Related
I have been compiling diagrams (pun intended) in hope of understanding the different implementations of common programming languages. I understand whether code is compiled or interpreted depends on the implementation of the code, and is not an aspect of the programming language itself.
I am interested in comparing Python interpretation with direct compilation (ex of C++)
and the virtual machine model (ex Java or C#)
In light of these two diagrams above, could you please help me develop a similar flowchart of how the .py file is converted to .pyc, uses the standard libraries (I gather they are called modules) and then actually run. Many programmers on SO indicate that python as a scripting language is not executed by the CPU but rather the interpreter, but that sounds quite impossible because ultimately hardware must be doing the computation.
First off, this is an implementation detail. I am limiting my answer to CPython and PyPy because I am familiar with them. Answers for Jython, IronPython, and other implementations will differ - probably radically.
Python is closer to the "virtual machine model". Python code is, contrary to the statements of some too-loud-for-their-level-of-knowledge people and despite everyone (including me) conflating it in casual discussion, never interpreted. It is always compiled to bytecode (again, on CPython and PyPy) when it is loaded. If it was loaded because a module was imported and was loaded from a .py file, a .pyc file may be created to cache the compilation output. This step is not mandatory; you can turn it off via various means, and program execution is not affected the tiniest bit (except that the next process to load the module has to do it again). However, the compilation to bytecode is not avoidable, the bytecode is generated in memory if it is not loaded from disk.
This bytecode (the exact details of which are an implementation detail and differ between versions) is then executed, at module level, which entails building function objects, class objects, and the like. These objects simply reuse (hold a pointer to) the bytecode which is already in memory. This is unlike C++ and Java, where code and classes are set in stone during/after compilation. During execution, import statements may be encountered. I lack the space, time and understanding to describe the import machinery, but the short story is:
If it was already imported once, you get that module object (another runtime construct for a thing static languages only have at compile time). A couple of builtin modules (well, all of them in PyPy, for reasons beyond the scope of this question) are already imported before any Python code runs, simply because they are so tightly integrated with the core of the interpreter and so fundamental. sys is such a module. Some Python code may also run beforehand, especially when you start the interactive interpreter (look up site.py).
Otherwise, the module is located. The rules for this are not our concern. In the end, these rules arrive at either a Python file or a dynamically-linked piece of machine code (.DLL on Windows, though Python modules specifically use the extension .pyd but that's just a name; on unix the equivalent .so is used).
The module is first loaded into memory (loaded dynamically, or parsed and compiled to bytecode).
Then, the module is initialized. Extension modules have a special function for that which is called. Python modules are simply run, from top to bottom. In well-behaved modules this just sets up global data, defines functions and classes, and imports dependencies. Of course, anything else can also happen. The resulting module object is cached (remember step one) and returned.
All of this applies to standard library modules as well as third party modules. That's also why you can get a confusing error message if you call a script of yours just like a standard library module which you import in that script (it imports itself, albeit without crashing due to caching - one of many things I glossed over).
How the bytecode is executed (the last part of your question) differs. CPython simply interprets it, but as you correctly note, that doesn't mean it magically doesn't use the CPU. Instead, there is a large ugly loop which detects what bytecode instruction shall be executed next, and then jumps to some native code which carries out the semantics of that instruction. PyPy is more interesting; it starts off interpreting but records some stats along the way. When it decides it's worth doing so, it starts recording what the interpreter does in detail, and generates some highly optimized native code. The interpreter is still used for other parts of the Python code. Note that it's the same with many JVMs and possibly .NET, but the diagram you cite glosses over that.
For the reference implementation of python:
(.py) -> python (checks for .pyc) -> (.pyc) -> python (execution dynamically loads modules)
There are other implementations. Most notable are:
jython which compiles (.py) to (.class) and follows the java pattern from there
pypy which employs a JIT as it compiles (.py). the chain from there could vary (pypy could be run in cpython, jython or .net environments)
Python is technically a scripted language but it is also compiled, python source is taken from its source file and fed into the interpreter which often compiles the source to bytecode either internally and then throws it away or externally and saves it like a .pyc
Yes python is a single virtual machine that then sits ontop of the actual hardware but all python bytecode is, is a series of instructions for the pvm (python virtual machine) much like assembler for the actual CPU.
Why should I use VM, like Parrot, for a dynamic language I use (Python, Perl, ...) if I already have an interpreter? What can I potentially gain, for the cost of having different VM between my code and my machine, and by using a separate interpreter?
(I am new in VM issue, so maybe the answer is obvious)
EDIT
What's the benefit of Parrot VM for end-users?
Why should I use VM, like Parrot, for a dynamic language I use (Python, Perl, ...) if I already have an interpreter?
First, if you're starting a project, then you may not already have an interpreter.
However, assuming you have an interpreter and are considering whether to add functionality to it or rewrite it to use Parrot, the tradeoffs that come to mind are:
In general, Parrot is probably better tested than the interpreter in question (better optimizer, better garbage collector, etc.)
In general, Parrot's developers know more about cross-platform issues than run-of-the-mill programmers
In general, Parrot has solved most problems you're likely to run into
Parrot was designed with complete generality in mind, and that added a ton of complexity; you may not need the extra generality
Personally, Parrot's optimizer (and register-based design, largely to make optimizations easier) and well tested cross platform codebase would be enough to convince me.
Parsing the ASCII source code is slow. It is faster if the source file gets parsed once, and then the interpreter uses a binary structure. In Python this structure gets stored in .pyc files for fast reuse.
There are two steps:
Parse the source, create byte code
Run (interpret) the byte code.
This is used by e.g. scala: There is no scala-VM. Scala is just a new syntax. The scala compiler creates java byte code.
If I write a python script, anyone can simply point an editor to it and read it. But for programming written in C, one would have to use decompilers and hex tables and such. Why is that? I mean I simply can't open up the Safari web browser and look at its code.
Note: The author disavows a deep expertise in this subject. Some assertions may be incorrect.
Python actually is compiled into bytecode, which is what gets run by the python interpreter. Whenever you use a Python module, Python will generate a .pyc file with a name corresponding to the module. This is the equivalent of the .o file that's generated when you compile a C file.
So if you want something to disassemble, the .pyc file would be it :)
The process that Python goes through when compiling a module is pretty similar to what gcc or another C compiler does with C source code. The major difference is that it happens transparently as part of execution of the file. It's also optional: when running a non-module, i.e. an end-user script, Python will just interpret the code rather than compiling it first.
So really your question is "Why are python programs distributed as source rather than as compiled modules?" Or, put another way, "Why are C applications distributed as compiled binaries rather than as source code?"
It used to be very common for C applications to be distributed as source code. This was back before operating systems and their various subentities (i.e. linux distributions) became more established. Some distros, for example gentoo, still distribute apps as source code. Apps which are a bit more cutting edge or obscure are still distributed as source code for all platforms they target.
The reason for this is compatibility, and dependencies. The reason you can run the precompiled binary Safari on a Mac, or Firefox on Ubuntu Linux, is because it's been specifically built for that operating system, architecture (e.g. x86_64), and set of libraries.
Unfortunately, compilation of a large app is pretty slow, and needs to be redone at least partially every time the app is updated. Thus the motivation for binary distributions.
So why not create a binary distribution of Python? For one thing, as Aaron mentions, modules would need to be recompiled for each new version of the Python bytecode. But this would be similar to rebuilding a C app to link with a newer version of a dynamic library — Python modules are analogous in this sense to C libraries.
The real reason is that Python compilation is very much quicker than C compilation. This is in part, I think, because of the dynamic nature of the language, and also because it's not as thorough of a compilation. This has its tradeoffs: in particular, Python apps run much more slowly than do their C counterparts, because Python has to interpret the compiled bytecode into instructions for the processor, whereas the C app already contains such instructions.
That all being said, there is a program called py2exe that will take a Python module and distribution and build a precompiled windows executable, including in it the logic of the module and its dependencies, including Python itself. I guess the point of this is to avoid having to coerce people into installing Python on their Windows system just to run your app. Under linux, or I think even OS/X, Python is usually already installed, so precompilation is not really necessary. Linux systems also have super-dandy package managers that will transparently install dependencies such as Python if they are not already installed.
Python is a script language, runs in a virtual machine through an interpeter.
C is a compiled language, the code compiled to binary code which the computer can run without all that extra stuff Python needs.
This is sorta a big topic. You should look into your local friendly Computer Science curriculum, you'll find a lot of great stuff on this subject there.
The short answer is the Python is an "interpreted" language, which means that it requires a machine language program (the python interpreter) to run the python program, adding a layer of indirection. C or C++ are different. They are compiled directly to machine code, which runs directly on your processor.
There is a lot of additional voodoo to be learned here, however. Technically Python is compiled to a bytecode, and modern interpreters do more and more "Just in Time" compilation, so the boundaries between compiled and interpreted code are getting fuzzier all the time.
In several comments you asked: "Is it then possible to compile python to an executable binary file and then simply distribute that?"
From a theoretical viewpoint, there's no question the answer is yes -- a Python program could be compiled to, and distributed as, fully compiled machine code.
From a practical viewpoint, it's open to a lot more question. There are a few things like Unladen Swallow, Psyco, Shed Skin, and PyPy that you might want to know about though.
Unladen Swallow is primarily an attempt at making Python run faster, but part of the plan to do so involves using LLVM for its back-end. LLVM can (among other things) produce native machine code output. The last couple of releases of Unladen Swallow have used LLVM for native code generation, but 1) the most recent update on the web site is from late 2009, and 2) the release notes for that version say: "The Unladen Swallow team does not recommend wide adoption of the 2009Q3 release."
Psyco works as a plug-in for Python that basically does JIT compilation, so even though it can speed up execution (quite a lot in some cases), it doesn't produce a machine-code executable you can distribute. In short, while it's sort of similar to what you want, it's not intended to do exactly what you've asked for.
Shed Skin Python-to-C++ produces C++ as its output, and you then compile the C++ and (potentially) distribute the result of that. Shedskin is currently at version 0.5 -- i.e., nobody's claiming that it's a finished, released product. On the other hand, development is ongoing, and each release does seem to include pretty substantial improvements.
PyPy is a Python implementation written in Python. Their intent is to allow code production to be "plugged in" without affecting the rest of the implementation -- but while they currently support 4 different code generation models, I don't believe any of them results in producing native machine code that runs directly on the hardware.
Bottom line: work has been done and is being done with the intent of doing what you asked about, but at least to my knowledge there's not really anything I could reasonably recommend as a finished product that you can really depend on to do the job right now. The primary emphasis is really on execution speed, not producing standalone executables.
Yes, you can - it's called disassembling, and allows you to look at the code of Safari perfectly well. The thing is, C, among other languages, compiles to native code, i.e. code that your CPU can "understand" and execute.
More or less obviously, the level of abstraction present in the instruction set of your CPU is much smaller than that of a high level language like Python. The CPU instructions are not concerned with "downloading that URI", but more "check if that bit is set in a hardware register".
So, in conclusion, the level of complexity present in a native application is much higher when looking at the machine code, so many people simply can't make any sense of what is going on there, it's hard to get the big picture. With experience and time at your hands, it is possible though - people do it all the time, reversing applications and all.
you can't open up and read the code that actually runs for python either. Try
import dis
def foo():
for i in range(100):
print i
print dis.dis(foo)
That will show you the (human readable) bytcode of the foo program. equivalently, you can save the file and import it from the interactive python interpreter. This will create a .pyc file with the same basename as the script. open that with a hex editor and you are looking at the actually python bytecode.
The reason for the difference is that python changes up it's byte code between releases so that you would either need to distribute a different version of a binary only release for each version of python. This would be a pain.
With C, it's compiled to native code and so the byte code is much more stable making binary only releases possible.
because C code is complied to object (machine) code and python code is compiled into an intermediate byte code. I am not sure if you are even referring to the byte code of python - you must be referring to the source file itself which is directly executable (hiding the byte code from you!). C needs to be compiled and linked.
Python scripts are parsed and converted to binary only when they're run - i.e., they're text files and you can read them with an editor.
C code is compiled and linked to an executable binary file before they can be run. Normally, only this executable binary file is distributed - hence you need a decompiler. You can always view the source code, if you've access to it.
Not all C programs require decompilers. There's lots of C code distributed in source form. And some Python programs do require decompilers, if distributed as bytecode (.pyc files).
But, to the extent that your assumptions are valid, it's because C is a compiled language while Python is an interpreted language.
Python scripts are analogous to a man looking at a to-do list written in English (or language he understands). The man has to do all the work, every time that list of things has to be done.
If the man, instead of doing the steps on his own each time, creates and programs a robot which can carry out those steps again and again (and probably faster than him), that robot is analogous to the C program.
The man in the python case is called the "interpreter" and in the C case is called the "compiler", and the C robot is called the compiled program/executable.
When you look at the python program source, you see the to-do list. In case of the robot, you see the gears, motors and batteries, etc, which look very different from the to-do list. If you could get hold of the C "to-do" list, it looks somewhat like the python code, just in a different language.
G-WAN executes ANSI C scripts on the fly -making it just like Python scripts.
This can be server-side scripts (using G-WAN as a Web server) or any general-purpose C program and you can link any existing library.
Oh, and G-WAN C scripts are much faster than Python, PHP or Java...
Python is an interpreted language. But why does my source directory contain .pyc files, which are identified by Windows as "Compiled Python Files"?
I've been given to understand that
Python is an interpreted language...
This popular meme is incorrect, or, rather, constructed upon a misunderstanding of (natural) language levels: a similar mistake would be to say "the Bible is a hardcover book". Let me explain that simile...
"The Bible" is "a book" in the sense of being a class of (actual, physical objects identified as) books; the books identified as "copies of the Bible" are supposed to have something fundamental in common (the contents, although even those can be in different languages, with different acceptable translations, levels of footnotes and other annotations) -- however, those books are perfectly well allowed to differ in a myriad of aspects that are not considered fundamental -- kind of binding, color of binding, font(s) used in the printing, illustrations if any, wide writable margins or not, numbers and kinds of builtin bookmarks, and so on, and so forth.
It's quite possible that a typical printing of the Bible would indeed be in hardcover binding -- after all, it's a book that's typically meant to be read over and over, bookmarked at several places, thumbed through looking for given chapter-and-verse pointers, etc, etc, and a good hardcover binding can make a given copy last longer under such use. However, these are mundane (practical) issues that cannot be used to determine whether a given actual book object is a copy of the Bible or not: paperback printings are perfectly possible!
Similarly, Python is "a language" in the sense of defining a class of language implementations which must all be similar in some fundamental respects (syntax, most semantics except those parts of those where they're explicitly allowed to differ) but are fully allowed to differ in just about every "implementation" detail -- including how they deal with the source files they're given, whether they compile the sources to some lower level forms (and, if so, which form -- and whether they save such compiled forms, to disk or elsewhere), how they execute said forms, and so forth.
The classical implementation, CPython, is often called just "Python" for short -- but it's just one of several production-quality implementations, side by side with Microsoft's IronPython (which compiles to CLR codes, i.e., ".NET"), Jython (which compiles to JVM codes), PyPy (which is written in Python itself and can compile to a huge variety of "back-end" forms including "just-in-time" generated machine language). They're all Python (=="implementations of the Python language") just like many superficially different book objects can all be Bibles (=="copies of The Bible").
If you're interested in CPython specifically: it compiles the source files into a Python-specific lower-level form (known as "bytecode"), does so automatically when needed (when there is no bytecode file corresponding to a source file, or the bytecode file is older than the source or compiled by a different Python version), usually saves the bytecode files to disk (to avoid recompiling them in the future). OTOH IronPython will typically compile to CLR codes (saving them to disk or not, depending) and Jython to JVM codes (saving them to disk or not -- it will use the .class extension if it does save them).
These lower level forms are then executed by appropriate "virtual machines" also known as "interpreters" -- the CPython VM, the .Net runtime, the Java VM (aka JVM), as appropriate.
So, in this sense (what do typical implementations do), Python is an "interpreted language" if and only if C# and Java are: all of them have a typical implementation strategy of producing bytecode first, then executing it via a VM/interpreter.
More likely the focus is on how "heavy", slow, and high-ceremony the compilation process is. CPython is designed to compile as fast as possible, as lightweight as possible, with as little ceremony as feasible -- the compiler does very little error checking and optimization, so it can run fast and in small amounts of memory, which in turns lets it be run automatically and transparently whenever needed, without the user even needing to be aware that there is a compilation going on, most of the time. Java and C# typically accept more work during compilation (and therefore don't perform automatic compilation) in order to check errors more thoroughly and perform more optimizations. It's a continuum of gray scales, not a black or white situation, and it would be utterly arbitrary to put a threshold at some given level and say that only above that level you call it "compilation"!-)
They contain byte code, which is what the Python interpreter compiles the source to. This code is then executed by Python's virtual machine.
Python's documentation explains the definition like this:
Python is an interpreted language, as
opposed to a compiled one, though the
distinction can be blurry because of
the presence of the bytecode compiler.
This means that source files can be
run directly without explicitly
creating an executable which is then
run.
There is no such thing as an interpreted language. Whether an interpreter or a compiler is used is purely a trait of the implementation and has absolutely nothing whatsoever to do with the language.
Every language can be implemented by either an interpreter or a compiler. The vast majority of languages have at least one implementation of each type. (For example, there are interpreters for C and C++ and there are compilers for JavaScript, PHP, Perl, Python and Ruby.) Besides, the majority of modern language implementations actually combine both an interpreter and a compiler (or even multiple compilers).
A language is just a set of abstract mathematical rules. An interpreter is one of several concrete implementation strategies for a language. Those two live on completely different abstraction levels. If English were a typed language, the term "interpreted language" would be a type error. The statement "Python is an interpreted language" is not just false (because being false would imply that the statement even makes sense, even if it is wrong), it just plain doesn't make sense, because a language can never be defined as "interpreted."
In particular, if you look at the currently existing Python implementations, these are the implementation strategies they are using:
IronPython: compiles to DLR trees which the DLR then compiles to CIL bytecode. What happens to the CIL bytecode depends upon which CLI VES you are running on, but Microsoft .NET, GNU Portable.NET and Novell Mono will eventually compile it to native machine code.
Jython: interprets Python sourcecode until it identifies the hot code paths, which it then compiles to JVML bytecode. What happens to the JVML bytecode depends upon which JVM you are running on. Maxine will directly compile it to un-optimized native code until it identifies the hot code paths, which it then recompiles to optimized native code. HotSpot will first interpret the JVML bytecode and then eventually compile the hot code paths to optimized machine code.
PyPy: compiles to PyPy bytecode, which then gets interpreted by the PyPy VM until it identifies the hot code paths which it then compiles into native code, JVML bytecode or CIL bytecode depending on which platform you are running on.
CPython: compiles to CPython bytecode which it then interprets.
Stackless Python: compiles to CPython bytecode which it then interprets.
Unladen Swallow: compiles to CPython bytecode which it then interprets until it identifies the hot code paths which it then compiles to LLVM IR which the LLVM compiler then compiles to native machine code.
Cython: compiles Python code to portable C code, which is then compiled with a standard C compiler
Nuitka: compiles Python code to machine-dependent C++ code, which is then compiled with a standard C compiler
You might notice that every single one of the implementations in that list (plus some others I didn't mention, like tinypy, Shedskin or Psyco) has a compiler. In fact, as far as I know, there is currently no Python implementation which is purely interpreted, there is no such implementation planned and there never has been such an implementation.
Not only does the term "interpreted language" not make sense, even if you interpret it as meaning "language with interpreted implementation", it is clearly not true. Whoever told you that, obviously doesn't know what he is talking about.
In particular, the .pyc files you are seeing are cached bytecode files produced by CPython, Stackless Python or Unladen Swallow.
These are created by the Python interpreter when a .py file is imported, and they contain the "compiled bytecode" of the imported module/program, the idea being that the "translation" from source code to bytecode (which only needs to be done once) can be skipped on subsequent imports if the .pyc is newer than the corresponding .py file, thus speeding startup a little. But it's still interpreted.
To speed up loading modules, Python caches the compiled content of modules in .pyc.
CPython compiles its source code into "byte code", and for performance reasons, it caches this byte code on the file system whenever the source file has changes. This makes loading of Python modules much faster because the compilation phase can be bypassed. When your source file is foo.py , CPython caches the byte code in a foo.pyc file right next to the source.
In python3, Python's import machinery is extended to write and search for byte code cache files in a single directory inside every Python package directory. This directory will be called __pycache__ .
Here is a flow chart describing how modules are loaded:
For more information:
ref:PEP3147
ref:“Compiled” Python files
THIS IS FOR BEGINNERS,
Python automatically compiles your script to compiled code, so called byte code, before running it.
Running a script is not considered an import and no .pyc will be created.
For example, if you have a script file abc.py that imports another module xyz.py, when you run abc.py, xyz.pyc will be created since xyz is imported, but no abc.pyc file will be created since abc.py isn’t being imported.
If you need to create a .pyc file for a module that is not imported, you can use the py_compile and compileall modules.
The py_compile module can manually compile any module. One way is to use the py_compile.compile function in that module interactively:
>>> import py_compile
>>> py_compile.compile('abc.py')
This will write the .pyc to the same location as abc.py (you can override that with the optional parameter cfile).
You can also automatically compile all files in a directory or directories using the compileall module.
python -m compileall
If the directory name (the current directory in this example) is omitted, the module compiles everything found on sys.path
Python (at least the most common implementation of it) follows a pattern of compiling the original source to byte codes, then interpreting the byte codes on a virtual machine. This means (again, the most common implementation) is neither a pure interpreter nor a pure compiler.
The other side of this is, however, that the compilation process is mostly hidden -- the .pyc files are basically treated like a cache; they speed things up, but you normally don't have to be aware of them at all. It automatically invalidates and re-loads them (re-compiles the source code) when necessary based on file time/date stamps.
About the only time I've seen a problem with this was when a compiled bytecode file somehow got a timestamp well into the future, which meant it always looked newer than the source file. Since it looked newer, the source file was never recompiled, so no matter what changes you made, they were ignored...
Python's *.py file is just a text file in which you write some lines of code. When you try to execute this file using say "python filename.py"
This command invokes Python Virtual Machine. Python Virtual Machine has 2 components: "compiler" and "interpreter". Interpreter cannot directly read the text in *.py file, so this text is first converted into a byte code which is targeted to the PVM (not hardware but PVM). PVM executes this byte code. *.pyc file is also generated, as part of running it which performs your import operation on file in shell or in some other file.
If this *.pyc file is already generated then every next time you run/execute your *.py file, system directly loads your *.pyc file which won't need any compilation(This will save you some machine cycles of processor).
Once the *.pyc file is generated, there is no need of *.py file, unless you edit it.
tldr; it's a converted code from the source code, which the python VM interprets for execution.
Bottom-up understanding: the final stage of any program is to run/execute the program's instructions on the hardware/machine. So here are the stages preceding execution:
Executing/running on CPU
Converting bytecode to machine code.
Machine code is the final stage of conversion.
Instructions to be executed on CPU are given in machine code. Machine code can be executed directly by CPU.
Converting Bytecode to machine code.
Bytecode is a medium stage. It could be skipped for efficiency, but sacrificing portability.
Converting Source code to bytecode.
Source code is a human readable code. This is what is used when working on IDEs (code editors) such as Pycharm.
Now the actual plot. There are two approaches when carrying any of these stages: convert [or execute] a code all at once (aka compile) and convert [or execute] the code line by line (aka interpret).
For example, we could compile a source code to bytecode, compile bytecode to machine code, interpret machine code for execution.
Some implementations of languages skip stage 3 for efficiency, i.e. compile source code into machine code and then interpret machine code for execution.
Some implementations skip all middle steps and interpret the source code directly for execution.
Modern languages often involve both compiling an interpreting.
JAVA for example, compiles source code to bytecode [that is how JAVA source is stored, as a bytecode, compile bytecode to machine code [using JVM], and interpret machine code for execution. [Thus JVM is implemented differently for different OSs, but the same JAVA source code could be executed on different OS that have JVM installed.]
Python for example, compile source code to bytecode [usually found as .pyc files accompanying the .py source codes], compile bytecode to machine code [done by a virtual machine such as PVM and the result is an executable file], interpret the machine code/executable for execution.
When can we say that a language is interpreted or compiled?
The answer is by looking into the approach used in execution. If it executes the machine code all at once (== compile), then it's a compiled language. On the other hand, if it executes the machine code line-by-line (==interpret) then it's an interpreted language.
Therefore, JAVA and Python are interpreted languages.
A confusion might occur because of the third stage, that's converting bytecode to machine code. Often this is done using a software called a virtual machine. The confusion occurs because a virtual machine acts like a machine, but it's actually not! Virtual machines are introduced for portability, having a VM on any REAL machine will allow us to execute the same source code. The approach used in most VMs [that's the third stage] is compiling, thus some people would say it's a compiled language. For the importance of VMs, we often say that such languages are both compiled and interpreted.
Python code goes through 2 stages. First step compiles the code into .pyc files which is actually a bytecode. Then this .pyc file(bytecode) is interpreted using CPython interpreter. Please refer to this link. Here process of code compilation and execution is explained in easy terms.
Its important distinguish language specification from language implementations:
Language specification is just a document with the formal specification of the language, with its context free grammar and definition of the semantic rules (like specifying primitive types and scope dynamics).
Language implementation is just a program (a compiler) that implement the use of the language according to its specification.
Any compiler consists of two independent parts: a frontend and backend. The frontend receives the source code, validate it and translate it into an intermediate code. After that, a backend translate it to machine code to run in a physical or a virtual machine.
An interpreter is a compiler, but in this case it can produce a way of executing the intermediate code directly in a virtual machine.
To execute python code, its necessary transform the code in a intermediate code, after that the code is then "assembled" as bytecode that can be stored in a file.pyc, so no need to compile modules of a program every time you run it.
You can view this assembled python code using:
from dis import dis
def a(): pass
dis(a)
Anyone can build a Compiler to static binary in Python language, as can build an interpreter to C language. There are tools (lex/yacc) to simplify and automate the proccess of building a compiler.
Machines don't understand English or any other languages, they understand only byte code, which they have to be compiled (e.g., C/C++, Java) or interpreted (e.g., Ruby, Python), the .pyc is a cached version of the byte code.
https://www.geeksforgeeks.org/difference-between-compiled-and-interpreted-language/
Here is a quick read on what is the difference between compiled language vs interpreted language, TLDR is interpreted language does not require you to compile all the code before run time and thus most of the time they are not strict on typing etc.
From the Google Open Source Blog:
PyPy is a reimplementation of Python
in Python, using advanced techniques
to try to attain better performance
than CPython. Many years of hard work
have finally paid off. Our speed
results often beat CPython, ranging
from being slightly slower, to
speedups of up to 2x on real
application code, to speedups of up to
10x on small benchmarks.
How is this possible? Which Python implementation was used to implement PyPy? CPython? And what are the chances of a PyPyPy or PyPyPyPy beating their score?
(On a related note... why would anyone try something like this?)
"PyPy is a reimplementation of Python in Python" is a rather misleading way to describe PyPy, IMHO, although it's technically true.
There are two major parts of PyPy.
The translation framework
The interpreter
The translation framework is a compiler. It compiles RPython code down to C (or other targets), automatically adding in aspects such as garbage collection and a JIT compiler. It cannot handle arbitrary Python code, only RPython.
RPython is a subset of normal Python; all RPython code is Python code, but not the other way around. There is no formal definition of RPython, because RPython is basically just "the subset of Python that can be translated by PyPy's translation framework". But in order to be translated, RPython code has to be statically typed (the types are inferred, you don't declare them, but it's still strictly one type per variable), and you can't do things like declaring/modifying functions/classes at runtime either.
The interpreter then is a normal Python interpreter written in RPython.
Because RPython code is normal Python code, you can run it on any Python interpreter. But none of PyPy's speed claims come from running it that way; this is just for a rapid test cycle, because translating the interpreter takes a long time.
With that understood, it should be immediately obvious that speculations about PyPyPy or PyPyPyPy don't actually make any sense. You have an interpreter written in RPython. You translate it to C code that executes Python quickly. There the process stops; there's no more RPython to speed up by processing it again.
So "How is it possible for PyPy to be faster than CPython" also becomes fairly obvious. PyPy has a better implementation, including a JIT compiler (it's generally not quite as fast without the JIT compiler, I believe, which means PyPy is only faster for programs susceptible to JIT-compilation). CPython was never designed to be a highly optimising implementation of the Python language (though they do try to make it a highly optimised implementation, if you follow the difference).
The really innovative bit of the PyPy project is that they don't write sophisticated GC schemes or JIT compilers by hand. They write the interpreter relatively straightforwardly in RPython, and for all RPython is lower level than Python it's still an object-oriented garbage collected language, much more high level than C. Then the translation framework automatically adds things like GC and JIT. So the translation framework is a huge effort, but it applies equally well to the PyPy python interpreter however they change their implementation, allowing for much more freedom in experimentation to improve performance (without worrying about introducing GC bugs or updating the JIT compiler to cope with the changes). It also means when they get around to implementing a Python3 interpreter, it will automatically get the same benefits. And any other interpreters written with the PyPy framework (of which there are a number at varying stages of polish). And all interpreters using the PyPy framework automatically support all platforms supported by the framework.
So the true benefit of the PyPy project is to separate out (as much as possible) all the parts of implementing an efficient platform-independent interpreter for a dynamic language. And then come up with one good implementation of them in one place, that can be re-used across many interpreters. That's not an immediate win like "my Python program runs faster now", but it's a great prospect for the future.
And it can run your Python program faster (maybe).
Q1. How is this possible?
Manual memory management (which is what CPython does with its counting) can be slower than automatic management in some cases.
Limitations in the implementation of the CPython interpreter preclude certain optimisations that PyPy can do (eg. fine grained locks).
As Marcelo mentioned, the JIT. Being able to on the fly confirm the type of an object can save you the need to do multiple pointer dereferences to finally arrive at the method you want to call.
Q2. Which Python implementation was used to implement PyPy?
The PyPy interpreter is implemented in RPython which is a statically typed subset of Python (the language and not the CPython interpreter). - Refer https://pypy.readthedocs.org/en/latest/architecture.html for details.
Q3. And what are the chances of a PyPyPy or PyPyPyPy beating their score?
That would depend on the implementation of these hypothetical interpreters. If one of them for example took the source, did some kind of analysis on it and converted it directly into tight target specific assembly code after running for a while, I imagine it would be quite faster than CPython.
Update: Recently, on a carefully crafted example, PyPy outperformed a similar C program compiled with gcc -O3. It's a contrived case but does exhibit some ideas.
Q4. Why would anyone try something like this?
From the official site. https://pypy.readthedocs.org/en/latest/architecture.html#mission-statement
We aim to provide:
a common translation and support framework for producing
implementations of dynamic languages, emphasizing a clean
separation between language specification and implementation
aspects. We call this the RPython toolchain_.
a compliant, flexible and fast implementation of the Python_
Language which uses the above toolchain to enable new advanced
high-level features without having to encode the low-level
details.
By separating concerns in this way, our implementation of Python - and
other dynamic languages - is able to automatically generate a
Just-in-Time compiler for any dynamic language. It also allows a
mix-and-match approach to implementation decisions, including many
that have historically been outside of a user's control, such as
target platform, memory and threading models, garbage collection
strategies, and optimizations applied, including whether or not to
have a JIT in the first place.
The C compiler gcc is implemented in C, The Haskell compiler GHC is written in Haskell. Do you have any reason for the Python interpreter/compiler to not be written in Python?
PyPy is implemented in Python, but it implements a JIT compiler to generate native code on the fly.
The reason to implement PyPy on top of Python is probably that it is simply a very productive language, especially since the JIT compiler makes the host language's performance somewhat irrelevant.
PyPy is written in Restricted Python. It does not run on top of the CPython interpreter, as far as I know. Restricted Python is a subset of the Python language. AFAIK, the PyPy interpreter is compiled to machine code, so when installed it does not utilize a python interpreter at runtime.
Your question seems to expect the PyPy interpreter is running on top of CPython while executing code.
Edit: Yes, to use PyPy you first translate the PyPy python code, either to C and build with gcc, to jvm byte code, or to .Net CLI code. See Getting Started