bk commit into 5.1 tree (svoj:1.2528) BUG#29464

[Sergey Vojtovich]

Below is the list of changes that have just been committed into a local
5.1 repository of svoj. When svoj does a push these changes will
be propagated to the main repository and, within 24 hours after the
push, to the public repository.
For information on how to access the public repository
see http://dev.mysql.com/doc/mysql/en/installing-source-tree.html

ChangeSet@stripped, 2007-07-05 15:17:06+05:00, svoj@stripped +3 -0
  BUG#29464 - load data infile into table with big5 chinese fulltext index
              hangs 100% cpu
  
  Fulltext parser may fall into infinite loop when it gets illegal
  multibyte sequence (or a sequence that doesn't have mapping to unicode).
  
  Affects 5.1 only.

  mysql-test/r/fulltext.result@stripped, 2007-07-05 15:17:04+05:00, svoj@stripped +4 -0
    A test case for BUG#29464.

  mysql-test/t/fulltext.test@stripped, 2007-07-05 15:17:04+05:00, svoj@stripped +9 -0
    A test case for BUG#29464.

  storage/myisam/ft_parser.c@stripped, 2007-07-05 15:17:04+05:00, svoj@stripped +10 -6
    ctype() may return negative value, which was stored in unsigned
    variable.
    
    Also ctype() may return negative length for correct multibyte
    sequence that doesn't have a mapping to unicode. These characters
    are skipped correctly with this patch.

diff -Nrup a/mysql-test/r/fulltext.result b/mysql-test/r/fulltext.result
--- a/mysql-test/r/fulltext.result	2007-04-18 13:27:01 +05:00
+++ b/mysql-test/r/fulltext.result	2007-07-05 15:17:04 +05:00
@@ -476,3 +476,7 @@ ALTER TABLE t1 DISABLE KEYS;
 SELECT * FROM t1 WHERE MATCH(a) AGAINST('test');
 ERROR HY000: Can't find FULLTEXT index matching the column list
 DROP TABLE t1;
+CREATE TABLE t1(a VARCHAR(2) CHARACTER SET big5 COLLATE big5_chinese_ci,
+FULLTEXT(a));
+INSERT INTO t1 VALUES(0xA3C2);
+DROP TABLE t1;
diff -Nrup a/mysql-test/t/fulltext.test b/mysql-test/t/fulltext.test
--- a/mysql-test/t/fulltext.test	2007-04-18 13:27:01 +05:00
+++ b/mysql-test/t/fulltext.test	2007-07-05 15:17:04 +05:00
@@ -399,4 +399,13 @@ ALTER TABLE t1 DISABLE KEYS;
 SELECT * FROM t1 WHERE MATCH(a) AGAINST('test');
 DROP TABLE t1;
 
+#
+# BUG#29464 - load data infile into table with big5 chinese fulltext index
+#             hangs 100% cpu
+#
+CREATE TABLE t1(a VARCHAR(2) CHARACTER SET big5 COLLATE big5_chinese_ci,
+FULLTEXT(a));
+INSERT INTO t1 VALUES(0xA3C2);
+DROP TABLE t1;
+
 # End of 4.1 tests
diff -Nrup a/storage/myisam/ft_parser.c b/storage/myisam/ft_parser.c
--- a/storage/myisam/ft_parser.c	2007-07-02 12:35:52 +05:00
+++ b/storage/myisam/ft_parser.c	2007-07-05 15:17:04 +05:00
@@ -111,7 +111,8 @@ uchar ft_get_word(CHARSET_INFO *cs, ucha
 {
   uchar *doc=*start;
   int ctype;
-  uint mwc, length, mbl;
+  uint mwc, length;
+  int mbl;
 
   param->yesno=(FTB_YES==' ') ? 1 : (param->quot != 0);
   param->weight_adjust= param->wasign= 0;
@@ -119,7 +120,7 @@ uchar ft_get_word(CHARSET_INFO *cs, ucha
 
   while (doc<end)
   {
-    for (; doc < end; doc+= (mbl > 0 ? mbl : 1))
+    for (; doc < end; doc+= (mbl > 0 ? mbl : (mbl < 0 ? -mbl : 1)))
     {
       mbl= cs->cset->ctype(cs, &ctype, (uchar*)doc, (uchar*)end);
       if (true_word_char(ctype, *doc))
@@ -157,7 +158,8 @@ uchar ft_get_word(CHARSET_INFO *cs, ucha
     }
 
     mwc=length=0;
-    for (word->pos= doc; doc < end; length++, doc+= (mbl > 0 ? mbl : 1))
+    for (word->pos= doc; doc < end; length++,
+         doc+= (mbl > 0 ? mbl : (mbl < 0 ? -mbl : 1)))
     {
       mbl= cs->cset->ctype(cs, &ctype, (uchar*)doc, (uchar*)end);
       if (true_word_char(ctype, *doc))
@@ -200,13 +202,14 @@ uchar ft_simple_get_word(CHARSET_INFO *c
                          FT_WORD *word, my_bool skip_stopwords)
 {
   uchar *doc= *start;
-  uint mwc, length, mbl;
+  uint mwc, length;
+  int mbl;
   int ctype;
   DBUG_ENTER("ft_simple_get_word");
 
   do
   {
-    for (;; doc+= (mbl > 0 ? mbl : 1))
+    for (;; doc+= (mbl > 0 ? mbl : (mbl < 0 ? -mbl : 1)))
     {
       if (doc >= end)
         DBUG_RETURN(0);
@@ -216,7 +219,8 @@ uchar ft_simple_get_word(CHARSET_INFO *c
     }
 
     mwc= length= 0;
-    for (word->pos= doc; doc < end; length++, doc+= (mbl > 0 ? mbl : 1))
+    for (word->pos= doc; doc < end; length++,
+         doc+= (mbl > 0 ? mbl : (mbl < 0 ? -mbl : 1)))
     {
       mbl= cs->cset->ctype(cs, &ctype, (uchar*)doc, (uchar*)end);
       if (true_word_char(ctype, *doc))