UTF-16 Sample

Open three terminal windows on UNIX, or three StreamBase Command Prompts on Windows. In each window, navigate to the directory where the sample is installed, or to your workspace copy of the sample, as described above.

In window 1, launch an instance of StreamBase Server on utf16-search.sbapp:

sbd -f sbd.sbconf utf16-search.sbapp

In window 2, enter the following command to dequeue tuples from the keywords-found output stream:

sbc dequeue keywords_found

Note: Initially, no output is displayed in the dequeue window. We must complete the next few steps before results are dequeued. The results will be messages confirming that matches were found in the StreamBase application between the Hebrew keywords and the Hebrew document's text.

In window 3, set the CLASSPATH, and double-check the other environment settings. Make sure the STREAMBASE_HOME environment variable is set to your StreamBase installation directory. These examples also assume a supported JDK is on your PATH.

Set the CLASSPATH to include this sample's JAR file as well as the standard StreamBase Client library.

On UNIX:

export CLASSPATH=$STREAMBASE_HOME/sample/utf-16/utf16.jar:
    $STREAMBASE_HOME/lib/sbclient.jar:$CLASSPATH

On Windows:

When you use a StreamBase Command Prompt, the CLASSPATH variable is pre-set to include the path to sbclient.jar for the current release. Thus, you only need to add utf16.jar to the CLASSPATH:

set CLASSPATH=%STREAMBASE_HOME%sample\utf-16\utf16.jar;%CLASSPATH%

In window 3, run the sample's Java client, which loads Unicode Hebrew keywords into the application's Query Table.

On UNIX and Windows:

To load UTF-16BE (big-endian) keywords into the running StreamBase application:

java com.streambase.UTF16Client sb://localhost:10000 
    bigend_keywords keywords_utf_16_big_endian.txt big

To load UTF-16LE (little-endian) keywords into the running StreamBase application:

java com.streambase.UTF16Client sb://localhost:10000
    littleend_keywords keywords_utf_16_littleendian.txt little

These commands display the length of the keywords being loaded into the running StreamBase application. For example:

Line len: 22
Line len: 16
Line len: 12
Line len: 10
Line len: 14
Line len: 6
Line len: 14
Line len: 16

A total of eight keywords were loaded.

Now that the Hebrew keywords have been loaded, you can run the same Java client with different command line parameters, this time to load a Hebrew document.

On UNIX and Windows:

To load the UTF-16BE (big-endian) document into the StreamBase application:

java com.streambase.UTF16Client sb://localhost:10000
    bigend_text doc_utf_16_big_endian.txt big

To load the UTF-16LE (little-endian) document into the StreamBase application:

java com.streambase.UTF16Client sb://localhost:10000
    littleend_text doc_utf_16_little_endian.txt little

In the running StreamBase application, a custom Java function (utf16) is used to find matches between the Hebrew keywords and the just-loaded Hebrew document.

Now look at window 2, the one running sbc dequeue. Look for evidence that the Hebrew characters were recognized properly, by virtue of the matches between the Hebrew keywords and the just-loaded Hebrew document. For example:

sbc dequeue keywords_found
1,5
2,7
1,1
2,6
1,1
1,4
1,2
1,1
5,1
1,3

In the dequeue results, you are seeing the count,id fields that comprise each tuple on the keywords_found Output Stream.

In window 3, type the following command to terminate the server and dequeuer:

sbadmin shutdown