Preparation for user-defined Unicode collations:

weights data now comes from a static variables
but from the charset structure.

include/m_ctype.h

@@ -188,7 +188,7 @@ typedef struct my_charset_handler_st
 } MY_CHARSET_HANDLER;
 
 extern MY_CHARSET_HANDLER my_charset_8bit_handler;
-
+extern MY_CHARSET_HANDLER my_charset_ucs2_handler;
 
 
 typedef struct charset_info_st
@@ -204,6 +204,7 @@ typedef struct charset_info_st
   uchar    *to_lower;
   uchar    *to_upper;
   uchar    *sort_order;
+  uint16   **sort_order_big;
   uint16      *tab_to_uni;
   MY_UNI_IDX  *tab_from_uni;
   uchar state_map[256];

mysys/charset.c

@@ -466,7 +466,8 @@ static my_bool init_available_charsets(myf myflags)
       if (*cs)
       {
         set_max_sort_char(*cs);
-        init_state_maps(*cs);
+        if (cs[0]->ctype)
+          init_state_maps(*cs);
       }
     }
     

strings/ctype-big5.c

@@ -6317,6 +6317,7 @@ CHARSET_INFO my_charset_big5_chinese_ci=
     to_lower_big5,
     to_upper_big5,
     sort_order_big5,
+    NULL,		/* sort_order_big*/
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     "",
@@ -6342,6 +6343,7 @@ CHARSET_INFO my_charset_big5_bin=
     to_lower_big5,
     to_upper_big5,
     sort_order_big5,
+    NULL,		/* sort_order_big*/
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     "",

strings/ctype-bin.c

@@ -384,7 +384,9 @@ CHARSET_INFO my_charset_bin =
     bin_char_array,		/* sort_order    */
     NULL,			/* tab_to_uni    */
     NULL,			/* tab_from_uni  */
-    "","",
+    NULL,			/* sort_order_big*/
+    "",
+    "",
     1,				/* strxfrm_multiply */
     1,				/* mbminlen      */
     1,				/* mbmaxlen      */

strings/ctype-czech.c

@@ -593,6 +593,7 @@ CHARSET_INFO my_charset_latin2_czech_ci =
     to_lower_czech,
     to_upper_czech,
     sort_order_czech,
+    NULL,		/* sort_order_big*/
     tab_8859_2_uni,	/* tab_to_uni   */
     idx_uni_8859_2,	/* tab_from_uni */
     "","",

strings/ctype-euc_kr.c

@@ -8685,6 +8685,7 @@ CHARSET_INFO my_charset_euckr_korean_ci=
     to_lower_euc_kr,
     to_upper_euc_kr,
     sort_order_euc_kr,
+    NULL,		/* sort_order_big*/
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     "",
@@ -8710,6 +8711,7 @@ CHARSET_INFO my_charset_euckr_bin=
     to_lower_euc_kr,
     to_upper_euc_kr,
     sort_order_euc_kr,
+    NULL,		/* sort_order_big*/
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     "",

strings/ctype-extra.c

@@ -28,6 +28,7 @@ CHARSET_INFO compiled_charsets[] = {
     NULL,
     NULL,
     NULL,
+    NULL,		/* sort_order_big*/
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     "","",

strings/ctype-gb2312.c

@@ -5738,6 +5738,7 @@ CHARSET_INFO my_charset_gb2312_chinese_ci=
     sort_order_gb2312,
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
+    NULL,		/* sort_order_big*/
     "",
     "",
     1,			/* strxfrm_multiply */
@@ -5762,6 +5763,7 @@ CHARSET_INFO my_charset_gb2312_bin=
     sort_order_gb2312,
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
+    NULL,		/* sort_order_big*/
     "",
     "",
     1,			/* strxfrm_multiply */

strings/ctype-gbk.c

@@ -9968,6 +9968,7 @@ CHARSET_INFO my_charset_gbk_chinese_ci=
     sort_order_gbk,
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
+    NULL,		/* sort_order_big*/
     "",
     "",
     1,			/* strxfrm_multiply */
@@ -9992,6 +9993,7 @@ CHARSET_INFO my_charset_gbk_bin=
     sort_order_gbk,
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
+    NULL,		/* sort_order_big*/
     "",
     "",
     1,			/* strxfrm_multiply */

strings/ctype-latin1.c

@@ -416,6 +416,7 @@ CHARSET_INFO my_charset_latin1=
     to_lower_latin1,
     to_upper_latin1,
     sort_order_latin1,
+    NULL,		/* sort_order_big*/
     cs_to_uni,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     "","",
@@ -693,6 +694,7 @@ CHARSET_INFO my_charset_latin1_german2_ci=
   to_lower_latin1,
   to_upper_latin1,
   sort_order_latin1_de,
+  NULL,					/* sort_order_big*/
   cs_to_uni,				/* tab_to_uni   */
   NULL,					/* tab_from_uni */
   "","",
@@ -717,6 +719,7 @@ CHARSET_INFO my_charset_latin1_bin=
   to_lower_latin1,
   to_upper_latin1,
   sort_order_latin1_de,
+  NULL,					/* sort_order_big*/
   cs_to_uni,				/* tab_to_uni   */
   NULL,					/* tab_from_uni */
   "",

strings/ctype-sjis.c

@@ -4585,6 +4585,7 @@ CHARSET_INFO my_charset_sjis_japanese_ci=
     sort_order_sjis,
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
+    NULL,		/* sort_order_big*/
     "",
     "",
     1,			/* strxfrm_multiply */
@@ -4609,6 +4610,7 @@ CHARSET_INFO my_charset_sjis_bin=
     sort_order_sjis,
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
+    NULL,		/* sort_order_big*/
     "",
     "",
     1,			/* strxfrm_multiply */

strings/ctype-tis620.c

@@ -957,6 +957,7 @@ CHARSET_INFO my_charset_tis620_thai_ci=
     sort_order_tis620,
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
+    NULL,		/* sort_order_big*/
     "",
     "",
     4,			/* strxfrm_multiply */
@@ -981,6 +982,7 @@ CHARSET_INFO my_charset_tis620_bin=
     sort_order_tis620,
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
+    NULL,		/* sort_order_big*/
     "",
     "",
     1,			/* strxfrm_multiply */

strings/ctype-uca.c

@@ -6436,7 +6436,7 @@ uint16 page0FFdata[]= { /* FF00 (3 weights per char) */
 0x0DC5,0x0000,0x0000, 0x0DC6,0x0000,0x0000, 0xFBC1,0xFFFE,0x0000,
 0xFBC1,0xFFFF,0x0000 };
 
-uchar ucal[256]={
+uchar uca_length[256]={
 4,3,3,4,3,3,3,3,0,3,3,3,3,3,3,3,
 3,3,3,3,3,2,3,3,3,3,0,0,0,3,3,3,
 5,5,4,3,5,2,3,3,2,2,5,3,0,0,3,3,
@@ -6454,7 +6454,7 @@ uchar ucal[256]={
 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
 0,0,0,0,0,0,0,0,0,3,3,4,3,9,3,3
 };
-uint16 *ucaw[256]={
+uint16 *uca_weight[256]={
 page000data,page001data,page002data,page003data,
 page004data,page005data,page006data,page007data,
 NULL       ,page009data,page00Adata,page00Bdata,
@@ -6533,6 +6533,8 @@ typedef struct my_uca_scanner_st
   const uint16 *wbeg;	/* Beginning of the current weight string */
   const uchar  *sbeg;	/* Beginning of the input string          */
   const uchar  *send;	/* End of the input string                */
+  uchar *uca_length;
+  uint16 **uca_weight;
   uint16 implicit[2];
   int page;
   int code;
@@ -6564,6 +6566,8 @@ static void my_uca_scanner_init(my_uca_scanner *scanner,
   scanner->sbeg= str;
   scanner->send= str + length - 2;
   scanner->wbeg= nochar; 
+  scanner->uca_length= cs->sort_order;
+  scanner->uca_weight= cs->sort_order_big;
 }
 
 
@@ -6620,6 +6624,9 @@ static int my_uca_scanner_next(my_uca_scanner *scanner)
   
   do 
   {
+    uint16 **ucaw= scanner->uca_weight;
+    uchar *ucal= scanner->uca_length;
+    
     if (scanner->sbeg > scanner->send)
       return -1;
     
@@ -6866,10 +6873,12 @@ static int my_strnxfrm_uca(CHARSET_INFO *cs,
   This fact allows us to use memcmp() safely, on both
   little-endian and big-endian machines.
 */
-static int my_uca_charcmp(my_wc_t wc1, my_wc_t wc2)
+static int my_uca_charcmp(CHARSET_INFO *cs, my_wc_t wc1, my_wc_t wc2)
 {
   size_t page1= wc1 >> MY_UCA_PSHIFT;
   size_t page2= wc2 >> MY_UCA_PSHIFT;
+  uchar *ucal= cs->sort_order;
+  uint16 **ucaw= cs->sort_order_big;
   size_t length1= ucal[page1];
   size_t length2= ucal[page2];
   uint16 *weight1= ucaw[page1] + (wc1 & MY_UCA_CMASK) * ucal[page1];
@@ -6943,7 +6952,7 @@ int my_wildcmp_uca(CHARSET_INFO *cs,
       }
       else
       {
-        if (my_uca_charcmp(s_wc,w_wc))
+        if (my_uca_charcmp(cs,s_wc,w_wc))
           return 1;
       }
       if (wildstr == wildend)
@@ -7006,7 +7015,7 @@ int my_wildcmp_uca(CHARSET_INFO *cs,
 			   (const uchar*)str_end)) <= 0)
             return 1;
           
-          if (!my_uca_charcmp(s_wc,w_wc))
+          if (!my_uca_charcmp(cs,s_wc,w_wc))
             break;
           str+= scan;
         }
@@ -7039,4 +7048,30 @@ MY_COLLATION_HANDLER my_collation_ucs2_uca_handler =
     my_hash_sort_uca
 };
 
+CHARSET_INFO my_charset_ucs2_general_uca=
+{
+    45,0,0,		/* number       */
+    MY_CS_COMPILED|MY_CS_STRNXFRM|MY_CS_UNICODE|MY_CS_NONTEXT,
+    "ucs2",		/* cs name    */
+    "ucs2_general_uca",	/* name         */
+    "",			/* comment      */
+    NULL,		/* ctype        */
+    NULL,		/* to_lower     */
+    NULL,		/* to_upper     */
+    uca_length,		/* sort_order   */
+    uca_weight,		/* sort_order_big*/
+    NULL,		/* tab_to_uni   */
+    NULL,		/* tab_from_uni */
+    "",
+    "",
+    8,			/* strxfrm_multiply */
+    2,			/* mbminlen     */
+    2,			/* mbmaxlen     */
+    9,			/* min_sort_char */
+    0xFFFF,		/* max_sort_char */
+    &my_charset_ucs2_handler,
+    &my_collation_ucs2_uca_handler
+};
+
+
 #endif

strings/ctype-ucs2.c

@@ -1397,7 +1397,7 @@ static MY_COLLATION_HANDLER my_collation_ucs2_bin_handler =
 };
 
 
-static MY_CHARSET_HANDLER my_charset_ucs2_handler=
+MY_CHARSET_HANDLER my_charset_ucs2_handler=
 {
     my_ismbchar_ucs2,	/* ismbchar     */
     my_mbcharlen_ucs2,	/* mbcharlen    */
@@ -1437,6 +1437,7 @@ CHARSET_INFO my_charset_ucs2_general_ci=
     to_upper_ucs2,	/* sort_order   */
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
+    NULL,		/* sort_order_big*/
     "",
     "",
     1,			/* strxfrm_multiply */
@@ -1448,30 +1449,6 @@ CHARSET_INFO my_charset_ucs2_general_ci=
     &my_collation_ucs2_general_ci_handler
 };
 
-CHARSET_INFO my_charset_ucs2_general_uca=
-{
-    45,0,0,		/* number       */
-    MY_CS_COMPILED|MY_CS_STRNXFRM|MY_CS_UNICODE|MY_CS_NONTEXT,
-    "ucs2",		/* cs name    */
-    "ucs2_general_uca",	/* name         */
-    "",			/* comment      */
-    ctype_ucs2,		/* ctype        */
-    to_lower_ucs2,	/* to_lower     */
-    to_upper_ucs2,	/* to_upper     */
-    to_upper_ucs2,	/* sort_order   */
-    NULL,		/* tab_to_uni   */
-    NULL,		/* tab_from_uni */
-    "",
-    "",
-    8,			/* strxfrm_multiply */
-    2,			/* mbminlen     */
-    2,			/* mbmaxlen     */
-    9,			/* min_sort_char */
-    0xFFFF,		/* max_sort_char */
-    &my_charset_ucs2_handler,
-    &my_collation_ucs2_uca_handler
-};
-
 CHARSET_INFO my_charset_ucs2_bin=
 {
     90,0,0,		/* number       */
@@ -1483,6 +1460,7 @@ CHARSET_INFO my_charset_ucs2_bin=
     to_lower_ucs2,	/* to_lower     */
     to_upper_ucs2,	/* to_upper     */
     to_upper_ucs2,	/* sort_order   */
+    NULL,		/* sort_order_big*/
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     "",

strings/ctype-ujis.c

@@ -8474,7 +8474,9 @@ CHARSET_INFO my_charset_ujis_japanese_ci=
     sort_order_ujis,
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
-    "","",
+    NULL,		/* sort_order_big*/
+    "",
+    "",
     1,			/* strxfrm_multiply */
     1,			/* mbminlen     */
     3,			/* mbmaxlen     */
@@ -8498,7 +8500,9 @@ CHARSET_INFO my_charset_ujis_bin=
     sort_order_ujis,
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
-    "","",
+    NULL,		/* sort_order_big*/
+    "",
+    "",
     1,			/* strxfrm_multiply */
     1,			/* mbminlen     */
     3,			/* mbmaxlen     */

strings/ctype-utf8.c

@@ -2090,6 +2090,7 @@ CHARSET_INFO my_charset_utf8_general_ci=
     to_upper_utf8,	/* sort_order   */
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
+    NULL,		/* sort_order_big*/
     "",
     "",
     1,			/* strxfrm_multiply */
@@ -2115,6 +2116,7 @@ CHARSET_INFO my_charset_utf8_bin=
     to_upper_utf8,	/* sort_order   */
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
+    NULL,		/* sort_order_big*/
     "",
     "",
     1,			/* strxfrm_multiply */

strings/ctype-win1250ch.c

@@ -627,6 +627,7 @@ CHARSET_INFO my_charset_cp1250_czech_ci =
   to_lower_win1250ch,
   to_upper_win1250ch,
   sort_order_win1250ch,
+  NULL,				/* sort_order_big*/
   tab_cp1250_uni,		/* tab_to_uni   */
   idx_uni_cp1250,		/* tab_from_uni */
   "","",