The Source of Magic
I was curious about what was going on behind the scenes. How many SQL SELECTs
were being issued? One way to find out is to turn on logging. You do this simply
by adding these statements in the top of your program:
import org.exolab.castor.util.Logger;
import java.io.*;
. . .
PrintWriter writer = new Logger(System.out).setPrefix("test");
jdo.setLogWriter(writer);
If we run the program once more we'll see this:
[test] SQL for creating hansen.playground.Movie: INSERT INTO movie ...
(more lines cut out here)
[test] SELECT actor.id,actor.name,actor_movie.id_movie FROM actor
LEFT OUTER JOIN actor_movie
ON actor.id=actor_movie.id_actor WHERE (actor.name = ?)
[test] Castor: Loading hansen.playground.Actor (23)
[test] Castor: Loading hansen.playground.Movie (12)
[test] Castor: Loading hansen.playground.Media (1)
[test] Castor: Loading hansen.playground.Actor (22)
[test] Castor: Loading hansen.playground.Movie (11)
[test] Castor: Loading hansen.playground.Media (2)
[test] Castor: Loading hansen.playground.Movie (13)
[test] Castor: Loading hansen.playground.Actor (21)
The first lines that are listed are all the SQL statements that Castor will
use when reading, creating, updating, and deleting objects from the
tables. I've only shown part of the first create statement above. Following this
is the (first) SELECT statement that is the result of the query for Keanu
Reeves. Finally there's a line each time an object is created. The keys are
listed in parenthesis. Here's a "translation":
[test] Castor: Loading hansen.playground.Actor (Ian Holm)
[test] Castor: Loading hansen.playground.Movie (Lord of the Rings)
[test] Castor: Loading hansen.playground.Media (DVD)
[test] Castor: Loading hansen.playground.Actor (Laurence Fishburne)
[test] Castor: Loading hansen.playground.Movie (The Matrix)
[test] Castor: Loading hansen.playground.Media (VHS)
[test] Castor: Loading hansen.playground.Movie (Chain Reaction)
[test] Castor: Loading hansen.playground.Actor (Keanu Reeves)
The lines seems to come in opposite order, but this might simply reflect the
way Castor is coded, e.g. by using recursive methods. You might also wonder why
all objects are loaded, even the movie where Keanu Reeves doesn't act. The
reason is that if you use the DVD object to get all movies in DVD format then
you'll need "Lord of the Rings". And when you get this movie you'll also need
Ian Holm. And so on.
Getting more information
If you're familiar with SQL then you'll see that one SELECT is not enough to
get all the objects we need. Let's therefore use a tool like P6Spy. It's a utility that sits between a JDBC
program and the database and watches and reports on the "traffic". To use it you
replace the name of the JDBC driver in Castor's database.xml file with
the P6Spy driver, and tell P6Spy to use the real JDBC driver. I won't go into
details about installation of P6Spy. Suffice to say, it's very simple to do.
What P6Spy tells you (in its spy.log file) when you run the
ReadActor program is this:
SELECT actor.id,actor.name,actor_movie.id_movie
FROM actor
LEFT OUTER JOIN actor_movie ON actor.id=actor_movie.id_actor
WHERE (actor.name = 'Keanu Reeves')
SELECT movie.title,movie.id_media,actor_movie.id_actor
FROM movie
LEFT OUTER JOIN actor_movie ON movie.id=actor_movie.id_movie
WHERE movie.id='11'
SELECT media.type,movie.id
FROM media
LEFT OUTER JOIN movie ON media.id=movie.id_media
WHERE media.id='1'
(5 more follows)
commit
The first select is the same as we've seen before. The others are used to get
data for all the objects in the model (3 actors, 3 films, 2 medias).
This brings up a question about performance. What if you don't have a need
for all the objects? Couldn't we postpone loading of objects until they're
actually needed? The answer is yes, there's a mechanism called "lazy loading",
which is one of the topics we'll look into in my next article. So stay
tuned on JavaBoutique.com!
New on the Java Boutique:
New Review:
Time Management Made Easy with the Quartz Enterprise Job Scheduler
Why not just use the Java timer API? This open source scheduling
API boasts simplicity, ease-of-integration, a well-rounded feature
set, and it's free!
New Applet:
Reverse Complement
Reverse Complement is a simple applet that converts DNA or RNA
sequences into three useful formats.
Elsewhere on internet.com:
WebDeveloper Java
Lots of Java information on webdeveloper.com
WDVL Java
Thorough Java resource at the Web Developer's Virtual Library.
ScriptSearch Java
Hundreds of free Java code files to download.
jGuru: Your View of the Java Universe
Customizable portal with online training, FAQs, regular news updates, and tutorials.
| 
