Optimization to use less memory.

18 files changed, 98 insertions, 92 deletions

diff --git a/include/m_ctype.h b/include/m_ctype.h
index 87b45bd4954..002b77b5310 100644
--- a/include/m_ctype.h
+++ b/include/m_ctype.h
@@ -208,8 +208,8 @@ typedef struct charset_info_st
   uint16   **sort_order_big;
   uint16      *tab_to_uni;
   MY_UNI_IDX  *tab_from_uni;
-  uchar state_map[256];
-  uchar ident_map[256];
+  uchar     *state_map;
+  uchar     *ident_map;
   uint      strxfrm_multiply;
   uint      mbminlen;
   uint      mbmaxlen;
diff --git a/mysys/charset.c b/mysys/charset.c
index a9c733e25cf..d2d71689d7b 100644
--- a/mysys/charset.c
+++ b/mysys/charset.c
@@ -406,12 +406,21 @@ static void set_max_sort_char(CHARSET_INFO *cs)
 }
 
 
-static void init_state_maps(CHARSET_INFO *cs)
+static my_bool init_state_maps(CHARSET_INFO *cs)
 {
   uint i;
-  uchar *state_map= cs->state_map;
-  uchar *ident_map= cs->ident_map;
+  uchar *state_map;
+  uchar *ident_map;
 
+  if (!(cs->state_map= (uchar*) my_once_alloc(256, MYF(MY_WME))))
+    return 1;
+    
+  if (!(cs->ident_map= (uchar*) my_once_alloc(256, MYF(MY_WME))))
+    return 1;
+
+  state_map= cs->state_map;
+  ident_map= cs->ident_map;
+  
   /* Fill state_map with states to get a faster parser */
   for (i=0; i < 256 ; i++)
   {
@@ -458,6 +467,7 @@ static void init_state_maps(CHARSET_INFO *cs)
   state_map[(uchar)'x']= state_map[(uchar)'X']= (uchar) MY_LEX_IDENT_OR_HEX;
   state_map[(uchar)'b']= state_map[(uchar)'b']= (uchar) MY_LEX_IDENT_OR_BIN;
   state_map[(uchar)'n']= state_map[(uchar)'N']= (uchar) MY_LEX_IDENT_OR_NCHAR;
+  return 0;
 }
 
 
@@ -582,7 +592,8 @@ static int simple_cs_copy_data(CHARSET_INFO *to, CHARSET_INFO *from)
 					     MY_CS_CTYPE_TABLE_SIZE,
 					     MYF(MY_WME))))
       goto err;
-    init_state_maps(to);
+    if (init_state_maps(to))
+      goto err;
   }
   if (from->to_lower)
     if (!(to->to_lower= (uchar*) my_once_memdup((char*) from->to_lower,
@@ -601,6 +612,8 @@ static int simple_cs_copy_data(CHARSET_INFO *to, CHARSET_INFO *from)
 						  MY_CS_SORT_ORDER_TABLE_SIZE,
 						  MYF(MY_WME))))
       goto err;
+
+    
     set_max_sort_char(to);
   }
   if (from->tab_to_uni)
@@ -1108,7 +1121,8 @@ static my_bool init_available_charsets(myf myflags)
       {
         set_max_sort_char(*cs);
         if (cs[0]->ctype)
-          init_state_maps(*cs);
+          if (init_state_maps(*cs))
+            *cs= NULL;
       }
     }
     
diff --git a/sql/sql_lex.cc b/sql/sql_lex.cc
index f98a6b43846..b6cb61fe10e 100644
--- a/sql/sql_lex.cc
+++ b/sql/sql_lex.cc
@@ -42,13 +42,6 @@ pthread_key(LEX*,THR_LEX);
 #define TOCK_NAME_LENGTH 24
 
 /*
-  Map to default keyword characters.  This is used to test if an identifer
-  is 'simple', in which case we don't have to do any character set conversions
-  on it
-*/
-uchar *bin_ident_map= my_charset_bin.ident_map;
-
-/*
   The following data is based on the latin1 character set, and is only
   used when comparing keywords
 */
@@ -566,13 +559,9 @@ int yylex(void *arg, void *yythd)
       else
 #endif
       {
-	result_state= bin_ident_map[c] ? IDENT : IDENT_QUOTED;
-        while (ident_map[c=yyGet()])
-	{
-	  /* If not simple character, mark that we must convert it */
-	  if (!bin_ident_map[c])
-	    result_state= IDENT_QUOTED;
-	}
+        for (result_state= c; ident_map[c= yyGet()]; result_state|= c);
+        /* If there were non-ASCII characters, mark that we must convert */
+        result_state= result_state & 0x80 ? IDENT_QUOTED : IDENT;
       }
       length= (uint) (lex->ptr - lex->tok_start)-1;
       if (lex->ignore_space)
@@ -674,12 +663,11 @@ int yylex(void *arg, void *yythd)
       }
       else
 #endif
-        while (ident_map[c = yyGet()])
-	{
-	  /* If not simple character, mark that we must convert it */
-	  if (!bin_ident_map[c])
-	    result_state= IDENT_QUOTED;
-	}
+      {
+        for (result_state=0; ident_map[c= yyGet()]; result_state|= c);
+        /* If there were non-ASCII characters, mark that we must convert */
+        result_state= result_state & 0x80 ? IDENT_QUOTED : IDENT;
+      }
       if (c == '.' && ident_map[yyPeek()])
 	lex->next_state=MY_LEX_IDENT_SEP;// Next is '.'
 
@@ -953,13 +941,11 @@ int yylex(void *arg, void *yythd)
 	We should now be able to handle:
 	[(global | local | session) .]variable_name
       */
-      result_state= IDENT;
-      while (ident_map[c=yyGet()])
-      {
-	/* If not simple character, mark that we must convert it */
-	if (!bin_ident_map[c])
-	  result_state= IDENT_QUOTED;
-      }
+      
+      for (result_state= 0; ident_map[c= yyGet()]; result_state|= c);
+      /* If there were non-ASCII characters, mark that we must convert */
+      result_state= result_state & 0x80 ? IDENT_QUOTED : IDENT;
+      
       if (c == '.')
 	lex->next_state=MY_LEX_IDENT_SEP;
       length= (uint) (lex->ptr - lex->tok_start)-1;
diff --git a/strings/ctype-big5.c b/strings/ctype-big5.c
index 7a3c4503d74..fb72dec7385 100644
--- a/strings/ctype-big5.c
+++ b/strings/ctype-big5.c
@@ -6321,8 +6321,8 @@ CHARSET_INFO my_charset_big5_chinese_ci=
     NULL,		/* sort_order_big*/
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     1,			/* mbminlen   */
     2,			/* mbmaxlen   */
@@ -6348,8 +6348,8 @@ CHARSET_INFO my_charset_big5_bin=
     NULL,		/* sort_order_big*/
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     1,			/* mbminlen   */
     2,			/* mbmaxlen   */
diff --git a/strings/ctype-bin.c b/strings/ctype-bin.c
index 48323018cca..7b3164bf438 100644
--- a/strings/ctype-bin.c
+++ b/strings/ctype-bin.c
@@ -386,8 +386,8 @@ CHARSET_INFO my_charset_bin =
     NULL,			/* tab_to_uni    */
     NULL,			/* tab_from_uni  */
     NULL,			/* sort_order_big*/
-    "",
-    "",
+    NULL,			/* state_map    */
+    NULL,			/* ident_map    */
     1,				/* strxfrm_multiply */
     1,				/* mbminlen      */
     1,				/* mbmaxlen      */
diff --git a/strings/ctype-czech.c b/strings/ctype-czech.c
index dede737f361..3218fdee673 100644
--- a/strings/ctype-czech.c
+++ b/strings/ctype-czech.c
@@ -597,7 +597,8 @@ CHARSET_INFO my_charset_latin2_czech_ci =
     NULL,		/* sort_order_big*/
     tab_8859_2_uni,	/* tab_to_uni   */
     idx_uni_8859_2,	/* tab_from_uni */
-    "","",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     4,			/* strxfrm_multiply */
     1,			/* mbminlen   */
     1,			/* mbmaxlen  */
diff --git a/strings/ctype-euc_kr.c b/strings/ctype-euc_kr.c
index 2d4c68978a3..c387246b4c6 100644
--- a/strings/ctype-euc_kr.c
+++ b/strings/ctype-euc_kr.c
@@ -8689,8 +8689,8 @@ CHARSET_INFO my_charset_euckr_korean_ci=
     NULL,		/* sort_order_big*/
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     1,			/* mbminlen   */
     2,			/* mbmaxlen   */
@@ -8716,8 +8716,8 @@ CHARSET_INFO my_charset_euckr_bin=
     NULL,		/* sort_order_big*/
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     1,			/* mbminlen   */
     2,			/* mbmaxlen   */
diff --git a/strings/ctype-extra.c b/strings/ctype-extra.c
index 51a9531fbf5..baf1d319b00 100644
--- a/strings/ctype-extra.c
+++ b/strings/ctype-extra.c
@@ -24,20 +24,22 @@ CHARSET_INFO compiled_charsets[] = {
     NullS,		/* cs name    */
     NullS,		/* name       */
     NullS,		/* comment    */
-    NULL,
-    NULL,
-    NULL,
-    NULL,
+    NULL,		/* tailoring  */
+    NULL,		/* ctype      */
+    NULL,		/* to_lower   */
+    NULL,		/* to_upper   */
+    NULL,		/* sort_order */
     NULL,		/* sort_order_big*/
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
-    "","",
-    0,
-    0,
-    0,
-    0,
-    0,
-    NULL,
-    NULL
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
+    0,			/* strxfrm_mul  */
+    0,			/* mbminlen     */
+    0,			/* mbmaxlen     */
+    0,			/* min_sort_ord */
+    0,			/* max_sort_ord */
+    NULL,		/* cset handler */
+    NULL		/* coll handler */
   }
 };
diff --git a/strings/ctype-gb2312.c b/strings/ctype-gb2312.c
index 49ca736a3c2..fe1f72e7eda 100644
--- a/strings/ctype-gb2312.c
+++ b/strings/ctype-gb2312.c
@@ -5740,8 +5740,8 @@ CHARSET_INFO my_charset_gb2312_chinese_ci=
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     NULL,		/* sort_order_big*/
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     1,			/* mbminlen   */
     2,			/* mbmaxlen   */
@@ -5766,8 +5766,8 @@ CHARSET_INFO my_charset_gb2312_bin=
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     NULL,		/* sort_order_big*/
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     1,			/* mbminlen   */
     2,			/* mbmaxlen   */
diff --git a/strings/ctype-gbk.c b/strings/ctype-gbk.c
index 0273feb4c2c..8b659cb55f9 100644
--- a/strings/ctype-gbk.c
+++ b/strings/ctype-gbk.c
@@ -9970,8 +9970,8 @@ CHARSET_INFO my_charset_gbk_chinese_ci=
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     NULL,		/* sort_order_big*/
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     1,			/* mbminlen   */
     2,			/* mbmaxlen */
@@ -9996,8 +9996,8 @@ CHARSET_INFO my_charset_gbk_bin=
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     NULL,		/* sort_order_big*/
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     1,			/* mbminlen   */
     2,			/* mbmaxlen */
diff --git a/strings/ctype-latin1.c b/strings/ctype-latin1.c
index fe39303e2ac..03d4e71377b 100644
--- a/strings/ctype-latin1.c
+++ b/strings/ctype-latin1.c
@@ -420,7 +420,8 @@ CHARSET_INFO my_charset_latin1=
     NULL,		/* sort_order_big*/
     cs_to_uni,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
-    "","",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     1,			/* mbminlen   */
     1,			/* mbmaxlen  */
@@ -699,7 +700,8 @@ CHARSET_INFO my_charset_latin1_german2_ci=
   NULL,					/* sort_order_big*/
   cs_to_uni,				/* tab_to_uni   */
   NULL,					/* tab_from_uni */
-  "","",
+  NULL,					/* state_map    */
+  NULL,					/* ident_map    */
   2,					/* strxfrm_multiply */
   1,					/* mbminlen   */
   1,					/* mbmaxlen  */
@@ -725,8 +727,8 @@ CHARSET_INFO my_charset_latin1_bin=
   NULL,					/* sort_order_big*/
   cs_to_uni,				/* tab_to_uni   */
   NULL,					/* tab_from_uni */
-  "",
-  "",
+  NULL,					/* state_map    */
+  NULL,					/* ident_map    */
   1,					/* strxfrm_multiply */
   1,					/* mbminlen   */
   1,					/* mbmaxlen  */
diff --git a/strings/ctype-sjis.c b/strings/ctype-sjis.c
index 22c58360348..b4a131d3410 100644
--- a/strings/ctype-sjis.c
+++ b/strings/ctype-sjis.c
@@ -4587,8 +4587,8 @@ CHARSET_INFO my_charset_sjis_japanese_ci=
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     NULL,		/* sort_order_big*/
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     1,			/* mbminlen   */
     2,			/* mbmaxlen */
@@ -4613,8 +4613,8 @@ CHARSET_INFO my_charset_sjis_bin=
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     NULL,		/* sort_order_big*/
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     1,			/* mbminlen   */
     2,			/* mbmaxlen */
diff --git a/strings/ctype-tis620.c b/strings/ctype-tis620.c
index b2b1ab98352..79ac2079720 100644
--- a/strings/ctype-tis620.c
+++ b/strings/ctype-tis620.c
@@ -959,8 +959,8 @@ CHARSET_INFO my_charset_tis620_thai_ci=
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     NULL,		/* sort_order_big*/
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     4,			/* strxfrm_multiply */
     1,			/* mbminlen   */
     1,			/* mbmaxlen  */
@@ -985,8 +985,8 @@ CHARSET_INFO my_charset_tis620_bin=
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     NULL,		/* sort_order_big*/
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     1,			/* mbminlen   */
     1,			/* mbmaxlen  */
diff --git a/strings/ctype-uca.c b/strings/ctype-uca.c
index 75e2c06eec2..e6b68b8c9b2 100644
--- a/strings/ctype-uca.c
+++ b/strings/ctype-uca.c
@@ -7063,8 +7063,8 @@ CHARSET_INFO my_charset_ucs2_general_uca=
     uca_weight,		/* sort_order_big*/
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     8,			/* strxfrm_multiply */
     2,			/* mbminlen     */
     2,			/* mbmaxlen     */
diff --git a/strings/ctype-ucs2.c b/strings/ctype-ucs2.c
index d1ba63b8b84..bdf9b0f9252 100644
--- a/strings/ctype-ucs2.c
+++ b/strings/ctype-ucs2.c
@@ -1439,8 +1439,8 @@ CHARSET_INFO my_charset_ucs2_general_ci=
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     NULL,		/* sort_order_big*/
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     2,			/* mbminlen     */
     2,			/* mbmaxlen     */
@@ -1465,8 +1465,8 @@ CHARSET_INFO my_charset_ucs2_bin=
     NULL,		/* sort_order_big*/
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     2,			/* mbminlen     */
     2,			/* mbmaxlen     */
diff --git a/strings/ctype-ujis.c b/strings/ctype-ujis.c
index 668dc7beb8b..f28ea165f80 100644
--- a/strings/ctype-ujis.c
+++ b/strings/ctype-ujis.c
@@ -8476,8 +8476,8 @@ CHARSET_INFO my_charset_ujis_japanese_ci=
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     NULL,		/* sort_order_big*/
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     1,			/* mbminlen     */
     3,			/* mbmaxlen     */
@@ -8503,8 +8503,8 @@ CHARSET_INFO my_charset_ujis_bin=
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     NULL,		/* sort_order_big*/
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     1,			/* mbminlen     */
     3,			/* mbmaxlen     */
diff --git a/strings/ctype-utf8.c b/strings/ctype-utf8.c
index 2d0feb1c890..39e9260ffed 100644
--- a/strings/ctype-utf8.c
+++ b/strings/ctype-utf8.c
@@ -2098,8 +2098,8 @@ CHARSET_INFO my_charset_utf8_general_ci=
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     NULL,		/* sort_order_big*/
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     1,			/* mbminlen     */
     3,			/* mbmaxlen     */
@@ -2125,8 +2125,8 @@ CHARSET_INFO my_charset_utf8_bin=
     NULL,		/* tab_to_uni   */
     NULL,		/* tab_from_uni */
     NULL,		/* sort_order_big*/
-    "",
-    "",
+    NULL,		/* state_map    */
+    NULL,		/* ident_map    */
     1,			/* strxfrm_multiply */
     1,			/* mbminlen     */
     3,			/* mbmaxlen     */
diff --git a/strings/ctype-win1250ch.c b/strings/ctype-win1250ch.c
index bb287eb695e..670318a082e 100644
--- a/strings/ctype-win1250ch.c
+++ b/strings/ctype-win1250ch.c
@@ -631,7 +631,8 @@ CHARSET_INFO my_charset_cp1250_czech_ci =
   NULL,				/* sort_order_big*/
   tab_cp1250_uni,		/* tab_to_uni   */
   idx_uni_cp1250,		/* tab_from_uni */
-  "","",
+  NULL,				/* state_map    */
+  NULL,				/* ident_map    */
   2,				/* strxfrm_multiply */
   1,				/* mbminlen  */
   1,				/* mbmaxlen  */