Book Review: PHP Web 2.0 Mashup Projects
You can find this review in podcast form on the Zend Developer Zone PHP Abstract Podcast.
I received an e-mail recently from a very nice gentleman at Packt Publishing, a UK-based publishing company focused on providing hands-on application-oriented publications to IT professionals, particularly those specific to open source technologies. Their representative asked if I would be willing to review one of their books, namely PHP Web 2.0 Mashup Projects by Shu-Wai Chow. Reviewing books is not something I had done before, so I thought I would give it a good old-fashioned college try.
In a supersaturated market, it is difficult to make an impression with a PHP book these days. The books of real value are those that focus on ways to apply the language to real world problems. These books delve into the depths of a particular application domain, showing PHP code and outlining design principles along the way. They are useful to current and prospective PHP programmers alike because they can introduce both not to PHP itself, but to an existing class of problems and how PHP can be applied to solve them. PHP Web 2.0 Mashup Projects is one of these books.
Most technology-related books on the shelves are several inches thick and an inherently daunting chore to sift through. Luckily, this book is not one of those. Do not let the size fool you, though; it is positively packed with useful information. It hits the high points of each topic it covers, giving you enough in the way of code samples and step-by-step explanations to get started, as well as resources to help you get better acquainted with topics that might be of particular interest to you.
The book is divided into six chapters, each of which covers a set of particular protocols, data formats, and APIs for acquiring and processing data in order to create a particular mashup application. These projects include:
- A search engine to find products on Amazon by their Universal Product Code
- A search engine to combine results from MSN and Yahoo!
- A video jukebox that pulls songs from Last.fm and videos from YouTube
- A traffic incident reporting application that sends SMS alerts
- An illustrated tube station line map using Google Maps and Flickr for related photos
The book's structure and layout make it easy to follow, whether you prefer to read it linearly or jump around to specific sections. It is an excellent reference that I can see myself returning to time and time again.
One of the strengths of the book is that it has a very wide base of coverage. It starts by introducing basics in interacting with web services and extracting the desired data from their responses using core PHP libraries. The REST, XML-RPC, and SOAP protocols and the WSDL standard are all covered in enough depth to get you started, so you can work with a web service regardless of the protocol or protocols it offers. The author does an excellent job of selecting example web services and data standards from large and well-known to small and obscure. For real world APIs, you will find the likes of Amazon, YouTube, Google, and Flickr, as well as sources that might not be household names, such as the Internet UPC Database. Data standards include general formats like XML, RDF, and JSON and more specialized formats like RSS and XSPF.
Another strength is that the book encourages good principles from the start. It advocates object-oriented design principles for code reuse and a DRY philosophy. It suggests using third-party libraries such as those in PEAR in order to avoid unnecessary reinvention of the wheel, but still shows you how to roll your own if and when it becomes necessary. The books also covers usability, particularly in the last chapter when it discusses AJAX and race conditions, and pays special attention to application security, an area of increasing concern in web applications. Unlike some books, this one includes tips for development outside its own showcased projects to alleviate you from having to spend your own time troubleshooting common issues or digging for solutions to "gotcha" situations.
And last but certainly not least, the book demonstrates that sometimes you have to be resourceful in locating and acquiring your data, particularly in Chapter 5 where one of my own areas of interest, web scraping, is covered. The topic is explained in plain language and supplemented with examples walking you through exactly how it can be used to acquire data for your own mashups. Web scraping is not a frequently broached topic and I applaud the author for making a point to include it. I believe it is a genuinely useful methodology that can help in data acquisition when no other options are available.
I cannot give the book an entirely glowing review, though. There are some errata present, both in content and code samples. Most are small, but some are enough to throw off a reader not already familiar with the material being covered. I've submitted some of these via the publisher's web site already, though I have yet to receive any related communications or see them show up on the web site at the time that I write this review. These issues are able to be corrected, though, and the quality of the book's content outshines them.
Overall, PHP Web 2.0 Mashup Projects is an excellent example of creativity in finding new ways to aggregate data sets in useful combinations. It is a testament to the possibilities of the internet when access to data is opened up and freedom to use that data enables developers to create exciting and inspiring new solutions. Mashups show the internet's potential increasing in leaps and bounds and this book can get you on your way to contributing to their future development.
Zend Framework and Remember The Milk
I've posted a few times on Twitter related to my latest project and a few people have already asked me about it, so I figured it was worth a blog post.
My first project for the Zend Framework was Zend_Service_Simpy, a service module providing a lightweight wrapper around the API for the Simpy social bookmarking service.
My latest project is another service module for the Zend Framework. This time, though, it's for the Remember The Milk API. RTM is basically a TODO list on serious steroids. It's the Swiss Army Knife of task management. It allows you to manage multiple lists of tasks. You can add them easily from a variety of mediums, tag them, prioritize them, set deadlines for them, have them repeat, get reminders for them, tie them to physical real world locations, and share them. RTM offers great support for integration with Google applications including Google Calendar, iGoogle, and Gmail (plus offline access powered by Google Gears). They're also very big into supporting mobile devices, including those running on Windows Mobile as well as the iPhone.
If you like, you can check out my original proposal for this module. I can already say that the API will end up changing a little, though, but it's good enough to give you a general idea of what the capabilities of the finished service module will be. I only actively started implementation recently and things are progressing at a fairly rapid pace. I still have unit tests and documentation to handle, but hopefully there's a shot at seeing it moved to core within the next two releases of the framework.
Google Reader and Yahoo! Pipes
I ran into a situation recently that I thought I'd share. I use Google Reader to manage the feeds that I read regularly. PHPDeveloper.org is among my favorite news syndication web sites. However, some of its posts, in particular those dealing with job posts or additions to CakePHP's Bakery, aren't interesting to me.
Eventually, I came to the conclusion that I could wrap the feed in a Yahoo! Pipe in order to filter out the uninteresting information. (I know, the irony of using Google and Yahoo products together might seem anything from ironic to downright unholy to some.)
Unfortunately, doing so meant that I had to remove the original PHPDeveloper.org feed from Google Reader and add the new pipe-wrapped feed in its place. Because (as best I can tell) certain things are tracked per feed rather than per URL (old items) or per item (read statuses), this meant losing all information specific to the old feed.
Granted, I only had to do this once, but I wish it had occurred to me earlier. Google Reader may have search capability (which took forever to be included), but that's not the same as being able to have content filtering automatically handled for me whenever I view the contents of a feed.
So my line of thought continued. It would be nice if there was an easy way to maintain the user experience of adding feeds through my preferred browser, Mozilla Firefox, but to have new feeds be automatically wrapped in a Yahoo! Pipe "behind the scenes." This would allow me to go back and manipulate feed content later if I saw a repeating pattern in specific content that didn't interest me.
Another unfortunate trait of this situation is that Yahoo! Pipes doesn't currently offer a web service API, or it might make implementing my idea significantly easier. While the AJAX interface exposes server interaction logic, it's obfuscated to the point where it makes reverse engineering attempts infeasible. It's unfortunate, because I think a marriage of the features of each of these services would make the result all the more useful for their users.