Bug#27345 Incorrect data returned when range-read from utf8_danish_ci indexes

Problem: like_range() returned wrong ranges for contractions (like 'ch' in Czech').
Fix: adding a special code to handle tricky cases:
- contraction head followed by a wild character
- full contraction
- contraction part followed by another contraction part,
  but they are not a contraction together.


mysql-test/r/ctype_uca.result:
  Adding test case
mysql-test/t/ctype_uca.test:
  Adding test case
strings/ctype-mb.c:
  Adding test case
strings/ctype-uca.c:
  Allocate additional 256 bytes for flags "is contraction part".
strings/ctype-ucs2.c:
  Adding test case

1 files changed, 10 insertions, 1 deletions

diff --git a/strings/ctype-uca.c b/strings/ctype-uca.c
index 4dbda0b9239..81fb9ee1970 100644
--- a/strings/ctype-uca.c
+++ b/strings/ctype-uca.c
@@ -7937,10 +7937,16 @@ static my_bool create_tailoring(CHARSET_INFO *cs, void *(*alloc)(uint))
   /* Now process contractions */
   if (ncontractions)
   {
-    uint size= 0x40*0x40*sizeof(uint16); /* 8K, for basic latin letter only */
+    /*
+      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 (rule[i].curr[1])
@@ -7966,6 +7972,9 @@ static my_bool create_tailoring(CHARSET_INFO *cs, void *(*alloc)(uint))
         
         /* 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;
       }
     }
   }