CODEWITHSUNDEEP
javacharacter-encodingsizevarcharstring-length
Sandip Bhoi
Sandip Bhoi

Reputation: 41

Java string .length() X does not fit into DB2 varchar(X)

I observed one issue that, I am trying to save a record in DB2 database with fields having length check in Java code. I have kept length check exactly equal to database varchar limit. And trying to save but getting SQL Exception DB2 SQL Error: SQLCODE=-302, SQLSTATE=22001, SQLERRMC=null, DRIVER=3.57.82;

Then I reduced the length (truncated) the length smaller than the database size. Truncated to substring (0, 900) apprx for varchar(1000).

Please let me what could be the reason of the same. Is it related to character encoding? How it needs to be handled?

What is default character encoding applied to String (input from request parameter of a text area field) and corresponding no. of bytes?

Upvotes: 4

Views: 3614

Answers (1)

Joni
Joni

Reputation: 111269

DB2 counts string length in bytes, not characters. The max length of a string you can store can therefore be a lot shorter than the size given for varchar.

Unfortunately the only way to truncate a string to a number of bytes is to encode it as bytes, truncating, and reconstructing the string. From what you say it sounds like a variable length encoding such as UTF-8 is being used. The difficult part is not producing an invalid character at the end, and the way to do that is using the NIO charset API:

import java.nio.*;
...
CharBuffer in = CharBuffer.wrap(stringToTruncate);
ByteBuffer out = ByteBuffer.allocate(maxLength);
Charset db2charset = Charset.forName("UTF-8");
CharsetEncoder db2encoder = db2charset.newEncoder();
db2encoder.encode(in, out, true);
out.flip();
return db2charset.decode(out).toString();

Upvotes: 5

Related Questions