bk commit into 4.1 tree (marko:1.2462) BUG#12178

[Marko Mäkelä]

Below is the list of changes that have just been committed into a local
4.1 repository of marko. When marko 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
  1.2462 05/09/30 13:58:00 marko@stripped +1 -0
  InnoDB: On UPDATE, trim UCS2 columns correctly. (Bug #12178)

  sql/ha_innodb.cc
    1.209 05/09/30 13:57:52 marko@stripped +30 -10
    innobase_convert_and_store_changed_col(), calc_row_difference():
     Replace parameter "is_unsigned" with "prtype".
    innobase_convert_and_store_changed_col():
     When trimming spaces, note that UCS2 spaces are 0x0020, not 0x20.

# This is a BitKeeper patch.  What follows are the unified diffs for the
# set of deltas contained in the patch.  The rest of the patch, the part
# that BitKeeper cares about, is below these diffs.
# User:	marko
# Host:	hundin.mysql.fi
# Root:	/home/marko/mysql-4.1

--- 1.208/sql/ha_innodb.cc	2005-09-28 14:14:40 +03:00
+++ 1.209/sql/ha_innodb.cc	2005-09-30 13:57:52 +03:00
@@ -2614,7 +2614,7 @@
 	mysql_byte*	data,	/* in: column data to store */
 	ulint		len,	/* in: data len */
 	ulint		col_type,/* in: data type in InnoDB type numbers */
-	ulint		is_unsigned)/* in: != 0 if an unsigned integer type */
+	ulint		prtype)	/* InnoDB precise data type and flags */
 {
 	uint	i;
 
@@ -2622,10 +2622,31 @@
 		data = NULL;
 	} else if (col_type == DATA_VARCHAR || col_type == DATA_BINARY
 		   || col_type == DATA_VARMYSQL) {
-	        /* Remove trailing spaces */
-        	while (len > 0 && data[len - 1] == ' ') {
-	                len--;
-	        }
+		/* Remove trailing spaces. */
+
+		/* Handle UCS2 strings differently.  As no new
+		collations will be introduced in 4.1, we hardcode the
+		charset-collation codes here.  In 5.0, the logic will
+		be based on mbminlen. */
+		ulint	cset	= dtype_get_charset_coll_noninline(prtype);
+		if (cset == 35/*ucs2_general_ci*/
+				|| cset == 90/*ucs2_bin*/
+				|| (cset >= 128/*ucs2_unicode_ci*/
+				&& cset <= 144/*ucs2_persian_ci*/)) {
+			/* space=0x0020 */
+			/* Trim "half-chars", just in case. */
+			len &= ~1;
+
+			while (len && data[len - 2] == 0x00
+					&& data[len - 1] == 0x20) {
+				len -= 2;
+			}
+		} else {
+			/* space=0x20 */
+			while (len && data[len - 1] == 0x20) {
+				len--;
+			}
+		}
 	} else if (col_type == DATA_INT) {
 		/* Store integer data in InnoDB in a big-endian
 		format, sign bit negated, if signed */
@@ -2634,7 +2655,7 @@
 			buf[len - 1 - i] = data[i];
 		}
 
-		if (!is_unsigned) {
+		if (!(prtype & DATA_UNSIGNED)) {
 			buf[0] = buf[0] ^ 128;
 		}
 
@@ -2677,7 +2698,7 @@
         byte*	        buf;
 	upd_field_t*	ufield;
 	ulint		col_type;
-	ulint		is_unsigned;
+	ulint		prtype;
 	ulint		n_changed = 0;
 	uint		i;
 
@@ -2702,8 +2723,7 @@
 		n_len = field->pack_length();
 
 		col_type = prebuilt->table->cols[i].type.mtype;
-		is_unsigned = prebuilt->table->cols[i].type.prtype &
-								DATA_UNSIGNED;
+		prtype = prebuilt->table->cols[i].type.prtype;
 		switch (col_type) {
 
 		case DATA_BLOB:
@@ -2743,7 +2763,7 @@
                           innobase_convert_and_store_changed_col(ufield,
 					  (mysql_byte*)buf,
 					  (mysql_byte*)n_ptr, n_len, col_type,
-						is_unsigned);
+						prtype);
 			ufield->exp = NULL;
 			ufield->field_no = prebuilt->table->cols[i].clust_pos;
 			n_changed++;