A Path Less Taken

In a recent post I touted the coolness factor of the Spot ORM. Based on a response to that post I prepared this short tutorial on how to get started. This is culled from personal experience and guidance that I received from the software’s author. I hope this is helpful to you if you are interested in giving Spot a try.

No, this is not about Doctrine and Propel, although I’m sure those are both really fine PHP ORM tools if you are willing to endure the learning curve. This is about 2 lesser known but notable ORM tools that deserve a nod of recognition. I’ve tried a number of PHP ORM tools over the last year and many of them had merit. In fairness, as you look at one attempt you begin to realize how much each of them owes to those that came before them. There are so many variations on a theme that you can easily get confused and give up in frustration. But my recent efforts have revealed two that I find very useful and note worthy depending on your project. Here is a quick run down on each.

I’m not an active Python developer at the moment, but I have been searching far and wide on the web for ORM solutions that might interest me. One Python based ORM which struck me as really interesting is SQLAlchemy. If you are interested in trying an ORM tool and you read the description on SQLAlchemy’s home page you would swear you had died and gone to ORM heaven! If this solution actually accomplishes what it claims then it could be a very interesting choice. I did not find a PHP port so I’ll have to keep looking for another PHP ORM that I’m interested in trying. Sigh.

Object Relational Mapping (ORM) tools allow developers to interact with databases through objects. The basic object in the typical solution is the model. The model is essentially an object representation of a single table in the database. In fact, implementations that subscribe to the activerecord design pattern will often view the model as an object representation of a single row of a specific table. Since the target data store in most cases is an RDBMS (excluding the recent surge in distributed Document based DB’s) the typical ORM needs some way to represent foreign key relationships in an object oriented way. To facilitate this the field has evolved a set of conventions or names for foreign key (FK) relationships in the DB based on the object’s (table’s) role in the given index. Here is a brief but hopefully useful explanation of four relationship types.

jonathangeiger of the Kohana PHP Framework community has been working hard lately on a new ORM that shows promise. The github repository can be found here if you would like to check the code out. I decided to point my PHP Code Explorer at it to figure out the Jelly object model. Below is a screen capture of what I found. You can click the image to see a larger view.

I hope this view is helpful to you. I know it helped me to get a better perspective of Jelly and how it was designed.

Update: The official Jelly web site can be found at http://jelly.jonathan-geiger.com/.