INF: SQL Server Truncation of a DBCS String (155723)

SUMMARY

If dual byte character set (DBCS) data is too long to fit into a char or varchar column, the data is truncated to fit into the column. If Microsoft SQL Server uses a DBCS code page and this truncation happens right in the middle of a double-byte character, the whole double-byte character is discarded.

The following scripts demonstrate this truncation. It is assumed that the code page that applies here is a DBCS code page.

CREATE TABLE test
(
  col1 char (10),
  col1 varchar (10)
)
GO

/* Let D represent a double character. */ 
/* Let L represent the leading byte.   */ 
/* Let T represent the trailing byte.  */ 
/* Let S represent a single character. */ 
/* Let s represent space (ASCII 20).   */ 

INSERT test VALUES ("SDDDDD", "SDDDDD")
GO
SELECT * FROM test
GO
				

col1       col2
---------- ----------
SLTLTLTLTs SLTLTLTLT
				

Because the truncated string is shorter than the maximum length, the char column that does not allow a null value and the char variable is padded with trailing blanks while the varchar column will not store trailing blanks.