Skip to content

M
MariaDB
Commit 4c14f9f2 authored by Michael Widenius's avatar Michael Widenius
Browse files
Options

Added more general support for sorting 2 characters as one (contractions) 

Added support for Croatian sorting orders utf8_croatian_ci and ucs2_croatian_ci.
Patch done by Alexander Barkov. See http://www.collation-charts.org/articles/croatian.htm

mysql-test/r/ctype_uca.result:
  Added testing of Croatian sort order
mysql-test/t/ctype_uca.test:
  Added testing of Croatian sort order
parent 84911a9f
Hide whitespace changes
Inline Side-by-side
Showing with 476 additions and 80 deletions
include/m_ctype.h
View file @ 4c14f9f2
......@@ -49,6 +49,24 @@ typedef struct unicase_info_st
extern MY_UNICASE_INFO *my_unicase_default[256];
extern MY_UNICASE_INFO *my_unicase_turkish[256];
#define MY_UCA_MAX_CONTRACTION 4
#define MY_UCA_MAX_WEIGHT_SIZE 8
typedef struct my_contraction_t
{
my_wc_t ch[MY_UCA_MAX_CONTRACTION]; /* Character sequence */
uint16 weight[MY_UCA_MAX_WEIGHT_SIZE];/* Its weight string, 0-terminated */
} MY_CONTRACTION;
typedef struct my_contraction_list_t
{
size_t nitems; /* Number of items in the list */
MY_CONTRACTION *item; /* List of contractions */
char *flags; /* Character flags, e.g. "is contraction head") */
} MY_CONTRACTIONS;
typedef struct uni_ctype_st
{
uchar pctype;
......@@ -262,7 +280,7 @@ typedef struct charset_info_st
uchar *to_lower;
uchar *to_upper;
uchar *sort_order;
uint16 *contractions;
MY_CONTRACTIONS *contractions;
uint16 **sort_order_big;
uint16 *tab_to_uni;
MY_UNI_IDX *tab_from_uni;
......@@ -475,6 +493,13 @@ my_bool my_charset_is_ascii_based(CHARSET_INFO *cs);
my_bool my_charset_is_8bit_pure_ascii(CHARSET_INFO *cs);
uint my_charset_repertoire(CHARSET_INFO *cs);
my_bool my_uca_have_contractions(CHARSET_INFO *cs);
my_bool my_uca_can_be_contraction_head(CHARSET_INFO *cs, my_wc_t wc);
my_bool my_uca_can_be_contraction_tail(CHARSET_INFO *cs, my_wc_t wc);
uint16 *my_uca_contraction2_weight(CHARSET_INFO *cs, my_wc_t wc1, my_wc_t wc2);
#define _MY_U 01 /* Upper case */
#define _MY_L 02 /* Lower case */
......
mysql-test/r/ctype_uca.result
View file @ 4c14f9f2
......@@ -159,6 +159,7 @@ insert into t1 values (_ucs2 0x01fc),(_ucs2 0x01fd),(_ucs2 0x01fe),(_ucs2 0x01ff
insert into t1 values ('AA'),('Aa'),('aa'),('aA');
insert into t1 values ('CH'),('Ch'),('ch'),('cH');
insert into t1 values ('DZ'),('Dz'),('dz'),('dZ');
insert into t1 values ('DŽ'),('Dž'),('dž'),('dŽ');
insert into t1 values ('IJ'),('Ij'),('ij'),('iJ');
insert into t1 values ('LJ'),('Lj'),('lj'),('lJ');
insert into t1 values ('LL'),('Ll'),('ll'),('lL');
......@@ -181,7 +182,7 @@ C,c,Ç,ç,Ć,ć,Ĉ,ĉ,Ċ,ċ,Č,č
CH,Ch,cH,ch
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DŽ,Dž,dž,DZ,Dz,dz
DZ,Dz,DŽ,Dž,dZ,dz,dŽ,dž,DŽ,Dž,dž,DZ,Dz,dz
Đ,đ
Ɖ
Ɗ
......@@ -286,7 +287,7 @@ C,c,Ç,ç,Ć,ć,Ĉ,ĉ,Ċ,ċ,Č,č
CH,Ch,cH,ch
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DŽ,Dž,dž,DZ,Dz,dz
DZ,Dz,DŽ,Dž,dZ,dz,dŽ,dž,DŽ,Dž,dž,DZ,Dz,dz
Ð,ð
Đ,đ
Ɖ
......@@ -400,6 +401,7 @@ CH,Ch,cH,ch
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DŽ,Dž,dž,DZ,Dz,dz
DŽ,Dž,dŽ,dž
Đ,đ
Ɖ
Ɗ
......@@ -513,7 +515,7 @@ C,c,Ç,ç,Ć,ć,Ĉ,ĉ,Ċ,ċ,Č,č
CH,Ch,cH,ch
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DŽ,Dž,dž,DZ,Dz,dz
DZ,Dz,DŽ,Dž,dZ,dz,dŽ,dž,DŽ,Dž,dž,DZ,Dz,dz
Đ,đ
Ɖ
Ɗ
......@@ -622,6 +624,7 @@ CH,Ch,cH,ch
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DŽ,Dž,dž,DZ,Dz,dz
DŽ,Dž,dŽ,dž
Đ,đ
Ɖ
Ɗ
......@@ -729,7 +732,7 @@ CH,Ch,cH,ch
Ć,ć
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DŽ,Dž,dž,DZ,Dz,dz
DZ,Dz,DŽ,Dž,dZ,dz,dŽ,dž,DŽ,Dž,dž,DZ,Dz,dz
Đ,đ
Ɖ
Ɗ
......@@ -840,6 +843,7 @@ CH,Ch,cH,ch
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz
DŽ,Dž,dŽ,dž
DŽ,Dž,dž,DZ,Dz,dz
Đ,đ
Ɖ
......@@ -951,7 +955,7 @@ C,c,Ç,ç,Ć,ć,Ĉ,ĉ,Ċ,ċ,Č,č
CH,Ch,cH,ch
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DŽ,Dž,dž,DZ,Dz,dz
DZ,Dz,DŽ,Dž,dZ,dz,dŽ,dž,DŽ,Dž,dž,DZ,Dz,dz
Đ,đ
Ɖ
Ɗ
......@@ -1056,7 +1060,7 @@ C,c,Ç,ç,Ć,ć,Ĉ,ĉ,Ċ,ċ,Č,č
CH,Ch,cH,ch
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DŽ,Dž,dž,DZ,Dz,dz
DZ,Dz,DŽ,Dž,dZ,dz,dŽ,dž,DŽ,Dž,dž,DZ,Dz,dz
Đ,đ
Ɖ
Ɗ
......@@ -1164,7 +1168,7 @@ CH,Ch,cH,ch
Ç,ç
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DŽ,Dž,dž,DZ,Dz,dz
DZ,Dz,DŽ,Dž,dZ,dz,dŽ,dž,DŽ,Dž,dž,DZ,Dz,dz
Đ,đ
Ɖ
Ɗ
......@@ -1275,6 +1279,7 @@ cH
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DŽ,Dž,dž,DZ,Dz,dz
DŽ,Dž,dŽ,dž
Đ,đ
Ɖ
Ɗ
......@@ -1382,7 +1387,7 @@ C,c,Ç,ç,Ć,ć,Ĉ,ĉ,Ċ,ċ,Č,č
CH,Ch,cH,ch
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DŽ,Dž,dž,DZ,Dz,dz
DZ,Dz,DŽ,Dž,dZ,dz,dŽ,dž,DŽ,Dž,dž,DZ,Dz,dz
Đ,đ
Ɖ
Ɗ
......@@ -1491,6 +1496,7 @@ cH
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DŽ,Dž,dž,DZ,Dz,dz
DŽ,Dž,dŽ,dž
Đ,đ
Ɖ
Ɗ
......@@ -1599,6 +1605,7 @@ cH
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DŽ,Dž,dž,DZ,Dz,dz
DŽ,Dž,dŽ,dž
Đ,đ
Ɖ
Ɗ
......@@ -1707,7 +1714,7 @@ cH
CH,Ch,ch
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DŽ,Dž,dž,DZ,Dz,dz
DZ,Dz,DŽ,Dž,dZ,dz,dŽ,dž,DŽ,Dž,dž,DZ,Dz,dz
Đ,đ
Ɖ
Ɗ
......@@ -1813,7 +1820,7 @@ C,c,Ç,ç,Ć,ć,Ĉ,ĉ,Ċ,ċ,Č,č
CH,Ch,cH,ch
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DŽ,Dž,dž,DZ,Dz,dz
DZ,Dz,DŽ,Dž,dZ,dz,dŽ,dž,DŽ,Dž,dž,DZ,Dz,dz
Đ,đ
Ɖ
Ɗ
......@@ -1921,7 +1928,7 @@ CH,Ch,cH,ch
Ĉ,ĉ
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DŽ,Dž,dž,DZ,Dz,dz
DZ,Dz,DŽ,Dž,dZ,dz,dŽ,dž,DŽ,Dž,dž,DZ,Dz,dz
Đ,đ
Ɖ
Ɗ
......@@ -2030,7 +2037,7 @@ C,c,Ç,ç,Ć,ć,Ĉ,ĉ,Ċ,ċ,Č,č
CH,Ch,cH,ch
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DŽ,Dž,dž,DZ,Dz,dz
DZ,Dz,DŽ,Dž,dZ,dz,dŽ,dž,DŽ,Dž,dž,DZ,Dz,dz
Đ,đ
Ɖ
Ɗ
......@@ -2121,6 +2128,118 @@ Z,z,Ź,ź,Ż,ż,Ž,ž
ǁ
ǂ
ǃ
select group_concat(c1 order by c1) from t1 group by c1 collate utf8_croatian_ci;
group_concat(c1 order by c1)
÷
×
A,a,À,Á,Â,Ã,Ä,Å,à,á,â,ã,ä,å,Ā,ā,Ă,ă,Ą,ą,Ǎ,ǎ,Ǟ,ǟ,Ǡ,ǡ,Ǻ,ǻ
AA,Aa,aA,aa
Æ,æ,Ǣ,ǣ,Ǽ,ǽ
B,b
ƀ
Ɓ
Ƃ,ƃ
C,c,Ç,ç,Ĉ,ĉ,Ċ,ċ
CH,Ch,cH,ch
Č,č
Ć,ć
Ƈ,ƈ
D,d,Ď,ď
DZ,Dz,dZ,dz,DZ,Dz,dz
dŽ
DŽ,Dž,dž,DŽ,Dž,dž
Đ,đ
Ɖ
Ɗ
Ƌ,ƌ
Ð,ð
E,e,È,É,Ê,Ë,è,é,ê,ë,Ē,ē,Ĕ,ĕ,Ė,ė,Ę,ę,Ě,ě
Ǝ,ǝ
Ə
Ɛ
F,f
Ƒ,ƒ
G,g,Ĝ,ĝ,Ğ,ğ,Ġ,ġ,Ģ,ģ,Ǧ,ǧ,Ǵ,ǵ
Ǥ,ǥ
Ɠ
Ɣ
Ƣ,ƣ
H,h,Ĥ,ĥ
ƕ,Ƕ
Ħ,ħ
I,i,Ì,Í,Î,Ï,ì,í,î,ï,Ĩ,ĩ,Ī,ī,Ĭ,ĭ,Į,į,İ,Ǐ,ǐ
IJ,Ij,iJ,ij,IJ,ij
ı
Ɨ
Ɩ
J,j,Ĵ,ĵ,ǰ
K,k,Ķ,ķ,Ǩ,ǩ
Ƙ,ƙ
L,l,Ĺ,ĺ,Ļ,ļ,Ľ,ľ
Ŀ,ŀ
lJ
LL,Ll,lL,ll
LJ,Lj,lj,LJ,Lj,lj
Ł,ł
ƚ
ƛ
M,m
N,n,Ñ,ñ,Ń,ń,Ņ,ņ,Ň,ň,Ǹ,ǹ
nJ
NJ,Nj,nj,NJ,Nj,nj
Ɲ
ƞ
Ŋ,ŋ
O,o,Ò,Ó,Ô,Õ,Ö,ò,ó,ô,õ,ö,Ō,ō,Ŏ,ŏ,Ő,ő,Ơ,ơ,Ǒ,ǒ,Ǫ,ǫ,Ǭ,ǭ
OE,Oe,oE,oe,Œ,œ
Ø,ø,Ǿ,ǿ
Ɔ
Ɵ
P,p
Ƥ,ƥ
Q,q
ĸ
R,r,Ŕ,ŕ,Ŗ,ŗ,Ř,ř
RR,Rr,rR,rr
Ʀ
S,s,Ś,ś,Ŝ,ŝ,Ş,ş,ſ
SS,Ss,sS,ss,ß
Š,š
Ʃ
ƪ
T,t,Ţ,ţ,Ť,ť
ƾ
Ŧ,ŧ
ƫ
Ƭ,ƭ
Ʈ
U,u,Ù,Ú,Û,Ü,ù,ú,û,ü,Ũ,ũ,Ū,ū,Ŭ,ŭ,Ů,ů,Ű,ű,Ų,ų,Ư,ư,Ǔ,ǔ,Ǖ,ǖ,Ǘ,ǘ,Ǚ,ǚ,Ǜ,ǜ
Ɯ
Ʊ
V,v
Ʋ
W,w,Ŵ,ŵ
X,x
Y,y,Ý,ý,ÿ,Ŷ,ŷ,Ÿ
Ƴ,ƴ
Z,z,Ź,ź,Ż,ż
ƍ
Ž,ž
Ƶ,ƶ
Ʒ,Ǯ,ǯ
Ƹ,ƹ
ƺ
Þ,þ
ƿ,Ƿ
ƻ
Ƨ,ƨ
Ƽ,ƽ
Ƅ,ƅ
ʼn
ǀ
ǁ
ǂ
ǃ
drop table t1;
SET NAMES utf8;
CREATE TABLE t1 (c varchar(255) NOT NULL COLLATE utf8_general_ci, INDEX (c));
......
mysql-test/t/ctype_uca.test
View file @ 4c14f9f2
......@@ -186,6 +186,7 @@ insert into t1 values (_ucs2 0x01fc),(_ucs2 0x01fd),(_ucs2 0x01fe),(_ucs2 0x01ff
insert into t1 values ('AA'),('Aa'),('aa'),('aA');
insert into t1 values ('CH'),('Ch'),('ch'),('cH');
insert into t1 values ('DZ'),('Dz'),('dz'),('dZ');
insert into t1 values ('DŽ'),('Dž'),('dž'),('dŽ');
insert into t1 values ('IJ'),('Ij'),('ij'),('iJ');
insert into t1 values ('LJ'),('Lj'),('lj'),('lJ');
insert into t1 values ('LL'),('Ll'),('ll'),('lL');
......@@ -213,6 +214,7 @@ select group_concat(c1 order by c1) from t1 group by c1 collate utf8_spanish2_ci
select group_concat(c1 order by c1) from t1 group by c1 collate utf8_roman_ci;
select group_concat(c1 order by c1) from t1 group by c1 collate utf8_esperanto_ci;
select group_concat(c1 order by c1) from t1 group by c1 collate utf8_hungarian_ci;
select group_concat(c1 order by c1) from t1 group by c1 collate utf8_croatian_ci;
drop table t1;
......
mysys/charset-def.c
View file @ 4c14f9f2
......@@ -42,6 +42,7 @@ extern CHARSET_INFO my_charset_ucs2_roman_uca_ci;
extern CHARSET_INFO my_charset_ucs2_persian_uca_ci;
extern CHARSET_INFO my_charset_ucs2_esperanto_uca_ci;
extern CHARSET_INFO my_charset_ucs2_hungarian_uca_ci;
extern CHARSET_INFO my_charset_ucs2_croatian_uca_ci;
#endif
#ifdef HAVE_CHARSET_utf8
......@@ -63,6 +64,7 @@ extern CHARSET_INFO my_charset_utf8_roman_uca_ci;
extern CHARSET_INFO my_charset_utf8_persian_uca_ci;
extern CHARSET_INFO my_charset_utf8_esperanto_uca_ci;
extern CHARSET_INFO my_charset_utf8_hungarian_uca_ci;
extern CHARSET_INFO my_charset_utf8_croatian_uca_ci;
#ifdef HAVE_UTF8_GENERAL_CS
extern CHARSET_INFO my_charset_utf8_general_cs;
#endif
......@@ -152,6 +154,7 @@ my_bool init_compiled_charsets(myf flags __attribute__((unused)))
add_compiled_collation(&my_charset_ucs2_persian_uca_ci);
add_compiled_collation(&my_charset_ucs2_esperanto_uca_ci);
add_compiled_collation(&my_charset_ucs2_hungarian_uca_ci);
add_compiled_collation(&my_charset_ucs2_croatian_uca_ci);
#endif
#endif
......@@ -186,6 +189,7 @@ my_bool init_compiled_charsets(myf flags __attribute__((unused)))
add_compiled_collation(&my_charset_utf8_persian_uca_ci);
add_compiled_collation(&my_charset_utf8_esperanto_uca_ci);
add_compiled_collation(&my_charset_utf8_hungarian_uca_ci);
add_compiled_collation(&my_charset_utf8_croatian_uca_ci);
#endif
#endif
......
strings/ctype-mb.c
View file @ 4c14f9f2
......@@ -567,8 +567,7 @@ my_bool my_like_range_mb(CHARSET_INFO *cs,
char *min_end= min_str + res_length;
char *max_end= max_str + res_length;
size_t maxcharlen= res_length / cs->mbmaxlen;
const char *contraction_flags= cs->contractions ?
((const char*) cs->contractions) + 0x40*0x40 : NULL;
my_bool have_contractions= my_uca_have_contractions(cs);
for (; ptr != end && min_str != min_end && maxcharlen ; maxcharlen--)
{
......@@ -636,8 +635,8 @@ my_bool my_like_range_mb(CHARSET_INFO *cs,
'ab\min\min\min\min' and 'ab\max\max\max\max'.
*/
if (contraction_flags && ptr + 1 < end &&
contraction_flags[(uchar) *ptr])
if (have_contractions && ptr + 1 < end &&
my_uca_can_be_contraction_head(cs, (uchar) *ptr))
{
/* Ptr[0] is a contraction head. */
......@@ -659,8 +658,8 @@ my_bool my_like_range_mb(CHARSET_INFO *cs,
is not a contraction, then we put only ptr[0],
and continue with ptr[1] on the next loop.
*/
if (contraction_flags[(uchar) ptr[1]] &&
cs->contractions[(*ptr-0x40)*0x40 + ptr[1] - 0x40])
if (my_uca_can_be_contraction_tail(cs, (uchar) ptr[1]) &&
my_uca_contraction2_weight(cs, (uchar) ptr[0], (uchar) ptr[1]))
{
/* Contraction found */
if (maxcharlen == 1 || min_str + 1 >= min_end)
......
strings/ctype-uca.c
View file @ 4c14f9f2
......@@ -6713,6 +6713,16 @@ static const char hungarian[]=
"&U < \\u00FC <<< \\u00DC << \\u0171 <<< \\u0170";
static const char croatian[]=
"&C < \\u010D <<< \\u010C < \\u0107 <<< \\u0106 "
"&D < d\\u017E <<< \\u01C6 <<< D\\u017E <<< \\u01C5 <<< D\\u017D <<< \\u01C4 "
" < \\u0111 <<< \\u0110 "
"&L < lj <<< \\u01C9 <<< Lj <<< \\u01C8 <<< LJ <<< \\u01C7 "
"&N < nj <<< \\u01CC <<< Nj <<< \\u01CB <<< NJ <<< \\u01CA "
"&S < \\u0161 <<< \\u0160 "
"&Z < \\u017E <<< \\u017D";
/*
Unicode Collation Algorithm:
Collation element (weight) scanner,
......@@ -6726,7 +6736,7 @@ typedef struct my_uca_scanner_st
const uchar *send; /* End of the input string */
uchar *uca_length;
uint16 **uca_weight;
uint16 *contractions;
MY_CONTRACTIONS *contractions;
uint16 implicit[2];
int page;
int code;
......@@ -6747,6 +6757,164 @@ typedef struct my_uca_scanner_handler_st
static uint16 nochar[]= {0,0};
#define MY_UCA_CNT_FLAG_SIZE 4096
#define MY_UCA_CNT_FLAG_MASK 4095
#define MY_UCA_CNT_HEAD 1
#define MY_UCA_CNT_TAIL 2
/********** Helper functions to handle contraction ************/
/**
Mark a character as a contraction part
@cs Pointer to CHARSET_INFO data
@wc Unicode code point
@flag flag: "is contraction head", "is contraction tail"
*/
static void
my_uca_add_contraction_flag(CHARSET_INFO *cs, my_wc_t wc, int flag)
{
cs->contractions->flags[wc & MY_UCA_CNT_FLAG_MASK]|= flag;
}
/**
Add a new contraction into contraction list
@cs Pointer to CHARSET_INFO data
@wc Unicode code points of the characters
@len Number of characters
@return New contraction
@retval Pointer to a newly added contraction
*/
static MY_CONTRACTION *
my_uca_add_contraction(CHARSET_INFO *cs,
my_wc_t *wc, int len __attribute__((unused)))
{
MY_CONTRACTIONS *list= cs->contractions;
MY_CONTRACTION *next= &list->item[list->nitems];
DBUG_ASSERT(len == 2); /* We currently support only contraction2 */
next->ch[0]= wc[0];
next->ch[1]= wc[1];
list->nitems++;
return next;
}
/**
Allocate and initialize memory for contraction list and flags
@cs Pointer to CHARSET_INFO data
@alloc Memory allocation function (typically points to my_alloc_once)
@n Number of contractions
@return Error code
@retval 0 - memory allocated successfully
@retval 1 - not enough memory
*/
static my_bool
my_uca_alloc_contractions(CHARSET_INFO *cs, void *(*alloc)(size_t), size_t n)
{
uint size= n * sizeof(MY_CONTRACTION);
if (!(cs->contractions= (*alloc)(sizeof(MY_CONTRACTIONS))))
return 1;
bzero(cs->contractions, sizeof(MY_CONTRACTIONS));
if (!(cs->contractions->item= (*alloc)(size)) ||
!(cs->contractions->flags= (char*) (*alloc)(MY_UCA_CNT_FLAG_SIZE)))
return 1;
bzero((void*) cs->contractions->item, size);
bzero((void*) cs->contractions->flags, MY_UCA_CNT_FLAG_SIZE);
return 0;
}
/**
Check if UCA data has contractions (public version)
@cs Pointer to CHARSET_INFO data
@retval 0 - no contraction, 1 - have contractions.
*/
my_bool
my_uca_have_contractions(CHARSET_INFO *cs)
{
return cs->contractions != NULL;
}
/**
Check if a character can be contraction head
@cs Pointer to CHARSET_INFO data
@wc Code point
@retval 0 - cannot be contraction head
@retval 1 - can be contraction head
*/
my_bool
my_uca_can_be_contraction_head(CHARSET_INFO *cs, my_wc_t wc)
{
return cs->contractions->flags[wc & MY_UCA_CNT_FLAG_MASK] & MY_UCA_CNT_HEAD;
}
/**
Check if a character can be contraction tail
@cs Pointer to CHARSET_INFO data
@wc Code point
@retval 0 - cannot be contraction tail
@retval 1 - can be contraction tail
*/
my_bool
my_uca_can_be_contraction_tail(CHARSET_INFO *cs, my_wc_t wc)
{
return cs->contractions->flags[wc & MY_UCA_CNT_FLAG_MASK] & MY_UCA_CNT_TAIL;
}
/**
Find a contraction and return its weight array
@cs Pointer to CHARSET data
@wc1 First character
@wc2 Second character
@return Weight array
@retval NULL - no contraction found
@retval ptr - contraction weight array
*/
uint16 *
my_uca_contraction2_weight(CHARSET_INFO *cs, my_wc_t wc1, my_wc_t wc2)
{
MY_CONTRACTIONS *list= cs->contractions;
MY_CONTRACTION *c, *last;
for (c= list->item, last= &list->item[list->nitems]; c < last; c++)
{
if (c->ch[0] == wc1 && c->ch[1] == wc2)
{
return c->weight;
}
}
return NULL;
}
#ifdef HAVE_CHARSET_ucs2
/*
Initialize collation weight scanner
......@@ -6766,7 +6934,7 @@ static uint16 nochar[]= {0,0};
*/
static void my_uca_scanner_init_ucs2(my_uca_scanner *scanner,
CHARSET_INFO *cs __attribute__((unused)),
CHARSET_INFO *cs,
const uchar *str, size_t length)
{
scanner->wbeg= nochar;
......@@ -6777,6 +6945,7 @@ static void my_uca_scanner_init_ucs2(my_uca_scanner *scanner,
scanner->uca_length= cs->sort_order;
scanner->uca_weight= cs->sort_order_big;
scanner->contractions= cs->contractions;
scanner->cs= cs;
return;
}
......@@ -6865,18 +7034,23 @@ static int my_uca_scanner_next_ucs2(my_uca_scanner *scanner)
if (scanner->contractions && (scanner->sbeg <= scanner->send))
{
int cweight;
my_wc_t wc1= ((scanner->page << 8) | scanner->code);
if (!scanner->page && !scanner->sbeg[0] &&
(scanner->sbeg[1] > 0x40) && (scanner->sbeg[1] < 0x80) &&
(scanner->code > 0x40) && (scanner->code < 0x80) &&
(cweight= scanner->contractions[(scanner->code-0x40)*0x40+scanner->sbeg[1]-0x40]))
if (my_uca_can_be_contraction_head(scanner->cs, wc1))
{
uint16 *cweight;
my_wc_t wc2= (((my_wc_t) scanner->sbeg[0]) << 8) | scanner->sbeg[1];
if (my_uca_can_be_contraction_tail(scanner->cs, wc2) &&
(cweight= my_uca_contraction2_weight(scanner->cs,
scanner->code,
scanner->sbeg[1])))
{
scanner->implicit[0]= 0;
scanner->wbeg= scanner->implicit;
scanner->sbeg+=2;
return cweight;
return *cweight;
}
}
}
if (!ucaw[scanner->page])
......@@ -6959,23 +7133,22 @@ static int my_uca_scanner_next_any(my_uca_scanner *scanner)
scanner->code= wc & 0xFF;
scanner->sbeg+= mb_len;
if (scanner->contractions && !scanner->page &&
(scanner->code > 0x40) && (scanner->code < 0x80))
if (my_uca_have_contractions(scanner->cs) &&
my_uca_can_be_contraction_head(scanner->cs, wc))
{
uint page1, code1, cweight;
my_wc_t wc2;
uint16 *cweight;
if (((mb_len= scanner->cs->cset->mb_wc(scanner->cs, &wc,
if (((mb_len= scanner->cs->cset->mb_wc(scanner->cs, &wc2,
scanner->sbeg,
scanner->send)) >=0) &&
(!(page1= (wc >> 8))) &&
((code1= (wc & 0xFF)) > 0x40) &&
(code1 < 0x80) &&
(cweight= scanner->contractions[(scanner->code-0x40)*0x40 + code1-0x40]))
my_uca_can_be_contraction_tail(scanner->cs, wc2) &&
(cweight= my_uca_contraction2_weight(scanner->cs, wc, wc2)))
{
scanner->implicit[0]= 0;
scanner->wbeg= scanner->implicit;
scanner->sbeg+= mb_len;
return cweight;
return *cweight;
}
}
......@@ -7012,6 +7185,33 @@ static my_uca_scanner_handler my_any_uca_scanner_handler=
my_uca_scanner_next_any
};
/**
Helper function:
Find address of weights of the given character.
@weights UCA weight array
@lengths UCA length array
@ch character Unicode code point
@return Weight array
@retval pointer to weight array for the given character,
or NULL if this page does not have implicit weights.
*/
static inline uint16 *
my_char_weight_addr(CHARSET_INFO *cs, uint wc)
{
uint page= (wc >> 8);
uint ofst= wc & 0xFF;
return cs->sort_order_big[page] ?
cs->sort_order_big[page] + ofst * cs->sort_order[page] :
NULL;
}
/*
Compares two strings according to the collation
......@@ -7683,8 +7883,8 @@ static my_coll_lexem_num my_coll_lexem_next(MY_COLL_LEXEM *lexem)
typedef struct my_coll_rule_item_st
{
uint base; /* Base character */
uint curr[2]; /* Current character */
my_wc_t base; /* Base character */
my_wc_t curr[2]; /* Current character */
int diff[3]; /* Primary, Secondary and Tertiary difference */
} MY_COLL_RULE;
......@@ -7834,6 +8034,7 @@ static int my_coll_rule_parse(MY_COLL_RULE *rule, size_t mitems,
static my_bool create_tailoring(CHARSET_INFO *cs, void *(*alloc)(size_t))
{
MY_COLL_RULE rule[MY_MAX_COLL_RULE];
MY_COLL_RULE *r, *rfirst, *rlast;
char errstr[128];
uchar *newlengths;
uint16 **newweights;
......@@ -7858,6 +8059,9 @@ static my_bool create_tailoring(CHARSET_INFO *cs, void *(*alloc)(size_t))
return 1;
}
rfirst= rule;
rlast= rule + rc;
if (!cs->caseinfo)
cs->caseinfo= my_unicase_default;
......@@ -7941,44 +8145,21 @@ static my_bool create_tailoring(CHARSET_INFO *cs, void *(*alloc)(size_t))
/* Now process contractions */
if (ncontractions)
{
/*
8K for weights for basic latin letter pairs,
plus 256 bytes for "is contraction part" flags.
*/
uint size= 0x40*0x40*sizeof(uint16) + 256;
char *contraction_flags;
if (!(cs->contractions= (uint16*) (*alloc)(size)))
return 1;
bzero((void*)cs->contractions, size);
contraction_flags= ((char*) cs->contractions) + 0x40*0x40;
for (i=0; i < rc; i++)
if (my_uca_alloc_contractions(cs, alloc, ncontractions))
return 1;
for (r= rfirst; r < rlast; r++)
{
if (rule[i].curr[1])
uint16 *to;
if (r->curr[1]) /* Contraction */
{
uint pageb= (rule[i].base >> 8) & 0xFF;
uint chb= rule[i].base & 0xFF;
uint16 *offsb= defweights[pageb] + chb*deflengths[pageb];
uint offsc;
if (offsb[1] ||
rule[i].curr[0] < 0x40 || rule[i].curr[0] > 0x7f ||
rule[i].curr[1] < 0x40 || rule[i].curr[1] > 0x7f)
{
/*
TODO: add error reporting;
We support only basic latin letters contractions at this point.
Also, We don't support contractions with weight longer than one.
Otherwise, we'd need much more memory.
*/
return 1;
}
offsc= (rule[i].curr[0]-0x40)*0x40+(rule[i].curr[1]-0x40);
/* Copy base weight applying primary difference */
cs->contractions[offsc]= offsb[0] + rule[i].diff[0];
/* Mark both letters as "is contraction part */
contraction_flags[rule[i].curr[0]]= 1;
contraction_flags[rule[i].curr[1]]= 1;
/* Mark both letters as "is contraction part" */
my_uca_add_contraction_flag(cs, r->curr[0], MY_UCA_CNT_HEAD);
my_uca_add_contraction_flag(cs, r->curr[1], MY_UCA_CNT_TAIL);
to= my_uca_add_contraction(cs, r->curr, 2)->weight;
/* Copy weight from the reset character */
to[0]= my_char_weight_addr(cs, r->base)[0];
/* Apply primary difference */
to[0]+= r->diff[0];
}
}
}
......@@ -8701,6 +8882,39 @@ CHARSET_INFO my_charset_ucs2_hungarian_uca_ci=
};
CHARSET_INFO my_charset_ucs2_croatian_uca_ci=
{
149,0,0, /* number */
MY_CS_COMPILED|MY_CS_STRNXFRM|MY_CS_UNICODE,
"ucs2", /* cs name */
"ucs2_croatian_ci", /* name */
"", /* comment */
croatian, /* tailoring */
NULL, /* ctype */
NULL, /* to_lower */
NULL, /* to_upper */
NULL, /* sort_order */
NULL, /* contractions */
NULL, /* sort_order_big*/
NULL, /* tab_to_uni */
NULL, /* tab_from_uni */
my_unicase_default, /* caseinfo */
NULL, /* state_map */
NULL, /* ident_map */
8, /* strxfrm_multiply */
1, /* caseup_multiply */
1, /* casedn_multiply */
2, /* mbminlen */
2, /* mbmaxlen */
9, /* min_sort_char */
0xFFFF, /* max_sort_char */
' ', /* pad char */
0, /* escape_with_backslash_is_dangerous */
&my_charset_ucs2_handler,
&my_collation_ucs2_uca_handler
};
#endif
......@@ -9358,6 +9572,38 @@ CHARSET_INFO my_charset_utf8_hungarian_uca_ci=
&my_collation_any_uca_handler
};
CHARSET_INFO my_charset_utf8_croatian_uca_ci=
{
213,0,0, /* number */
MY_CS_COMPILED|MY_CS_STRNXFRM|MY_CS_UNICODE,
"utf8", /* cs name */
"utf8_croatian_ci", /* name */
"", /* comment */
croatian, /* tailoring */
ctype_utf8, /* ctype */
NULL, /* to_lower */
NULL, /* to_upper */
NULL, /* sort_order */
NULL, /* contractions */
NULL, /* sort_order_big*/
NULL, /* tab_to_uni */
NULL, /* tab_from_uni */
my_unicase_default, /* caseinfo */
NULL, /* state_map */
NULL, /* ident_map */
8, /* strxfrm_multiply */
1, /* caseup_multiply */
1, /* casedn_multiply */
1, /* mbminlen */
3, /* mbmaxlen */
9, /* min_sort_char */
0xFFFF, /* max_sort_char */
' ', /* pad char */
0, /* escape_with_backslash_is_dangerous */
&my_charset_utf8_handler,
&my_collation_any_uca_handler
};
#endif /* HAVE_CHARSET_utf8 */
#endif /* HAVE_UCA_COLLATIONS */
strings/ctype-ucs2.c
View file @ 4c14f9f2
......@@ -1526,8 +1526,7 @@ my_bool my_like_range_ucs2(CHARSET_INFO *cs,
char *min_org=min_str;
char *min_end=min_str+res_length;
size_t charlen= res_length / cs->mbmaxlen;
const char *contraction_flags= cs->contractions ?
((const char*) cs->contractions) + 0x40*0x40 : NULL;
my_bool have_contractions= my_uca_have_contractions(cs);
for ( ; ptr + 1 < end && min_str + 1 < min_end && charlen > 0
; ptr+=2, charlen--)
......@@ -1567,8 +1566,9 @@ my_bool my_like_range_ucs2(CHARSET_INFO *cs,
return 0;
}
if (contraction_flags && ptr + 3 < end &&
ptr[0] == '\0' && contraction_flags[(uchar) ptr[1]])
if (have_contractions && ptr + 3 < end &&
ptr[0] == '\0' &&
my_uca_can_be_contraction_head(cs, (uchar) ptr[1]))
{
/* Contraction head found */
if (ptr[2] == '\0' && (ptr[3] == w_one || ptr[3] == w_many))
......@@ -1581,8 +1581,9 @@ my_bool my_like_range_ucs2(CHARSET_INFO *cs,
Check if the second letter can be contraction part,
and if two letters really produce a contraction.
*/
if (ptr[2] == '\0' && contraction_flags[(uchar) ptr[3]] &&
cs->contractions[(ptr[1]-0x40)*0x40 + ptr[3] - 0x40])
if (ptr[2] == '\0' &&
my_uca_can_be_contraction_tail(cs, (uchar) ptr[3]) &&
my_uca_contraction2_weight(cs,(uchar) ptr[1], (uchar) ptr[3]))
{
/* Contraction found */
if (charlen == 1 || min_str + 2 >= min_end)
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7