Question

If I persist a windows-1252 string (e.g. something with “smart quotes”) from C# to a record into a SQLServer db, is the encoding preserved if I read it back out into another string variable?

Answer 1

Yes, no or maybe.

Each VARCHAR column in SQL Server has a “collation”, roughly equivalent to a codepage (character set) plus a sort order.

If the codepage of the collation can store all the characters in the string they should round-trip properly. If not, not.

If your collation is a variant of Latin1_General you should be able to store the whole windows-1252 character set.

For more, see here:

• Windows Collation Name:
  http://msdn.microsoft.com/en-us/library/ms188046.aspx
• Collation and International Terminology: http://msdn.microsoft.com/en-us/library/ms143726.aspx
• Char and Varchar: http://msdn.microsoft.com/en-us/library/ms176089.aspx

Note that if you haven’t set a collation on the field, it will use the database’s default collation, which is usually Latin1_General_CI_AS.

It is important to be aware of character set issues. If you need to store characters in one field which cannot all be fitted into the same codepage you need to use NVARCHAR. In an NVARCHAR field, all characters can be stored and the collation only controls the sort order and equality comparisions.