I have lot of .c file which is autogenerated by simulink coder. Analyzing those file to understand simulink model is some what time consuming, As per my knowledge there is no reverse tool which converts c code to simulink model.
so is there any possiblity to convert the auto generated c file to simulink model using any parser?
Short answer: No.
The real question is: why would you want to do that? The golden reference (the design if you want) is the model from which the code is generated, it's a one-way system. If you want to change the design, change the model, re-generate the code. That's how it's designed and intended to work, not the other way round.
You can however integrate C code (functions) into a Simulink model the legacy code tool, but that's a different exercise to what you are asking. You can also call external C code from a Simulink model using a MATLAB Function block and the coder.ceval command, see Integrate C Code Using the MATLAB Function Block in the documentation for more detail.s.
There is some research like this one in that area, but any approach I know attempts to convert general c code to a model. The "best" performance I have ever heard of was a ready to use and no longer available prototype which generated incredible huge simulink models out of some lines of c code.
To have any practical use in your use case, a translator which recognizes the original model blocks would be required. Such a software does not exist.
You can try Modelify.
Modelify is a new tool that converts legacy C code into Simulink models. Instead of a verbatim conversion, Modelify transforms low-level logic from C into higher level constructs in Simulink.
Website Link
I am building a Simulink model in order it generates the C-code equivalent. Actually, the C-code doing that already exists but I want to visualize it with Simulink. I know that the code Simulink will generate will not be the same than the one it is based on, but there are some functionalities I want to keep.
For example, I use a lot of Macro (#define) in the C code, to be more readable and I really do not know how to force Simulink to define these macro's. So my question is, which object/structure/data should I create and where (model workspace or Matlab workspace?) such that when C code will be generated, all my macro's will be defined with preprocessor commands.
I thank you a lot for any lead you could propose.
Finally I found my answer in this huge user guide
https://engineering.purdue.edu/~dionysis/EE452/Lab2/Lab2_Supporting_Materials/Embedded%20Coder_Users%20Guide.pdf
I had to use mpt.Parameter Imported from file and check Inline parameter option.
So let's say I have a string containing some code in C, predictably read from a file that has other things in it besides normal C code. How would I turn this string into code usable by the program? Do I have to write an entire interpreter, or is there a library that already does this for me? The code in question may call subroutines that I declared in my actual C file, so one that only accounts for stock C commands may not work.
Whoo. With C this is actually pretty hard.
You've basically got a couple of options:
interpret the code
To do this, you'll hae to write an interpreter, and interpreting C is a fairly hard problem. There have been C interpreters available in the past, but I haven't read about one recently. In any case, unless you reallY really need this, writing your own interpreter is a big project.
Googling does show a couple of open-source (partial) C interpreters, like picoc
compile and dynamically load
If you can capture the code and wrap it so it makes a syntactically complete C source file, then you can compile it into a C dynamically loadable library: a DLL in Windows, or a .so in more variants of UNIX. Then you could load the result at runtime.
Now, what normally would lead someone to do this is a need to be able to express some complicated scripting functions. Have you considered the possibility of using a different language? Python, Scheme (guile) and Lua are easily available to add as a scripting language to a C application.
C has nothing of this nature. That's because C is compiled, and the compiler needs to do a lot of building of the code before the code starts running (hence receives a string as input) that it can't really change on the fly that easily. Compiled languages have a rigidity to them while interpreted languages have a flexibility.
You're thinking of Perl, Python PHP etc. and so called "fourth generation languages." I'm sure there's a technical term in c.s. for this flexibility, but C doesn't have it. You'll need to switch to one of these languages (and give up performance) if you have a task that requires this sort of string use much. Check out Perl's /e flag with regexes, for instance.
In C, you'll need to design your application so you don't need to do this. This is generally quite doable, as for its non-OO-ness and other deficiencies many huge, complex applications run on well-written C just fine.
For teaching purpose we are building a javascript step by step interpreter for (a subset of) C code.
Basically we have : int,float..., arrays, functions, for, while... no pointers.
The javascript interpreter is done and allow us to explain how a boolean expression is evaluated, will show the variables stack...
For now, we are manually converting our C examples to some javascript that will run and build a stack of actions (affectation, function call...) that can later on be used to do the step by step stuff. Since we are limiting ourselves to a subset of C it's quite easy to do.
Now we would like to compile the C code to our javascript representation. All we need is a Abstract-syntax tree of the C code and the javascript generation is straightforward.
Do you know a good C-parser that could generate a such tree ? No need to be in javascript (but that would be perfect), any language is alright as this can be done offline.
I've looked at Emscripten ( https://github.com/kripken/emscripten ) but it's more a C=>javascript compiler and that's not what we want.
I've recently used Eli Bendersky's pycparser to mess with ASTs of C code. I think it'd work well for your purposes.
I think that ANTLR has a full C parser.
To do your translation task, I suspect you will need full symbol table support; you have to know what the symbols mean. Here most "parsers" will fail you; they don't build a full symbol table. I think ANTLR does not, but I could be wrong.
Our DMS Software Reengineering Toolkit with its C Front End provides a full C arser, and builds complete symbol tables. (You may not need it for your application, but it includes a full C preprocessor, too). It also provide control flow, data flow, points-to-analysis and call graph construction, all of which can be useful in translating C to whatever your target virtual machine is.
I've heard that there are some things one cannot do as a computer programmer, but I don't know what they are. One thing that occurred to me recently was: wouldn't it be nice to have a class that could make a copy of the source of the program it runs, modify that program and add a method to the class that it is, and then run the copy of the program and terminate itself. Is it possible for code to write code?
If you want to learn about the limits of computability, read about the halting problem
In computability theory, the halting
problem is a decision problem which
can be stated as follows: given a
description of a program and a finite
input, decide whether the program
finishes running or will run forever,
given that input.
Alan Turing proved in 1936 that a
general algorithm to solve the halting problem for all
possible program-input pairs cannot exist
Start by looking at quines, then at Macro-Assemblers and then lex & yacc, and flex & bison. Then consider self-modifying code.
Here's a quine (formatted, use the output as the new input):
#include<stdio.h>
main()
{
char *a = "main(){char *a = %c%s%c; int b = '%c'; printf(a,b,a,b,b);}";
int b = '"';
printf(a,b,a,b,b);
}
Now if you're just looking for things programmers can't do look for the opposite of np-complete.
Sure it is. That's how a lot of viruses work!
Get your head around this: computability theory.
Yes, that's what most Lisp macros do (for just one example).
Yes it certainly is, though maybe not in the context you are referring to check out this post on t4.
If you look at Functional Programming that has many opportunities to write code that generates further code, the way that a language like Lisp doesn't differentiate between code and data is a significant part of it's power.
Rails generates the various default model and controller classes from the database schema when it's creating a new application. It's quite standard to do this kind of thing with dynamic languages- I have a few bits of PHP around that generate php files, just because it was the simplest solution to the problem I was dealing with at the time.
So it is possible. As for the question you are asking, though- that is perhaps a little vague- what environment and language are you using? What do you expect the code to do and why does it need to be added to? A concrete example may bring more directly relevant responses.
Yes it is possible to create code generators.
Most of the time they take user input and produce valid code. But there are other possibilities.
Self modifying programes are also possible. But they were more common in the dos era.
Of course you can! In fact, if you use a dynamic language, the class can change itself (or another class) while the program is still running. It can even create new classes that didn't exist before. This is called metaprogramming, and it lets your code become very flexible.
You are confusing/conflating two meanings of the word "write". One meaning is the physical writing of bytes to a medium, and the other is designing software. Of course you can have the program do the former, if it was designed to do so.
The only way for a program to do something that the programmer did not explicitly intend it to do, is to behave like a living creature: mutate (incorporate in itself bits of environment), and replicate different mutants at different rates (to avoid complete extinction, if a mutation is terminal).
Sure it is. I wrote an effect for Paint.NET* that gives you an editor and allows you to write a graphical effect "on the fly". When you pause typing it compiles it to a dll, loads it and executes it. Now, in the editor, you only need to write the actual render function, everything else necessary to create a dll is written by the editor and sent to the C# compiler.
You can download it free here: http://www.boltbait.com/pdn/codelab/
In fact, there is even an option to see all the code that was written for you before it is sent to the compiler. The help file (linked above) talks all about it.
The source code is available to download from that page as well.
*Paint.NET is a free image editor that you can download here: http://getpaint.net
In relation to artificial intelligence, take a look at Evolutionary algorithms.
make a copy of the source of the program it runs, modify that program and add a method to the class that it is, and then run the copy of the program and terminate itself
You can also generate code, build it into a library instead of an executable, and then dynamically load the library without even exiting the program that is currently running.
Dynamic languages usually don't work quite as you suggest, in that they don't have a completely separate compilation step. It isn't necessary for a program to modify its own source code, recompile, and start from scratch. Typically the new functionality is compiled and linked in on the fly.
Common Lisp is a very good language to practice this in, but there are others where you can created code and run it then and there. Typically, this will be through a function called "eval" or something similar. Perl has an "eval" function, and it's generally common for scripting languages to have the ability.
There are a lot of programs that write other programs, such as yacc or bison, but they don't have the same dynamic quality you seem to be looking for.
Take a look at Langtom's loop. This is the simplest example of self-reproducing "program".
There is a whole class of such things called "Code Generators". (Although, a compiler also fits the description as you set it). And those describe the two areas of these beasts.
Most code generates, take some form of user input (most take a Database schema) and product source code which is then compiled.
More advanced ones can output executable code. With .NET, there's a whole namespace (System.CodeDom) dedicated to the create of executable code. The these objects, you can take C# (or another language) code, compile it, and link it into your currently running program.
I do this in PHP.
To persist settings for a class, I keep a local variable called $data. $data is just a dictionary/hashtable/assoc-array (depending on where you come from).
When you load the class, it includes a php file which basically defines data. When I save the class, it writes the PHP out for each value of data. It's a slow write process (and there are currently some concurrency issues) but it's faster than light to read. So much faster (and lighter) than using a database.
Something like this wouldn't work for all languages. It works for me in PHP because PHP is very much on-the-fly.
It has always been possible to write code generators. With XML technology, the use of code generators can be an essential tool. Suppose you work for a company that has to deal with XML files from other companies. It is relatively straightforward to write a program that uses the XML parser to parse the new XML file and write another program that has all the callback functions set up to read XML files of that format. You would still have to edit the new program to make it specific to your needs, but the development time when a new XML file (new structure, new names) is cut down a lot by using this type of code generator. In my opinion, this is part of the strength of XML technology.
Lisp lisp lisp lisp :p
Joking, if you want code that generates code to run and you got time to loose learning it and breaking your mind with recursive stuff generating more code, try to learn lisp :)
(eval '(or true false))
wouldn't it be nice to have a class that could make a copy of the source of the program it runs, modify that program and add a method to the class that it is, and then run the copy of the program and terminate itself
There are almost no cases where that would solve a problem that cannot be solved "better" using non-self-modifying code..
That said, there are some very common (useful) cases of code writing other code.. The most obvious being any server-side web-application, which generates HTML/Javascript (well, HTML is markup, but it's identical in theory). Also any script that alters a terminals environment usually outputs a shell script that is eval'd by the parent shell. wxGlade generates code to that creates bare-bone wx-based GUIs.
See our DMS Software Reengineering Toolkit. This is general purpose machinery to read and modify programs, or generate programs by assembling fragments.
This is one of the fundamental questions of Artificial Intelligence. Personally I hope it is not possible - otherwise soon I'll be out of a job!!! :)
It is called meta-programming and is both a nice way of writing useful programs, and an interesting research topic. Jacques Pitrat's Artificial Beings: the conscience of a conscious machine book should interest you a lot. It is mostly related to meta-knowledge based computer programs.
Another related term is multi-staged programming (because there are several stages of programs, each generating the next one).