Naming Your Adapter When Using the Adapter Pattern


At some point in your project you may need to use a third party API.  One common approach for using third party APIs is to use the Adapter Pattern.  The API that you use may be very specific to a particular technology.

When you do create these adapter classes that utilize that API you should avoid names for your adapter that relate to the particular technology.

 For example, let's say that you have an API for the Sony Toughpad barcode scanner.  You wouldn't want your adapter classes to be named SonyToughpadBarcodeScannerAdapter. You should avoid naming your adapter like this in favor of something more generic like BarcodeScannerAdapter.

 This is because in the future you may need to adapt your adapter to a different API such as the Intermec Handheld Device barcode scanner.  The name for you adapter will be adaptable to this API but the name will be misleading.  At this point you have the choice of renaming your adpter to the more generic name.  However, renaming comes with a cost.

 Renaming files in version  control systems means that you lose the history of those files.  That is because in some version control systems the renaming of a files appears as two separate operations; delete (old file name), create (new file).  Renaming files becomes problematic in version control systems, specially when you have branches for different versions of the applicaiton.

 Things will be much easier if you simply avoid naming your proprietary classes to something relating to the underlying technology.

 This concept applies to the adapter as well as any associated classes upstream.

 The same thing would apply if you created an internal API for a reporting system or file transfer system.

 So please avoid even hinting at the underlying technology brand when naming your adapters. Your maintenance programmers will thank you.

Comments

Tuukka Haapaniemi said…
Basing the decision not to use the technology names on the fact that the version control history breaks, is not very valid. Modern distributed version control systems, like Git and Mercury, support file renaming and relocation. If you're not using those, I seriously recommend you to give them a try. That reasoning would also lead to resisting file renaming altogether, which hinders refactoring considerably. Seldom do you get the names right from the get-go, and renaming is a major part of refactoring, just so that the maintenance does not become too cumbersome.

Version control aside, I also advocate the use of very specific, technology based names for the adapters. The reason for this is that the software projects I've been working with tend to need more than one adapter for the same task. Say that you have an online store, from which you post packages to the customers. This store could be integrated to a number of courier service providers, each doing the same thing, i.e. registering the package for delivery, but through a set of separate interfaces. Each interface and hence each adapter has the same semantic meaning, but each adapter is different due to different service providers. Naming the adapters accordingly will save you the headache of wondering which adapter is used with which interface.
January 30, 2016 at 11:17 PM
Post a Comment

Popular posts from this blog

Putting Files on the Rackspace File Cloud

Recently, I’ve been playing around with Asp.Net MVC and I (along with some friends) am creating an Asp.Net MVC application that can access files from our website. This means that files must display on the site as well as provide a way to download/upload them. I won’t give away all the details of the project but I will say this: Sharing photos is one of the driving forces for our project.  The files are only image type files( jpg, etc) The files must be uploaded by the users and stored somewhere within our system.  We don’t necessarily know at this time how much storage space we need but we think that we will need a little at first but possibly a lot more down the road. This means that the storage of these files must be scalable. The uploading of files must be controlled by our system.  Right off the bat we started our application and were uploading files to the web server with no problem.  We only did it this way to kind of prove our system model. We were just basically writing
Read more

Setting Up XnaMobileUnit to run unit tests in Windows Phone 7 XNA Games

Image
Introducing XnaMobileUnit I decided to roll my own unit test framework for Windows Phone 7 XNA games.   I have a couple of reasons why I decided to roll my own.  Among them was the fact that you cannot reference regular .Net assemblies from Windows Mobile Framework.  That means that referencing NUnit was out of the question.  Also using some of the other libraries that can be used for Silverlight Applications didn't quite feel right for me. The framework that I have created, XnaMobileUnit, is very simple to use and provides all the necessary functionality to run unit tests.  I am still working on the framework to provide a few more capabilities and features, but for the most part it is ready to use out of the box.  XnaMobileUnit is open source licensed under the Apache 2.0 License and can be downloaded from http://code.google.com/p/xna-mobile-unit/ .  I recommend downloading the latest source and compiling it then referencing the new DLL. Otherwise you can simply download the l
Read more

Reasons to Stay Away from CSLA

We've been using CSLA - Component-based Scalable Logical Architecture in a product for approximately one year and while it solves quite a few problems it violates some major good practices. No Clear Separation of Layers While at the time that we chose CSLA we knew very little about domain driven design, it is no excuse for the framework to possess the data access layer within the business object. This makes it extremely difficult to separate the concerns. As a result, it is difficult to mock the data-access layer creating many problems when attempting to perform unit tests. This also puts the additional responsibility of data access on the business object class, violating the Single Responsibility Principle( SRP ). Data access layer should be totally separated from the business object period. Having the data access within the within the BO class is not the only additional responsibility placed on the class but is the most significant one I can think of. No Context Validation
Read more