Looking at the future of PHP from the design blueprint of Zend Engine 2.0 (1)

Some miscellaneous remarks

First of all, the original intention of writing this article. It has been a while since I got the design blueprint document about Zend Engine 2.0. After reading it, I had the urge to write a review - because according to the description of the document, the next generation of PHP will be a more consistent object-oriented version. The language used for development is at least given more object-oriented features. But the question that comes with it is whether a language like PHP, whose initial goal is rapid Web development, is worth modifying itself to be comprehensive? This issue seems to require some review articles to discuss, and I am willing to put forward my own opinions. But then some busy things brought this impulse back to my stomach little by little. It wasn’t until recently that I thought of it again, so I read the design blueprint document several times, and this is how I came up with this article. (When I first got this English document, I had plans to translate this design blueprint document into Chinese, but considering that first, it is not the final blueprint, second, everyone should develop the habit of reading the original text directly, and third, the translation is always It will inevitably lead to some obscurities, so I gave it up for the time being. However, when I wrote this article, I decided to "take the risk" of translating it. Please feel free to correct me if I can't convey the meaning accurately.)

The second thing is. Explanation of the focus of this article - In this review I will mainly review the object-oriented features that will be greatly enhanced in PHP in the future. If you are a PHP developer, then I guess you should know some of the object-oriented features in the PHP language; but since PHP is generally used in "extreme rapid development environments" (this is a word I coined myself, indicating some Customer-oriented website development is characterized by a very short construction period and unclear customer requirements), so there are not many developers and development projects that really use its object features extensively; in addition, the existing PHP object model is relatively weak compared to C++ and The weakness of Java also limits the use of this feature. However, in future versions of PHP, the focus of modifications will be on the object-oriented model in the language, improving many bad features in the existing version and adding other features. Therefore, discussing the future of PHP focuses on discussing the object-oriented features of PHP.

Okay, let’s get down to business and take a look at the new features of Zend Engine 2.0.

Summary of the Zend Engine 2.0 design blueprint (draft)

It can be clearly seen from the design blueprint (draft) that the next generation Zend Engine is based on a new object-oriented model-based. If you have ever used the object-oriented features of the existing PHP 4, you may feel a little awkward while finding a little bit of Java or C++ - not only the lack of object-oriented syntax, but also sometimes you will get unexpected results. Results - This is all due to the less elegant object-oriented model in the existing Zend Engine 1.0 that supports PHP 4.

To put it simply, the next generation of Zend Engine will move closer to Java and draw heavily on its object-oriented model. Judging from the types of improved features, they can be divided into three categories: The first category is the improvement and enhancement of the existing object-oriented model, including the definition of builders and destructors, and the addition of private member variables and static members. Object-oriented features such as variables, multiple inheritance, and overloading; the second category is the modification, addition and deletion of the control process, such as adding try/catch/throw exception handling mechanisms; the third category is the modification, addition, and deletion of functions, such as Increment function for string offset. (For details of each type of improvement, you can consult the documents listed in the reference materials.) It can be seen that through the first and second types of improvements, PHP is gradually improving itself into an object-oriented programming language. language.

But this is where the problem arises:

On the positive side, languages with object-oriented features are more welcome in today’s programming world (even if the language itself is not built on object-oriented On top of this, the language can also be prevented from falling behind the trend by adding defined objects, etc.) - In this sense, Zend Engine 2.0 changes PHP's support for object-oriented from the current tentative contact. For full support in the future, it seems to be more in line with the trend of programming language development; in addition, when building enterprise-level applications (which is what PHP is often criticized for now), it is already a fact to use object-oriented methods for modeling and implementation. standards, and this improvement of PHP may cater to this need and solve the weakness of the language itself in this regard.

On the negative side, there seems to be little point in dedicating a new version of PHP to developers that is more Java-like. The reason why PHP is widely used, in addition to open source code and cross-platform, is probably also an important point to adapt to the simple and fast web programming characteristics of Internet website construction. The extremely short learning time, friendly language style (especially if you are familiar with C) and a large number of extended class library functions are enough to prove its power; but if the next version of such a language is transformed into a similar object-oriented language, Not only will a large number of existing developers be at a loss in a short period of time, but it will also be very detrimental to attracting new developers to join - since there is a language like Java, why bother learning PHP?

The above are some of my own "objective" analyses--the so-called "objective" means that when writing about positive aspects, I disguise myself as a loyal supporter of Zend Engine 2.0; when writing about negative aspects, Quite the opposite (I hope you can feel objective after reading my analysis). However, "objective" actually just pave the way for my "subjective" argument -

What do we expect PHP to be like in the future?

In fact, the key to the problem may be in which direction we expect PHP to develop, or which field PHP will focus on.

The reason why PHP is so popular, especially when there is almost no commercial support (Zend currently provides support for PHP, but its strength is really negligible compared with Microsoft and Sun.) It has become a programming language that can compete with ASP and JSP because it is completely oriented to the reality of fast web programming environment. This situation can be clearly seen in many PHP tutorials - to implement some common web functions, using PHP often means less code and complexity; at the same time, for programming in a specific field, PHP also has extensions The functions of the module are available for use (although there is no commercial support, PHP has the support of many advocates of open source software. They not only contributed to the generation and development of PHP, but also provided various extension module functions.), so that it is very convenient for developers. In other words, more ready-to-use and free function libraries mean less work. All that is required is to consult the function manual and use them properly. PHP's competitors appear to be a bit "pedantic" - not flexible enough but more rigorous. Perhaps the products of large commercial companies are indeed not as easy to use as the masterpieces of hackers.

However, in order to adapt to the real fast web programming environment, PHP gave up some things when focusing on convenience and ease of use: such as the writing and packaging of developers' own modules (non-source code level), and object-oriented features (The existing version only incompletely supports some features) and so on - what these PHP lacks is what its competitors have (for example, ASP can communicate with COM components, and JSP can easily use Java Beans; JSP was born out of Java, a completely object-oriented language, etc.), so PHP is usually excluded from the candidate list for building enterprise-level applications.

The core of the problem has become relatively clear - the current development of PHP has encountered a fork in the road: should we continue to give full play to the characteristics of our fast web programming language, or should we improve and modify ourselves into a serious language? What programming language is needed in a business environment? 1

Looking at the future of PHP from the design blueprint of Zend Engine 2.0 (1)_PHP Tutorial