summaryrefslogtreecommitdiff
path: root/strings
diff options
context:
space:
mode:
authorunknown <serg@serg.mylan>2004-06-10 21:18:57 +0200
committerunknown <serg@serg.mylan>2004-06-10 21:18:57 +0200
commit1e24da548bee6359ef93423afd412287a0a7da4f (patch)
tree3cc0388adca3b345d5907760ca63b1bb0e6f6bfc /strings
parent2e8fc0cae2af64fb6f9c840b97ef6bf1ca6914cf (diff)
downloadmariadb-git-1e24da548bee6359ef93423afd412287a0a7da4f.tar.gz
bug#3964 and related issues: FTB problems with charsets where one byte can match many
correct prefix compare with my_strnncoll include/m_ctype.h: 6th argument to my_strncoll to handle prefix comparison myisam/ft_boolean_search.c: bug#3964 and related issues: problems with charsets where one byte can match many *correct* prefix compare with my_strnncoll *correct* backup of info->lastkey mysql-test/r/fulltext.result: 6th argument to my_strncoll to handle prefix comparison mysql-test/t/fulltext.test: 6th argument to my_strncoll to handle prefix comparison mysys/my_handler.c: 6th argument to my_strncoll to handle prefix comparison sql/sql_parse.cc: cleanup strings/ctype-big5.c: 6th argument to my_strncoll to handle prefix comparison strings/ctype-bin.c: 6th argument to my_strncoll to handle prefix comparison strings/ctype-czech.c: 6th argument to my_strncoll to handle prefix comparison strings/ctype-gbk.c: 6th argument to my_strncoll to handle prefix comparison strings/ctype-latin1.c: 6th argument to my_strncoll to handle prefix comparison strings/ctype-mb.c: 6th argument to my_strncoll to handle prefix comparison strings/ctype-simple.c: 6th argument to my_strncoll to handle prefix comparison strings/ctype-sjis.c: 6th argument to my_strncoll to handle prefix comparison strings/ctype-tis620.c: 6th argument to my_strncoll to handle prefix comparison strings/ctype-uca.c: 6th argument to my_strncoll to handle prefix comparison strings/ctype-ucs2.c: 6th argument to my_strncoll to handle prefix comparison strings/ctype-utf8.c: 6th argument to my_strncoll to handle prefix comparison strings/ctype-win1250ch.c: 6th argument to my_strncoll to handle prefix comparison
Diffstat (limited to 'strings')
-rw-r--r--strings/ctype-big5.c5
-rw-r--r--strings/ctype-bin.c18
-rw-r--r--strings/ctype-czech.c8
-rw-r--r--strings/ctype-gbk.c5
-rw-r--r--strings/ctype-latin1.c5
-rw-r--r--strings/ctype-mb.c19
-rw-r--r--strings/ctype-simple.c7
-rw-r--r--strings/ctype-sjis.c5
-rw-r--r--strings/ctype-tis620.c6
-rw-r--r--strings/ctype-uca.c5
-rw-r--r--strings/ctype-ucs2.c30
-rw-r--r--strings/ctype-utf8.c5
-rw-r--r--strings/ctype-win1250ch.c15
13 files changed, 96 insertions, 37 deletions
diff --git a/strings/ctype-big5.c b/strings/ctype-big5.c
index f024fa0cc14..af8f5d49654 100644
--- a/strings/ctype-big5.c
+++ b/strings/ctype-big5.c
@@ -251,11 +251,12 @@ static int my_strnncoll_big5_internal(const uchar **a_res,
static int my_strnncoll_big5(CHARSET_INFO *cs __attribute__((unused)),
const uchar *a, uint a_length,
- const uchar *b, uint b_length)
+ const uchar *b, uint b_length,
+ my_bool b_is_prefix)
{
uint length= min(a_length, b_length);
int res= my_strnncoll_big5_internal(&a, &b, length);
- return res ? res : (int) (a_length - b_length);
+ return res ? res : (int)((b_is_prefix ? length : a_length) - b_length);
}
diff --git a/strings/ctype-bin.c b/strings/ctype-bin.c
index 7cac8c7c337..35630eb854f 100644
--- a/strings/ctype-bin.c
+++ b/strings/ctype-bin.c
@@ -91,10 +91,20 @@ static uchar bin_char_array[] =
static int my_strnncoll_binary(CHARSET_INFO * cs __attribute__((unused)),
const uchar *s, uint slen,
- const uchar *t, uint tlen)
+ const uchar *t, uint tlen,
+ my_bool t_is_prefix)
{
- int cmp= memcmp(s,t,min(slen,tlen));
- return cmp ? cmp : (int) (slen - tlen);
+ uint len=min(slen,tlen);
+ int cmp= memcmp(s,t,len);
+ return cmp ? cmp : (int)((t_is_prefix ? len : slen) - tlen);
+}
+
+
+static int my_strnncollsp_binary(CHARSET_INFO * cs __attribute__((unused)),
+ const uchar *s, uint slen,
+ const uchar *t, uint tlen)
+{
+ return my_strnncoll_binary(cs,s,slen,t,tlen,0);
}
@@ -334,7 +344,7 @@ skip:
MY_COLLATION_HANDLER my_collation_8bit_bin_handler =
{
my_strnncoll_binary,
- my_strnncoll_binary,
+ my_strnncollsp_binary,
my_strnxfrm_bin,
my_like_range_simple,
my_wildcmp_bin,
diff --git a/strings/ctype-czech.c b/strings/ctype-czech.c
index 2eb2fac46e9..9175451ceb7 100644
--- a/strings/ctype-czech.c
+++ b/strings/ctype-czech.c
@@ -242,12 +242,16 @@ while (1) \
static int my_strnncoll_czech(CHARSET_INFO *cs __attribute__((unused)),
const uchar * s1, uint len1,
- const uchar * s2, uint len2)
+ const uchar * s2, uint len2,
+ my_bool s2_is_prefix)
{
int v1, v2;
const uchar * p1, * p2, * store1, * store2;
int pass1 = 0, pass2 = 0;
+ if (s2_is_prefix && len1 > len2)
+ len1=len2;
+
p1 = s1; p2 = s2;
store1 = s1; store2 = s2;
@@ -276,7 +280,7 @@ int my_strnncollsp_czech(CHARSET_INFO * cs,
{
for ( ; slen && s[slen-1] == ' ' ; slen--);
for ( ; tlen && t[tlen-1] == ' ' ; tlen--);
- return my_strnncoll_czech(cs,s,slen,t,tlen);
+ return my_strnncoll_czech(cs,s,slen,t,tlen,0);
}
diff --git a/strings/ctype-gbk.c b/strings/ctype-gbk.c
index 0dc00a73fa3..18024db3f35 100644
--- a/strings/ctype-gbk.c
+++ b/strings/ctype-gbk.c
@@ -2614,11 +2614,12 @@ int my_strnncoll_gbk_internal(const uchar **a_res, const uchar **b_res,
int my_strnncoll_gbk(CHARSET_INFO *cs __attribute__((unused)),
const uchar *a, uint a_length,
- const uchar *b, uint b_length)
+ const uchar *b, uint b_length,
+ my_bool b_is_prefix)
{
uint length= min(a_length, b_length);
int res= my_strnncoll_gbk_internal(&a, &b, length);
- return res ? res : (int) (a_length - b_length);
+ return res ? res : (int) ((b_is_prefix ? length : a_length) - b_length);
}
diff --git a/strings/ctype-latin1.c b/strings/ctype-latin1.c
index 0b439964c7c..257d8e37eee 100644
--- a/strings/ctype-latin1.c
+++ b/strings/ctype-latin1.c
@@ -525,7 +525,8 @@ uchar combo2map[]={
static int my_strnncoll_latin1_de(CHARSET_INFO *cs __attribute__((unused)),
const uchar *a, uint a_length,
- const uchar *b, uint b_length)
+ const uchar *b, uint b_length,
+ my_bool b_is_prefix)
{
const uchar *a_end= a + a_length;
const uchar *b_end= b + b_length;
@@ -558,7 +559,7 @@ static int my_strnncoll_latin1_de(CHARSET_INFO *cs __attribute__((unused)),
A simple test of string lengths won't work -- we test to see
which string ran out first
*/
- return ((a < a_end || a_extend) ? 1 :
+ return ((a < a_end || a_extend) ? (b_is_prefix ? 0 : 1) :
(b < b_end || b_extend) ? -1 : 0);
}
diff --git a/strings/ctype-mb.c b/strings/ctype-mb.c
index 9b02cd3b3da..129e2f2d9a9 100644
--- a/strings/ctype-mb.c
+++ b/strings/ctype-mb.c
@@ -322,7 +322,7 @@ uint my_instr_mb(CHARSET_INFO *cs,
int mblen;
if (!cs->coll->strnncoll(cs, (unsigned char*) b, s_length,
- (unsigned char*) s, s_length))
+ (unsigned char*) s, s_length, 0))
{
if (nmatch)
{
@@ -352,10 +352,19 @@ uint my_instr_mb(CHARSET_INFO *cs,
static int my_strnncoll_mb_bin(CHARSET_INFO * cs __attribute__((unused)),
const uchar *s, uint slen,
- const uchar *t, uint tlen)
+ const uchar *t, uint tlen,
+ my_bool t_is_prefix)
{
- int cmp= memcmp(s,t,min(slen,tlen));
- return cmp ? cmp : (int) (slen - tlen);
+ uint len=min(slen,tlen);
+ int cmp= memcmp(s,t,len);
+ return cmp ? cmp : (int) ((t_is_prefix ? len : slen) - tlen);
+}
+
+static int my_strnncollsp_mb_bin(CHARSET_INFO * cs __attribute__((unused)),
+ const uchar *s, uint slen,
+ const uchar *t, uint tlen)
+{
+ return my_strnncoll_mb_bin(cs,s,slen,t,tlen,0);
}
@@ -513,7 +522,7 @@ static int my_wildcmp_mb_bin(CHARSET_INFO *cs,
MY_COLLATION_HANDLER my_collation_mb_bin_handler =
{
my_strnncoll_mb_bin,
- my_strnncoll_mb_bin,
+ my_strnncollsp_mb_bin,
my_strnxfrm_mb_bin,
my_like_range_simple,
my_wildcmp_mb_bin,
diff --git a/strings/ctype-simple.c b/strings/ctype-simple.c
index ba1fc1c424a..3ef8a9d5997 100644
--- a/strings/ctype-simple.c
+++ b/strings/ctype-simple.c
@@ -47,16 +47,19 @@ int my_strnxfrm_simple(CHARSET_INFO * cs,
}
int my_strnncoll_simple(CHARSET_INFO * cs, const uchar *s, uint slen,
- const uchar *t, uint tlen)
+ const uchar *t, uint tlen,
+ my_bool t_is_prefix)
{
int len = ( slen > tlen ) ? tlen : slen;
uchar *map= cs->sort_order;
+ if (t_is_prefix && slen > tlen)
+ slen=tlen;
while (len--)
{
if (map[*s++] != map[*t++])
return ((int) map[s[-1]] - (int) map[t[-1]]);
}
- return (int) (slen-tlen);
+ return (int) (slen - tlen);
}
diff --git a/strings/ctype-sjis.c b/strings/ctype-sjis.c
index 72666175a1f..cfe59bb2852 100644
--- a/strings/ctype-sjis.c
+++ b/strings/ctype-sjis.c
@@ -232,9 +232,12 @@ static int my_strnncoll_sjis_internal(CHARSET_INFO *cs,
static int my_strnncoll_sjis(CHARSET_INFO *cs __attribute__((unused)),
const uchar *a, uint a_length,
- const uchar *b, uint b_length)
+ const uchar *b, uint b_length,
+ my_bool b_is_prefix)
{
int res= my_strnncoll_sjis_internal(cs, &a, a_length, &b, b_length);
+ if (b_is_prefix && a_length > b_length)
+ a_length= b_length;
return res ? res : (int) (a_length - b_length);
}
diff --git a/strings/ctype-tis620.c b/strings/ctype-tis620.c
index e2a138300c3..b9ee8825eef 100644
--- a/strings/ctype-tis620.c
+++ b/strings/ctype-tis620.c
@@ -529,12 +529,16 @@ static uint thai2sortable(uchar *tstr, uint len)
static
int my_strnncoll_tis620(CHARSET_INFO *cs __attribute__((unused)),
const uchar * s1, uint len1,
- const uchar * s2, uint len2)
+ const uchar * s2, uint len2,
+ my_bool s2_is_prefix)
{
uchar buf[80] ;
uchar *tc1, *tc2;
int i;
+ if (s2_is_prefix && len1 > len2)
+ len1= len2;
+
tc1= buf;
if ((len1 + len2 +2) > (int) sizeof(buf))
tc1= (uchar*) malloc(len1+len2);
diff --git a/strings/ctype-uca.c b/strings/ctype-uca.c
index 81073d47554..ba7d308ec5c 100644
--- a/strings/ctype-uca.c
+++ b/strings/ctype-uca.c
@@ -6704,7 +6704,8 @@ implicit:
static int my_strnncoll_uca(CHARSET_INFO *cs,
const uchar *s, uint slen,
- const uchar *t, uint tlen)
+ const uchar *t, uint tlen,
+ my_bool t_is_prefix)
{
my_uca_scanner sscanner;
my_uca_scanner tscanner;
@@ -6720,7 +6721,7 @@ static int my_strnncoll_uca(CHARSET_INFO *cs,
t_res= my_uca_scanner_next(&tscanner);
} while ( s_res == t_res && s_res >0);
- return ( s_res - t_res );
+ return (t_is_prefix && t_res < 0) ? 0 : (s_res - t_res);
}
diff --git a/strings/ctype-ucs2.c b/strings/ctype-ucs2.c
index 67340fdd4f4..040fc5f65fd 100644
--- a/strings/ctype-ucs2.c
+++ b/strings/ctype-ucs2.c
@@ -182,7 +182,8 @@ static void my_casedn_str_ucs2(CHARSET_INFO *cs __attribute__((unused)),
static int my_strnncoll_ucs2(CHARSET_INFO *cs,
const uchar *s, uint slen,
- const uchar *t, uint tlen)
+ const uchar *t, uint tlen,
+ my_bool t_is_prefix)
{
int s_res,t_res;
my_wc_t s_wc,t_wc;
@@ -213,7 +214,14 @@ static int my_strnncoll_ucs2(CHARSET_INFO *cs,
s+=s_res;
t+=t_res;
}
- return ( (se-s) - (te-t) );
+ return t_is_prefix ? t-te : ((se-s) - (te-t));
+}
+
+static int my_strnncollsp_ucs2(CHARSET_INFO *cs,
+ const uchar *s, uint slen,
+ const uchar *t, uint tlen)
+{
+ return my_strnncoll_ucs2(cs,s,slen,t,tlen,0);
}
@@ -1224,8 +1232,9 @@ int my_wildcmp_ucs2_bin(CHARSET_INFO *cs,
static
int my_strnncoll_ucs2_bin(CHARSET_INFO *cs,
- const uchar *s, uint slen,
- const uchar *t, uint tlen)
+ const uchar *s, uint slen,
+ const uchar *t, uint tlen,
+ my_bool t_is_prefix)
{
int s_res,t_res;
my_wc_t s_wc,t_wc;
@@ -1250,7 +1259,14 @@ int my_strnncoll_ucs2_bin(CHARSET_INFO *cs,
s+=s_res;
t+=t_res;
}
- return ( (se-s) - (te-t) );
+ return t_is_prefix ? t-te : ((se-s) - (te-t));
+}
+
+static int my_strnncollsp_ucs2_bin(CHARSET_INFO *cs,
+ const uchar *s, uint slen,
+ const uchar *t, uint tlen)
+{
+ return my_strnncoll_ucs2_bin(cs,s,slen,t,tlen,0);
}
@@ -1374,7 +1390,7 @@ my_bool my_like_range_ucs2(CHARSET_INFO *cs,
static MY_COLLATION_HANDLER my_collation_ucs2_general_ci_handler =
{
my_strnncoll_ucs2,
- my_strnncoll_ucs2,
+ my_strnncollsp_ucs2,
my_strnxfrm_ucs2,
my_like_range_ucs2,
my_wildcmp_ucs2_ci,
@@ -1387,7 +1403,7 @@ static MY_COLLATION_HANDLER my_collation_ucs2_general_ci_handler =
static MY_COLLATION_HANDLER my_collation_ucs2_bin_handler =
{
my_strnncoll_ucs2_bin,
- my_strnncoll_ucs2_bin,
+ my_strnncollsp_ucs2_bin,
my_strnxfrm_ucs2_bin,
my_like_range_simple,
my_wildcmp_ucs2_bin,
diff --git a/strings/ctype-utf8.c b/strings/ctype-utf8.c
index 29d2c5d1358..48882f356fd 100644
--- a/strings/ctype-utf8.c
+++ b/strings/ctype-utf8.c
@@ -1802,7 +1802,8 @@ static void my_casedn_str_utf8(CHARSET_INFO *cs, char * s)
static int my_strnncoll_utf8(CHARSET_INFO *cs,
const uchar *s, uint slen,
- const uchar *t, uint tlen)
+ const uchar *t, uint tlen,
+ my_bool t_is_prefix)
{
int s_res,t_res;
my_wc_t s_wc,t_wc;
@@ -1833,7 +1834,7 @@ static int my_strnncoll_utf8(CHARSET_INFO *cs,
s+=s_res;
t+=t_res;
}
- return ( (se-s) - (te-t) );
+ return t_is_prefix ? t-te : ((se-s) - (te-t));
}
diff --git a/strings/ctype-win1250ch.c b/strings/ctype-win1250ch.c
index 2eefb570170..65d20ceedc9 100644
--- a/strings/ctype-win1250ch.c
+++ b/strings/ctype-win1250ch.c
@@ -448,20 +448,25 @@ static struct wordvalue doubles[] = {
static int my_strnncoll_win1250ch(CHARSET_INFO *cs __attribute__((unused)),
const uchar * s1, uint len1,
- const uchar * s2, uint len2)
+ const uchar * s2, uint len2,
+ my_bool s2_is_prefix)
{
int v1, v2;
const uchar * p1, * p2;
int pass1 = 0, pass2 = 0;
int diff;
+ if (s2_is_prefix && len1 > len2)
+ len1=len2;
+
p1 = s1; p2 = s2;
- do {
+ do
+ {
NEXT_CMP_VALUE(s1, p1, pass1, v1, (int)len1);
NEXT_CMP_VALUE(s2, p2, pass2, v2, (int)len2);
- diff = v1 - v2;
- if (diff != 0) return diff;
+ if ((diff = v1 - v2))
+ return diff;
} while (v1);
return 0;
}
@@ -478,7 +483,7 @@ int my_strnncollsp_win1250ch(CHARSET_INFO * cs,
{
for ( ; slen && s[slen-1] == ' ' ; slen--);
for ( ; tlen && t[tlen-1] == ' ' ; tlen--);
- return my_strnncoll_win1250ch(cs,s,slen,t,tlen);
+ return my_strnncoll_win1250ch(cs,s,slen,t,tlen,0);
}