For a project at work I needed to parse some XML. The xml contained several entries of entries. Only part of the information given for each entry was actually needed by the software. Therefor the object used did not have a property for each field in the entry. Just doing the follow code will not work:


[currentObject setValue:(id)elemValue forKey:objProperty];


This will raise an exception as soon as an attempt is made to set a property that does not exist. Wrapping the setter in a check to see if the object has this property will solve this problem. Using responseToSelector a test can be made to see if the object reacts to a selector


if ([currentObject responseToSelector:NSSelectorFromString(objProperty)]) {
[currentObject setValue:(id)elemValue forKey:objProperty];
}


This should do it. Except it doesn’t completely. If your object has a function that is named like objProperty, responseToSelector will return true, because your object does response to it. Its just not a property that is set-able, but instead, a method. Actually, reponseToSelector only controls for function. Since synthesizing a property creates a method with the name of the property, it works for detecting properties as well. To avoid getting an exception on setting values for keys. Make sure that objProperty does not contain a value equal to a method in currentObject that does not represent a property.

NSXMLParser requires a NSData object to init with data. Following the conversion cod eI found on Apple’s site I did the following:


NSData *xmlData = [NSData dataWithBytes:[xml dataUsingEncoding:NSUTF8StringEncoding] length:[xml lengthOfBytesUsingEncoding:NSUTF8StringEncoding]];


Passing the xmlData through to the parser ended in me getting an error 4, which means document empty. I tried everything. Printed out the xml string to see it was full, compared length of data and string, releasing the parser. Still error 4. For the heck of it I tried giving the parser a nil for data. That gave an error 5. Trying to give an empty NSData object also gave error code 4. After 3 hours spent in google and stack overflow I was not even a small step further.

I really didn’t know what to do when, while scanning all these pages I saw a piece of code of converting the string to data without giving the length. for the heck of it I decided to try this:


NSData* xmlData = [xml dataUsingEncoding:NSUTF8StringEncoding];


What do you know it worked. No idea why. Naturally. I wanted to know what the length of the data object was now. I was fully expecting it to be bigger than original. To my surprise it was the same length as the string. The same length I was giving it in the first call.

Why this now wrks I have no idea. The first and the second calls should give the same results, yet somehow they don’t. If you got this problem, then converting without giving the function a length. Good luck and happy programming.

sudo su - postgres
This will log you into the postgres server user and move you to its directory.

bin/pg_ctl -D data stop
This will stop the server from running.

Step 1: PostgreSQL

You can download a binary installer from the PostgreSQL site. It will install Postgre to your /Library folder. This should take care of the Postgre part.

Step 2: Sql-Ledger

Ledger is written in perl, and requires perl to run. Luckily Snow Leopard comes standard with perl. What it does not come standard with is the perl binding for PostgreSQL. The following 3 commands will take care of that:

sudo perl -MCPAN -e "install +YAML"
YAML is not necessary but it can be helpful, and lacking YAML might give you problems with DBI.

sudo perl -MCPAN -e "install DBI"

sudo perl -MCPAN -e "install DBD::Pg"

This will install the perl binding for PostreSQL.

Download Sql-Ledger from here. As of this writing, the latest version is 2.8.31

Unzip and untar the archive and follow the instructions in the readme file. I did not manage to build using the automated script and had to do everything by hand. There is also this page on the ledger page with a somewhat oldish instructions for OS X.

Once Ledger is tested and it is working, its time to get the python binding

Step 3: Python binding

The pyhton binding is for PostgreSQL. You can get the binding library, called Psycopg2 (don’t ask me why) from here. Download and upack it. Installing is done by running:
python setup.py build
sudo python setup.py install
If you are not able to build it might be a problem with your path. Python is trying to link to the PostgreSQL bin files. Try to add it to your path and then to build again.

And your done. Enjoy (or not) working with Django-Sql-Ledger

Solr is a “popular, blazing fast open source enterprise search platform from the Apache Lucene project”. Drupal search has traditionally been less then adequate. Using Solr to index and search you Drupal site is therefore quite helpful.

To start, download the Solr package (link) and the Solr drupal module (link). the current standard distribution is Solr 1.4 and Drupal 6.
Unarchive the tar files in a location that is reachable from your server.

We will assume that tomcat is installed in /opt/tomcat

In the Solr package there is an example application. That is the Solr application we are going to use. If you can write your own, you probably don’t need this tutorial.

Step 1:
Copy the example application, found in the solr package under example/webapps to the tomecat webapps folder which should be /opt/tomcat/webapps. when copying the file rename it to something other the solr. this will enable you to have multiplie solr installations running under one Tomcat container. We will assume the app is renamed new_solr_inst. If you are in the tomcat root folder the command should like like this:
cp /path/to/solr/package/apache-solr-1.4.0/apache-solr-1.4.0/example/webapps/solr.war webapps/new_solr_inst.war

Tomcat will pick up the war file and open it automatically after a restart.

Step 2:
In the tomcat root folder create a folder to hold your Solr configuration and data.
mkdir new_solr_inst
Copy the example app configuration into that folder
cp -R /path/to/solr/package/apache-solr-1.4.0/apache-solr-1.4.0/example/solr/* new_solr_inst

Step 3:
You need to replace the original schema.xml and solrconfig.xml with the ones provided by the drupal module. Copy these two files to your Solr configuration folder. Assuming you are currently at the tomcat root folder
cp /path/to/solr/module/schema.xml new_solr_inst/conf
cp /path/to/solr/module/solrconfig.xml new_solr_inst/conf
The Solr app is now ready for use.

Step 4:
The last step is needed to inform Tomcat about the app and its settings. Go to conf/Catalina/localhost and create the file new_solr_inst.xml. In that file fill the following content:

Step 5:
You need to restart tomcat
/opt/tomcat/bin/shutdown.sh
/opt/tomcat/bin/startup.sh
And your Solr should be running. There are three tests we can use to check everything went well.

• Tomcat has opened the war file. There is now a folder new_solr_inst in the webapps folder of Tomcat.
• Solr has made a data directory in the Solr home folder (that is /opt/tomcat/new_solr_inst)
• In your browser go to the Solr app http://domain:tomcat_port/new_solr_inst and you should see a welcome screen

If all is well you are ready to combine this Solr installation with your Drupal

Step 6:
Activate the Apache-Solr module and go to the Solr module settings page admin/settings/apachesolr
Assuming tomcat and drupal are running on the same server fill in localhost as the host name, fill in the port tomcat is running on and solr path should be your app name in the tomcat webapps folder. In our case is it new_solr_inst.

Save the settings. Drupal should inform you that a connection is made with Solr

Notes

• It is not nessecary to put the app configuration in the Tomcat root folder. you can probably (though I have not tested this) place it anywhere you want. you do need to make sure that the environment tag in the Tomact configuration file (the one in the conf/Catalina/localhost folder) is pointing to the right place.
• If you make changes to the schema you may need to dump all the indices Solr has made. you can do this via the Drupal admin interface for the Solr module.

Using cascading delete I can define that if an object is removed from its table, all relation tables entries with that object id are also removed. Quite handy. the way this achieves is MySQL is through the use of foreign keys. If I define in the table members the unique id of a user as a foreign key I can also define tell the RDMS to delete the entries in members where that foreign key exist. Here is an example of the definitions:


CREATE TABLE members (
usr_id VARCHAR(30) NOT NULL,
grp_id VARCHAR(30) NOT NULL,
FOREIGN KEY usr_id REFERENCE user (usr_id) ON DELETE CASCADE,
FOREIGN KEY grp_id REFERENCE group (grp_id) ON DELETE CASCADE);


As you can see the table is created normally by defining the fields, then the foreign keys are named. The REFERENCE shows to which column in what table they refer to.

Before I started working all I heard from the working people is how they miss the student time. How much more demanding working life is. It won’t be an overstatment to say I went to my first day a bit worried.

My first day was like Christmas.  I got a mobile phone, a laptop and a car. And that was pretty much it. The next two days have been just as nice. Together with all the people who started working for Getronics-Pinkroccade this month we had a 2 day into plan. Some presentation, some group talks, and beer. Yes every building in the company has its own beer tap with a barman. I like that. Did I mention, its free beer? The people I met were intellegent, interesting and fun. If this is how most people in the company are, I made a good choice coming to work here.

From the forth day on, and for the coming 2 weeks I’m in courses in Amersfoort. Its quite relaxed. I’m doing the courses alone, with a tutor I can turn to for questions. My tempo is quite high (im going almost at twice the expected rate) so I might be finished earlier then expected.

After I finish the courses I have a month of stage to build some web app (interesting ideas for what exactly are welcome). Then it will be time to actually work on real projects.