#endif
***************
*** 9942,9952 ****
* Return TRUE if word class indicates a word character.
* Only for characters above 255.
* Unicode subscript and superscript are not considered word characters.
*/
static int
! spell_mb_isword_class(cl)
! int cl;
{
return cl >= 2 && cl != 0x2070 && cl != 0x2080;
}
--- 9951,9966 ----
* Return TRUE if word class indicates a word character.
* Only for characters above 255.
* Unicode subscript and superscript are not considered word characters.
+ * See also dbcs_class() and utf_class() in mbyte.c.
*/
static int
! spell_mb_isword_class(cl, wp)
! int cl;
! win_T *wp;
{
+ if (wp->w_s->b_cjk)
+ /* East Asian characters are not considered word characters. */
+ return cl == 2 || cl == 0x2800;
return cl >= 2 && cl != 0x2070 && cl != 0x2080;
}
***************
*** 9971,9979 ****
if (*s > 255)
{
if (enc_utf8)
! return spell_mb_isword_class(utf_class(*s));
if (enc_dbcs)
! return dbcs_class((unsigned)*s >> 8, *s & 0xff) >= 2;
return 0;
}
return spelltab.st_isw[*s];
--- 9985,9994 ----
if (*s > 255)
{
if (enc_utf8)
! return spell_mb_isword_class(utf_class(*s), wp);
if (enc_dbcs)
! return spell_mb_isword_class(
! dbcs_class((unsigned)*s >> 8, *s & 0xff), wp);
return 0;
}
return spelltab.st_isw[*s];
***************
*** 10193,10205 ****
line = ml_get_curline();
p = line + curwin->w_cursor.col;
/* Backup to before start of word. */
! while (p > line && spell_iswordp_nmw(p))
mb_ptr_back(line, p);
/* Forward to start of word. */
! while (*p != NUL && !spell_iswordp_nmw(p))
mb_ptr_adv(p);
! if (!spell_iswordp_nmw(p)) /* No word found. */
{
beep_flush();
return;
--- 10208,10220 ----
line = ml_get_curline();
p = line + curwin->w_cursor.col;
/* Backup to before start of word. */
! while (p > line && spell_iswordp_nmw(p, curwin))
mb_ptr_back(line, p);
/* Forward to start of word. */
! while (*p != NUL && !spell_iswordp_nmw(p, curwin))
mb_ptr_adv(p);
! if (!spell_iswordp_nmw(p, curwin)) /* No word found. */
{
beep_flush();
return;
***************
*** 10436,10442 ****
for (;;)
{
mb_ptr_back(line, p);
! if (p == line || spell_iswordp_nmw(p))
break;
if (vim_regexec(®match, p, 0)
&& regmatch.endp[0] == line + endcol)
--- 10451,10457 ----
for (;;)
{
mb_ptr_back(line, p);
! if (p == line || spell_iswordp_nmw(p, curwin))
break;
if (vim_regexec(®match, p, 0)
&& regmatch.endp[0] == line + endcol)
***************
*** 11645,11651 ****
/* When appending a compound word after a word character don't
* use Onecap. */
! if (p != NULL && spell_iswordp_nmw(p))
c &= ~WF_ONECAP;
make_case_word(tword + sp->ts_splitoff,
preword + sp->ts_prewordlen, c);
--- 11660,11666 ----
/* When appending a compound word after a word character don't
* use Onecap. */
! if (p != NULL && spell_iswordp_nmw(p, curwin))
c &= ~WF_ONECAP;
make_case_word(tword + sp->ts_splitoff,
preword + sp->ts_prewordlen, c);
***************
*** 11895,11901 ****
* character when the word ends. But only when the
* good word can end. */
if (((!try_compound && !spell_iswordp_nmw(fword
! + sp->ts_fidx))
|| fword_ends)
&& fword[sp->ts_fidx] != NUL
&& goodword_ends)
--- 11910,11917 ----
* character when the word ends. But only when the
* good word can end. */
if (((!try_compound && !spell_iswordp_nmw(fword
! + sp->ts_fidx,
! curwin))
|| fword_ends)
&& fword[sp->ts_fidx] != NUL
&& goodword_ends)
***************
*** 14226,14232 ****
}
else
{
! if (spell_iswordp_nmw(s))
*t++ = *s;
++s;
}
--- 14242,14248 ----
}
else
{
! if (spell_iswordp_nmw(s, curwin))
*t++ = *s;
++s;
}
***************
*** 14521,14527 ****
else
{
did_white = FALSE;
! if (!spell_iswordp_nmw(t))
continue;
}
}
--- 14537,14543 ----
else
{
did_white = FALSE;
! if (!spell_iswordp_nmw(t, curwin))
continue;
}
}
***************
*** 16045,16051 ****
for (p = line + startcol; p > line; )
{
mb_ptr_back(line, p);
! if (spell_iswordp_nmw(p))
break;
}
--- 16061,16067 ----
for (p = line + startcol; p > line; )
{
mb_ptr_back(line, p);
! if (spell_iswordp_nmw(p, curwin))
break;
}
*** ../vim-7.4.087/src/structs.h 2013-11-09 05:30:18.000000000 +0100
--- src/structs.h 2013-11-12 03:55:50.000000000 +0100
***************
*** 1310,1315 ****
--- 1310,1318 ----
regprog_T *b_cap_prog; /* program for 'spellcapcheck' */
char_u *b_p_spf; /* 'spellfile' */
char_u *b_p_spl; /* 'spelllang' */
+ # ifdef FEAT_MBYTE
+ int b_cjk; /* all CJK letters as OK */
+ # endif
#endif
#if !defined(FEAT_SYN_HL) && !defined(FEAT_SPELL)
int dummy;
*** ../vim-7.4.087/src/version.c 2013-11-11 23:17:31.000000000 +0100
--- src/version.c 2013-11-12 03:59:03.000000000 +0100
***************
*** 740,741 ****
--- 740,743 ----
{ /* Add new patch number below this line */
+ /**/
+ 88,
/**/