Php internal architecture

Editor's Notes

• #2: PHP is an ever evolving piece of software that is useful and easily hacked without touching the core The features in PHP that make it so pluggable are not unique to PHP, other languages use them as well Understanding how PHP is structured and works in comparison to other languages can help you understand what your code is doing You don’t have to know C to see what’s going on “beneath the hood”
• #3: PHP has a pretty standard architecture for an interpreted language actually. We’ll take a look at how it compares to other similar languages By similar I mean – no compilation step So Ruby, Python, and NodeJS We’re not going to compare apples to oranges here – what use would that be? So don’t get me started on compiled languages (yes, that includes JAVA, C#, go and Rust, sorry)
• #4: It’s very important to understand the “big picture” of how your system fits together. You don’t have to know exactly how the parser works, but knowing you have a parser is important. It’s not that important to know how modules get loaded, but you should understand they exist and are loadable A lot of people make excuses that “you don’t know need to know how an internal combustion engine works to drive a car” Well, that’s bologna. I would hope you know that when you turn the key, it uses electricity in your battery (which needs to be working) to ignite the gasoline (which you also need enough of) in your car – you need to know the rules of the road and how to use the break No, you don’t need to know what the right mix ratio is for fuel to air in your cylinder But by George you better know that spark plugs make your engine go and they need to be replaced!
• #5: So what do all these pieces do? The sapi and the streams do input and output control, no matter what we’re using for input or for output The engine actually processes our PHP files – doing the work Extensions add functionality to PHP (if you ever want an adventure – don’t just do disable-all, try forcing all extensions off by fiddling with the .m4 files for date, pcre, standard, tokenizer, and a few others) You will have almost NO functionality in PHP itself Streams are a bit broken because although parts of them are integrated into PHP’s sapi layer, other parts are integrated into the engine and still other parts into the standard extension (ugh)
• #6: PHP is designed to be basic – the sapi says “run this script” and the engine runs it and returns output via the sapi That’s it The sapi gets to decide how things are handled There are ways of accessing input and output aside from our main sapi process, but they’re started from the main process They don’t exist independantly
• #7: PHP gets away with mod_php because of it’s concept of shared nothing Do you notice a trend yet? The rack specification and wsgi have a LOT in common They’re weirdly (and purposely) very similar to PSR-7
• #8: Ok, at this point things are getting a wee bit ridiculous Do you see the pattern? Yes, that’s right. There is no real difference between any of these things That’ s because they’re all solving the exact same issue
• #9: A lot of this has been reinterated for years, but these are the reasons that PHP IS different What’s interesting to me that these are often cited as DOWNSIDES to php, and the last one? Half the ruby and python developers I talk to do NOT understand what I’m talking about Problem is most of the PHP folks don’t either 
• #10: Why is this important? This IS The thing The choices you make depend on what you are doing! Do you want easy web dev and deployment? Do you want good horizontal scalability? Then PHP is your answer Do you have another question: well that’s your problem
• #11: PHP is designed in a way so that not only do we have a pluggable infrastructure for adding functionality, we have a pluggable infrastructure for speaking input and output This is usually the number one question people are asked when they have a bug … and half can’t answer SAPIS are (one of) the most underutilized features in the system, basically because few people know they exist, fewer can use them And their API is not as robust as it could be because it hasn’t HAD to be
• #13: So cgi has a lot of benefits actually Except for the fact that it’s slow as dirt Or molasses in july
• #14: Make sure to talk about what an mpm is in apache – a multi-processing model
• #15: Fastcgi really does work, and work well, mostly as advertised
• #16: Basically smacks a process manager on top of the fastcgi protocol, this is not necessarily needed for webservers like IIS which have features like worker pools, but for Nginx and apache raw fastcgi is … not fabulous. This adds a bunch of management features
• #17: Yes, the PHP cli is a sapi 
• #22: Yeah, so PHP actually has an http stream and code to help write a server in it that takes care of a LOT of hard stuff for you, just in case you really want to write your own appserver and then reverse proxy into it ;) IF YOU’RE CRAZY
• #23: Streams are pretty much the most awesomesauce part of the package
• #24: Streams are a huge underlying component of PHP Streams were introduced with PHP 4.3.0 – they are old, but underuse means they can have rough edges… so TEST TEST TEST But they are more powerful then almost anything else you can use Why is this better ? Lots and lots of data in small chunks lets you do large volumes without maxing out memory and cpu
• #25: All input and output comes into PHP It gets pushed through a streams filter Then through the streams wrapper During this point the stream context is available for the filter and wrapper to use Streams themselves are the “objects” coming in Wrappers are the “classes” defining how to deal with the stream
• #26: What is streamable behavorior? We’ll get to that in a bit Protocol: set of rules which is used by computers to communicate with each other across a network Resource: A resource is a special variable, holding a reference to an external resource Talk about resources in PHP and talk about general protocols, get a list from the audience of protocols they can name (yes http is a protocol) A socket is a special type of stream – pound this into their heads A socket is an endpoint of communication to which a name can be bound. A socket has a type and one associated process. Sockets were designed to implement the client-server model for interprocess communication where: In php , a wrapper ties the stream to the transport – so your http wrapper ties your PHP data to the http transport and tells it how to behave when reading and writing data
• #27: Internet Domain sockets expect a port number in addition to a target address. In the case of fsockopen() this is specified in a second parameter and therefore does not impact the formatting of transport URL. With stream_socket_client() and related functions as with traditional URLs however, the port number is specified as a suffix of the transport URL delimited by a colon. unix:// provides access to a socket stream connection in the Unix domain. udg:// provides an alternate transport to a Unix domain socket using the user datagram protocol. Unix domain sockets, unlike Internet domain sockets, do not expect a port number. In the case of fsockopen() the portno parameter should be set to 0.
• #30: Lexical Analysis Converts the source from a sequence of characters into a sequence of tokens
• #31: Syntax Analysis Analyzes a sequence of tokens to determine their grammaticalstructure
• #32: 5.6 and 7+
• #33: Generate bytecode based on the information gathered byanalyzing the sourcecode
• #34: so zend is actually a “virtual machine” it interprets OPCODES and does stuff with them reads each opcode and does a specific action – like a giant state machine
• #35: Extensions are the soul of what makes PHP great If you take away the extensions there’s not a lot left in PHP Sadly there isn’t a nice API to go with extensions, and that’s something PHP should address
• #36: PHP IS a glue language, it was not designed, it grew… and it was designed to be Cish and tie TIGHTLY to C without making the poor dev worry about the hard stuff! PHP has probably the sharpest delineation between a library (PHP code) and an extension (C code you have to compile) of any interpreted language out there. It is also arguably the most modular and extensible. This is not necessarily a bad thing, except that because so many people do PHP “on the cheap” they’re absolutely terrified of extensions! If C (or C++) can do it – PHP can be glued to it. The extension architecture introduced was probably one of the BEST decisions ever made for PHP! – And a bit of trivia – it didn’t exist before PHP 3.0
• #37: I lied in the slide here – you can bind to non C/C++ stuff too, there are some clever bindings to .NET managed assemblies and cocoa bindings and all kinds of evil – but those are not common use case and are hard to do extensions have to be compiled for the major.minor version you’re using – and there’s lots of different flags and such that can make binaries incompatible, this is one of the strengths and weaknesses of a compiled language, it’s optimized for what it’s meant to run on but doesn’t work at all elsewhere
• #38: Porting C libraries in pure PHP could be done – but why in the world would you do that!! Unless you’re doing something incredibly stupid – C code is going to be faster than PHP No matter what you do, there are parts of the engine you can’t touch – in userland Extensions do NOT have this limitation, you can do all kinds of evil. Yes there are extensions for threading, debugging, compiling php opcodes, intercepting function calls, profiling, you name it. C is a POWERFUL thing.
• #39: Two types of extensions zend extensions can dig right into the engine itself php module cannot A PHP extension can be both! (see xdebug) So what are “core default” extensions – these are the ones that are distributed with the PHP source code and are turned on by default – there are a few that are “always on” (as of 5.3) – PCRE, date, standard (duh), reflection and spl There are bunch that are generally turned “on by default” – PHP tries to build them even if you don’t flag them as on (not having the libs necessary will turn them off) – these include things like libxml2, zlib, iconv There are still a bunch more included in the core source – some are excellent and some – some suck
• #41: pickle or peh cul (depending on your side of the pond) is the ongoing argument for pronunciation – bottom line is the logo IS a pickle yo ;) Benefits: ▫▫code hosting and distribution (git! and github mirror!) ▫pecl install (PEAR packaging and installer support) ▫community ▫advertising Split off from PEAR in 2003 •WezFurlong is “instigator” ▫see famous letter ▫http://news.php.net/article.php?group=php.pecl.dev&article=5 •Many of the points raised still need work –mainly PECL needs manpower, bodies to do the work
• #45: I get all the time “PHP is so slow” If you’re connecting to a remote memcache server, and a remote redis server, and a remote webservice, and a remote database… even if all that code is fast you’re screwed – scale right, scale truly horizontal with localhost connections for everything you can get away with At the end of the day
• #47: FFI is a cool concept but the number one rule of FFI is… you can totally shoot yourself in the foot with it! The basic idea (and ruby and python use this extensively) is to “wrap” your C stuff with the FFI extension, and then Write python/ruby code to unstupid the APIS
• #49: So many languages support this idea of calling into usually C code Then they usually put a layer of regular code “on top” to make APIS non painful This can be useful it also tends to be slower
• #50: before you get into extension writing – if you just want an ffi wrapper and are just going to call the exact C calls from an existing library why go to the trouble of writing an extension? ffi is pretty great but a bit of a flakey extension yet, but it’s identical to python’s “ctypes” which is a stupid name, it’s really ffi I hear al lthe time about how “great” python is because of ctypes, frankly I beg to differ. Part of wrappign a C extension is translating the C calls into something far more “phpish”
• #51: oh ffi is cool and so needs a maintainer
• #52: Some of my wishlist of how I’d love to see you get involved 
• #53: Would love to do some code for evil on this 