Problem creating SQL commands in JDBC

Question

I want to send the following seemingly very simple SQL statements to the mysql server via JDBC inside a servlet:
querry1: SELECT * FROM japeng_math WHERE (kanji='userInput' OR kana='userInput')
querry2: SELECT * FROM majors WHERE major='majorcode'     
userInput and majorcode are variables of type String. They are assigned values before using in the SQL.
I coded the commands as follows:
querry1 = "select * from japeng_math where (kanji=" + "'" userInput "'" + "or kana=" + "'" userInput "'" + ")" ;
System.out.println("querry 1: " + querry1);
querry2 = "select * from majors where major=" + "'" + majorcode + "'";
System.out.println("querry 2: " + querry2);
I can't find what is going wrong here, but I'm not getting the desired strings for querry1 and querry2.
I found that error in querry string occurs only when two-byte Japanese characters (which is the commone case) are input.
Here is the output at console:
1. when a two-byte Japanese character was input in the form field
querry 1: select * from japeng_math where (kanji='���a'or kana='���a')
'uerry 2: select * from majors where major='geom
(There is an apostrophe in place of q of querry. And also major='geom. But the expected value should have been : major='geom')
2. when single-byte (hankaku) Japanese character was input
querry 1: select * from japeng_math where (kanji='����'or kana='����')
querry 2: select * from majors where major='geom'
(both OK)
3. when English was input, both OK.
I'm stuck. Please help. How can I make it work seamlessly, without worrying what type of charcters are input?
By the way, my database & server setting are in Shift-JIS. I've also converted the input string to Shift-JIS in the program
before using it.
Thank you very much.

Answer 1

>
I'm stuck. Please help. How can I make it work
seamlessly, without worrying what type of charcters
are input?Characters in java are in unicode.
When you display characters, such as when System.out is used, the unicode is converted into something else.
Similarily when you go through the JDBC interface they are converted to something else.
And depending on the source of the characters they might not be unicode to start with.
So the first step is to determine that the input strings actually contain the characters that you expect. The only way to do that is to print each character as an integer (hex format) and then look it up in a unicode chart to determine that it is the correct character. Until you do this you there is no point in trying anything else.
Once you determine this then the next step is to use a PreparedStatement to create the SQL. This might work. Or it might not. You should read the following thread including the links posted there ....
http://forum.java.sun.com/thread.jsp?forum=48&thread=407345