Adding Unit Test in XCode 3.2.

Some Objective C links:

Learning Objective C
About Unit Tests

Adding Unit Test in XCode 3.2 is very easy, here is a simple note on this:

1. Create an objective C application, name it ‘MyApp’, build and run, you see an empty window pops up.
2. Right click ‘Targets’, Add/New Target, choose ‘Unit Test Bundle’, set Target name to ‘Unit Tests’
3. When Target “unit Tests” info pops up, under ‘General’, ‘Direct Dependencies’, add ‘MyApp’.
4. Right click ‘Unit Tests’ under ‘Targets’, build ‘Unit Tests’.
5. You see following error, because there is no test yet.

"Command /bin/sh failed with exit code 1"

6. Add a new group, name it ‘Tests’.
7. Right click, add file, choose ‘Objective C test case class’, name it ‘MyTest’, be sure target is: ‘Unit Tests’, uncheck ‘MyApp’ if it is checked. do the step 4 again, passed? if yes, proceed to next step.
8. Add a test to MyTest.m

-(void)testCase1;
{
	STAssertEquals(10, 20, @"equal?");
}

build ‘Unit Tests’ again, you will see an error reported:

testCase1] : '10' should be equal to '20': equal?

that means our test is working, set the statement above to:

STAssertEquals(10, 10, @"equal?");

build ‘Unit Tests’ again, passed!

As subject says:)

To fix slow connection problem with my realtek usb wireless card, I upgraded to 9.26.31rc6, the connection is perfect, but saa7134 module is not loaded by default any more, for that I have to manually compile and install it, since this is my first time installing a module, it’s quite basic, here are steps:

1)sudo apt-get install mercurial
hg clone http://linuxtv.org/hg/v4l-dvb

2) make -C /home/sqllyw/v4l-dvb/v4l
3) cd /home/sqllyw/v4l-dvb/v4l
sudo make install
sudo reboot

saa7134 is back.

Updated on Aug 31, 2009: upgraded to Kernel 2.6.31 rc7, saa7134 is part of that.

This is about 9.04 but should be applicable to other versions as well.

1) get your kernel version from http://kernel.ubuntu.com/~kernel-ppa/mainline/
I downloaded following filels:
linux-headers-2.6.30-020630_2.6.30-020630_all.deb
linux-headers-2.6.30-020630-generic_2.6.30-020630_i386.deb
linux-image-2.6.30-020630-generic_2.6.30-020630_i386.deb

2) put the files under one directory, cd to that directory
3) sudo dpkg -i *.deb
4) sudo reboot, choose the kernel version from boot menu, that’s it.

My personal experience in the setting up of mythtv, tested PCI Card
KWorld Plus TV Analog Lite PCI(PVR 7134E) and ITO23 REV:4E(a 7135 based card)

Got a need lately to view TV from a laptop computer and decided to set up Mythtv for that, it took a while to be able to do this, the main problem is the TV tuner card, the tip I can give is, find a card fully supported by Mythtv before starting, this will save a lot of time. anyway, I finally got my box up and running almost 99%, still have 1% that I’ll figure out later, steps to setup:

0) hardware part, I use a simple approach, connect the video out and audio out of my set top box to the video in and audio in of the tuner card, this make it the composite input(the card has RF and S-Video inputs, but I’m not using it), for the audio, since my set top box uses RCA, so I have to use a RCA to mini 3.5 jack adaptor. optionally you can connect from the audio out of tv tuner to the line in port of your pc’s sound card, this is not really necessary if you want to use only mythtv for TV viewing.

1) install first Ubuntu 9.04 desktop

2) sudo apt-get update

3) dmesg | grep saa
you should see your card there, but type of card is set to UNKNOWN or GENERIC(can’t remember exactly), if you find the card there, proceed to #4 otherwise Google:)

4) sudo pico /etc/modprobe.d/alsa-base.conf
look for the line: install saa7134 /sbin/modprobe –ignore -install saa7134 $CMDLINE_OPTS{….}
change the $CMDLINE_OPTS to card=59 tuner=56, save and reboot

5) dmesg |grep Kworld

[ 12.104209] saa7134[0]: subsystem: 17de:7128, board: Kworld/Tevion V-Stream Xpert TV PVR7134 [card=59,insmod option]
[ 12.108898] input: saa7134 IR (Kworld/Tevion V-Str as /devices/pci0000:00/0000:00:1e.0/0000:02:04.0/input/input5
[ 13.104060] saa7134 IR (Kworld/Tevion V-Str: unknown key: key=0×03 raw=0×03 down=1

ok, you card is found and identified, let’s do some testing without mythtv for the meantime

6) sudo apt-get install xawtv
this installs the tv program, click /Applications/Sound & Video/XawTV, you see the xawtv screen, right click and set video source to ‘composite1′, you should see the TV live, if you connect an audio cable from the card’s audio out port to your pc’s audio in, you should hear the sound, if no sound, play with the volume control, look for PCM/line in, do some tinkering there until you heard the sound, however we are not going to use this sound from line in port of pc for mythtv.

up to this point, our hardware/software for the TV is up, let’s install mythtv:

7) sudo apt-get install mythtv
click System/Adminstration/MythTV Backend Setup
basically the audio has to be set to dsp1
9) click Applications/Sound & Video/MythTV FrontEnd
audio has to be set to dsp
10) let’s try a remote FrontEnd
as long as the password for the mythtv is correct, this does not require too much setup, audio is different from #9

more on 8 to 10 when I’m free.

Note: just tried a new PCI TV board, it is very similar to this one:

http://www.pacificgeek.com/product.asp?id=49178

At back of the card, it says ITO23 REV:4E
took a while to figure out: card=79 tuner=43

Entity relations

@SecondaryTables

One to Many relationship

@Entity
public class Publisher {

 	private Long id;

	@Id @GeneratedValue
    public Long getId() { return id; }
}

@Entity
public class Book {
	private Long id;
	private Publisher publisher;

	@Id @GeneratedValue
	public Long getId() {return id;}

	public Publisher getPublisher();
}

Publisher has many books, so publisher is in the ONE side while book in the MANY SIDE.
To have a unidirectional relation we can annotate the Book class as:

	@ManyToOne
	public Publisher getPublisher();

or

	@ManyToOne(fetch = FetchType.LAZY)
    @JoinColumn(name="publisher_id")		// this will use publisher_id as attribute in the table, otherwise it will be publisher
	public Publisher getPublisher();

Notice there is no need to have MappedBy as it is optional in the unidirectional ManyToOne relation.

with this approach, we can find the publisher from the Book, but finding books from an instance of Publisher object is not possible unless you use some queries. to be able to do that, we need to make the relationship bidirectional, just add this annotation in the Publisher side(ONE side):

@Entity
public class Publisher {

 	private Long id;
 	private List<Book> books;

	@Id @GeneratedValue
    public Long getId() { return id; }

	@OneToMany
	@OneToMany(mappedBy="publiser")
	public List<Book> getBooks() { return books;}
}

It is mandatory to have mappedBy annotation in a bidirectional relationship and it has to be in the ONE side(maybe I’m wrong?), the ‘publisher’ here refer to the ‘publisher’ property in the MANY side. you can specify other attributes in the OneToMany annotation:

	@OneToMany(cascade= CascadeType.ALL, mappedBy="publisher")
    @OrderBy("Area desc")

Note: delete orphan to be added.

One to One relationship

@Entity
public class Door {
    private Long id;
    @Id
    public Long getId() {return id;}
    public void setId(Long id) { this.id = id;}
}

@Entity
public class Key {
    private Long id;
    @Id
    public Long getId() { return id; }
    public void setId(Long id) { this.id = id;}
}

their generated sql:

create table Door (
      id bigint not null,
      primary key (id)
  );

create table Key (
      id bigint not null,
      primary key (id)
);

The only thing required to setup a one to one relationship is: to annotate in the one side of the relationship, let’s do it in the Door:

@Entity
public class Door {

    private Long id;
    private Key key;
    @Id

    public Long getId() { return id; }
    public void setId(Long id) { this.id = id;}

    @OneToOne
    public Key getKey() { return key; }
    public void setKey(Key key) { this.key = key;}
}

This change will trigger a change in the SQL as well:

create table Door (
      id bigint not null,
      key_id bigint,
      primary key (id)
  );

alter table Door
      add index FK2097CE549DF154 (key_id),
      add constraint FK2097CE549DF154
      foreign key (key_id)
      references Key (id);

Door owns the relationship in this case, and it is unidirectional, you can access Key from the Door, but not vice versa. to make it bi-directional you can annotate the other end:

@Entity
public class Key {
    private Long id;
    private Door door;

    @Id
    public Long getId() {
        return id;
    }

    public void setId(Long id) {
        this.id = id;
    }

    @OneToOne
    public Door getDoor() {
        return door;
    }

    public void setDoor(Door door) {
        this.door = door;
    }
}

and you got the new sql:

create table Key (
      id bigint not null,
      door_id bigint,
      primary key (id)
  );

You will notice Key now has door_id as a foreign key, both end owns the relation, this isn’t too good. A preferable way is, one end is the owning entity while the other end is a owned entity, in example above we can make the Key a owned entity while Door is a owning entity, to make this happen, add a mappedBy in the Key class:

@OneToOne(mappedBy = "key")

the “key” field is referring to the key field in the Door class, this will generate a new SQL:

create table Key (
 	id bigint not null,
    primary key (id)
);

We don’t have the door_id in this case as Key is only owned by Door, the relationship is maintained by Door with key_id as foreign key stored in the Door class. this allows us to travel from Key to Class as well.

If we want some operations cascaded from the owning class to the owned class, we can do:

  @OneToOne(cascade = CascadeType.REMOVE, fetch = FetchType.LAZY)
  public Key getKey() {
      return key;
  }

If you delete the Door, then the related Key object will be deleted as well. however this cascade operation does not happen when you do a batch delete on Door. If I understand correctly, the cascade and fetch type should always be specified in the owning entity side, thus it should be in the class that does not have a ‘mappedBy=’ clause, might not be correct. class with ‘mappedBy=’ annotation does not maintain the foreign key, it is in the other class.

Main entry text

mvn install:install-file -Dfile=jai_core.jar -DgroupId=javax.media -DartifactId=jai-core -Dversion=1.3.3 -Dpackaging=jar -DgeneratePom=true
mvn install:install-file -Dfile=jai_codec.jar -DgroupId=javax.media -DartifactId=jai-codec -Dversion=1.3.3 -Dpackaging=jar -DgeneratePom=true

to check if tomcat is running:

ps -ef | grep java

getting mysql schema:

mysqldump –no-data -u user -ppassword dbname >schema.sql

Tomcat Plugin Home Page

Tomcat plugin is good for the deployment of servlet applications into a Tomcat server, it has three major functions:

1. Deployment

This is the functions I used often, I’d consider this as the most important functions of this plug in

2. Container related goals

Using this group of goals you can query the status of a running Tomcat server.

3. Running an app in an embedded Tomcat Server

This is very handy in development, you can run your app even you do not have a tomcat server installed, the embedded server is version 5.5.25 at the time of this writing

To get started you should have a servelet application based on Maven 2 running first, go to the directory of the application where pom.xml is located then try:

maven tomcat:run

pointing your browser at http://localhost:8080/myproject

You should see the program running. that’s the #3 feature listed above.

The easiest way to get started with Plugin is, have a tomcat running at port 8080, add a user ‘admin’ without password to $CATALINA_HOME/conf/tomcat_users.xml：

  <user username="admin" password="" roles="standard,manager"/>

try following container related goals:

mvn tomcat:info

mvn tomcat:list

To deploy your app:

mvn tomcat:deploy

mvn tomcat:redeploy

mvn tomcat:undeploy

Now the plug in is running, let’s do something more, I’d prefer to have a installed Tomcat Server running at port 80 and the embedded server at its default port 8080, with that, I can test my app first in the embedded server at port 8080 and deploy it later into the installed Tomcat server and test it in the port 80, of course you can choose whatever port number you like as long as the installed Tomcat server’s port is different from the embedded Tomcat’s port.

1. update ./m2/settings.xml

<servers>
    <server>
      <id>local_tomcat</id>
      <username>admin</username>
      <password></password>
    </server>
</servers>

2. Add this blog to pom.xml

		<plugin>
		        <groupId>org.codehaus.mojo</groupId>
		        <artifactId>tomcat-maven-plugin</artifactId>
		        <version>1.0-beta-1</version>
		        <configuration>
		          <server>local_tomcat</server>
		          <url>http://localhost:80/manager</url>
		        </configuration>
		 	</plugin>

the entry in the settings.xml is actually to specify the user name and password while the port number is specified in the pom.xml’s url part.

• To set global environment variables

One way is, update /etc/bashrc
for local user, you can ~/.MacOSX/environment.plist

• To install tomcat 6

Some simple steps in setting up Exim4 in Ubuntu 8.10.x, might be applicable to other distributions too.

Some related Links:

Setting up Exim SMTP Auth
SMTP Relaying Via a Smarthost
Gmail POP3 with Fetchmail

apt-get install mailx
apt-get install exim4

Note: this will remove the postfix if it is installed

To configure:

dpkg-reconfigure exim4-config

restart exim4:

invoke-rc.d exim4 restart

To relay with a smarthost, you need to update:
/etc/exim4/passwd.client, put something like this in the file:
example.com:abc@example.com:mypassword

restart the exim4 and do a test sending:

echo “my test” |mail -s “testing exim4″ abc@abc.com

check /var/log/exim4/mainlog and you should see some messages in the log about this test mail.

So far the exim was setup without smtp authorization, you need to enable SMTP Auth if you want to use RoundCube web mail in the same server.

Here is a guide about setting up the SMTP Auth.

Create a password for smtp and save it in /etc/exim4/passwd(create if not exists)

htpasswd -nd usernameforsmtp

copy and paste it to /etc/exim4/passwd (this has to be done when a new user is added, you do not have to restart exim4)

Update /etc/exim4/exim4.conf.template:

AUTH_SERVER_ALLOW_NOTLS_PASSWORDS=1
uncomment plain_server
sudo update-exim4.conf

restart exim4, this time the basic auth is on at least in the server side.

The title is not misleading, everybody said it will take a few min to setup Apache Continuum, but I spent more than two days to have it up and running, looking back, I’d like to document steps of how to really set it up in 10 minutes.

First, take a look at the Apache Continuum’s home page, then forget about it, follow the steps below, if you get it up and running, then you go back to the link above:

Download Apache Continuun

make sure it is 1.3.1, unzip it to a directory.

Configure and run Continuum:

you need to update two files only:

update the application.xml(/path_to_continuum/apps/continuum/WEB-INF/classes/META-INF/plexus/)

uncomment this line:

<allowedScheme>file</allowedScheme>

update this two lines:

<from-mailbox>developer@gmail.com</from-mailbox>
<from-name>Developer</from-name>

update jetty.xml(/path_to_continuum/apps/continuum/conf):

 <New class="org.mortbay.naming.factories.MailSessionReference">
    <Set name="user">sample@sample.com</Set>
    <Set name="password">password</Set>
    <Set name="properties">
      <New class="java.util.Properties">
        <Put name="mail.smtp.host">smtp.sample.com</Put>
      </New>
    </Set>
  </New>

start Continuum, ./continuum start

point browser at http://localhost:8080/continuum, if everything ok, you should see the Continuum’s page.

Add a Maven 2 project:
the POM url is kind of confusing, newbie will think this is actually the pom.xml in the svn repository and end up using this:

svn+ssh://myname@myserverserver/var/svn/repos/project1/pom.xml

this is wrong, the POM url here refers to an actual pom.xml file and the svn path should be specified in the pom.xml. to add a Maven 2 project, you need to have a project imported already into svn, you can check out the pom.xml, and make sure you have following entry:

<scm>
 	<connection>
   	scm:svn:file:///Users/guest/svn_dir/repos/project1
   </connection>
 </scm>

if you do not have this entry, add it, and make sure you commit the change back before proceeding.

the path to svn repoistory here is only a sample, you have to use whatever applicable to you. now put this file in a place that you can access, it can be in a web site or can be in the hard disk, if it is in the hard disk, then you specify POM url as follow when adding a Maven 2 project:

file:/Users/guest/test789/pom.xml

then click add, if you do not see any error, it’s already up and running, hope it is within 10 minutes.

You can find the Continuum forum here

Additioal Tips:

To change port number used by Continuum for its web interface, update jetty.xml, look for:

<Set name="port"><SystemProperty name="jetty.port" default="8000"/></Set>