Prepared statements and streaming columns

setXXXStream requests stream data between the application and the database.

JDBC allows an IN parameter to be set to a Java input stream for passing in large amounts of data in smaller chunks. When the statement is executed, the JDBC driver makes repeated calls to this input stream, reading its contents and transmitting those contents as the parameter data.

Derby supports the three types of streams that JDBC 1.2 provides. These three streams are:
  • setBinaryStream

    for streams containing uninterpreted bytes

  • setAsciiStream

    for streams containing ASCII characters

  • setUnicodeStream

    for streams containing Unicode characters

JDBC 2.0 and JDBC 3.0 require that you specify the length of the stream and Derby enforces this requirement if your application runs on JDK 1.5 or earlier. If your application runs on JDK 1.6, then Derby exposes a JDBC 4.0 implementation, which lets you use the streaming interfaces without having to specify the stream length. The stream object passed to these three methods can be either a standard Java stream object or the user's own subclass that implements the standard java.io.InputStream interface.

According to the JDBC standard, streams can be stored only in columns of the data types shown in Streamable JDBC Data Types. Streams cannot be stored in columns of the other built-in data types or of user-defined data types.
Table 1. Streamable JDBC Data Types
Column Values Type Correspondent AsciiStream UnicodeStream BinaryStream
CLOB java.sql.Clob x x  
CHAR   x x  
VARCHAR   x x  
LONGVARCHAR   X X  
BINARY   x x x
BLOB java.sql.Blob x x x
VARBINARY   x x x
LONGVARBINARY   x x X
A large X indicates the preferred target data type for the type of stream. (See Mapping of java.sql.Types to SQL Types.)
Note: If the stream is stored in a column of a type other than LONG VARCHAR or LONG VARCHAR FOR BIT DATA, the entire stream must be able to fit into memory at one time. Streams stored in LONG VARCHAR and LONG VARCHAR FOR BIT DATA columns do not have this limitation.
The following example shows how a user can store a streamed java.io.File in a LONG VARCHAR column:
Statement s = conn.createStatement();
s.executeUpdate("CREATE TABLE atable (a INT, b LONG VARCHAR)");
conn.commit();
java.io.File file = new java.io.File("derby.txt");
int fileLength = (int) file.length();
// first, create an input stream
java.io.InputStream fin = new java.io.FileInputStream(file);
PreparedStatement ps = conn.prepareStatement(
    "INSERT INTO atable VALUES (?, ?)");
ps.setInt(1, 1);
// set the value of the input parameter to the input stream
ps.setAsciiStream(2, fin, fileLength);
ps.execute();
conn.commit();