|
|
Hello to all, I have the following topic/question.
During my career (in big c/cpp-based projects) I found out, that managing dependencies between files is "almost all" what architecture means.
But there is an open topic for which I have no answer:
All tools I'm aware of (e.g. Doxygen) are parsing dependencies between h-files only. Why do they not search for include-statements in c/cpp-files? I think these includes in c/cpp-files are equally important.modified on Saturday, February 13, 2010 10:22 AM
|
|
|
|
|
For me, an Architect does not (always) go as deep as to define what goes into which header file.
As far as I know, doxygen will pre-process all files, cpp and h, to extract information to build the documentation.
so, what is the question ?
 Watched code never compiles.
|
|
|
|
|
I also do not define what header goes into which h- oder cpp-file, but I define the dependencies of the components: In a tree-like component-hierarchy. This means, if a component uses the other, it has to include some of it's headers.
By parsing the include-structure (and also what the cpp-files are including - if possible) I can easily verify if the implementation fits to my defined architecture, because if nothing is included, it can also not depend on, because the compiler will not compile.
In this way I can easily control the quality of implementation independent of the size of a project, let it have 10 files or 10.000 files.
Basically I want to avoid cyclic dependencies which cause the "sudden death" for any non-trivial sw.
But now I'm experimenting with Doxygen and I do not really understand it.
|
|
|
|
|
Now I found out what the problem is with the parsing of cpp-files with doxygen.
But I do not have a solution.
For the following please assume, that a.h,cpp are in directory A and b.h is in directory B
-root-Dir
- A-Dir (a.h, a.cpp)
- B-Dir (b.h)
If I have in the file a.cpp, then Doxygen works very well and shows that Direcotry A depends on Directory B
// File: a.cpp
#include "a.h"
#include "b.h"
But I I include b.h in a relative way, then Doxygen does not show the dependency
// File: a.cpp
#include "a.h"
#include ".../B/b.h"
Ok, I found the problem, but I do not have a solution for configuring Doxygen
|
|
|
|
|
I will be better if you ask this quetion in Doxygen forum[^].
-Saurabh
|
|
|
|
|
Hi! I have such a problem: I need to calculate the work time per day(24 hours) when computer works: user log in in the system...working for some time and then log out. Is there any ways how to get the time when he logs in and logs out the system? OS Windows XP, Vista.
Thanks!
|
|
|
|
|
|
Thanks!!
|
|
|
|
|
Hi to all
Please look at the following code:
class base
{
public:
void show_base()
{
cout<<"executing base"<<endl;
}
};
class derived:public base
{
public:
void show_derived()
{
show_base();
cout<<"code is added at derived class":
}
};
void main()
{
derived d;
d.show_derived();
}
=============================
Required output:
executing base
code is added at derived class
=============================
In the above my problem is "I want to access a base class member function.i.e when i called show_derived(), it needs first calls show_base() and then show_derived() remaining code should execute. Is it possible to call like this. Or please show me any other possibilities
Thanks in advance......................To invent something, you need a mountain of junk in your mind.
---------------------Thomas alva edison
|
|
|
|
|
sampath-padamatinti wrote: cout<<"code is added at derived class":
Excluding a minor syntax error in the above line (colon instead of semicolon at the end of the statement), your code,as it stands, produces exactly the output you require.
 If the Lord God Almighty had consulted me before embarking upon the Creation, I would have recommended something simpler.
-- Alfonso the Wise, 13th Century King of Castile.
This is going on my arrogant assumptions. You may have a superb reason why I'm completely wrong.
-- Iain Clarke
[My articles]
|
|
|
|
|
Hello,
The code is giving required output. What is your question? Is that something which you need different an output?
|
|
|
|
|
Thanks for reply,
sorry, I am not check my code properly. It s giving me correct answer.To invent something, you need a mountain of junk in your mind.
---------------------Thomas alva edison
|
|
|
|
|
Hi
Is there a way to migrate a C++ application to web-based application? All GUI and user interfaces should use web and others use C++ code. Like a web-based vector editor.
Thanks,
|
|
|
|
|
If you mean "with an automatic tool" I'm diffident: A standalone application and a web-based application have a completely different rendering and communication logic.
If you mean "is it possible to make in a browser an applet that renders a remote desktop executing an application", well ... I doubt about performances, and that's not a web app., it's just masked as it!
You can -in some particular cases- for "form based apps" migrate them into "web based" since form-based have a communication mimic that's similar to the web one. But that's a very particular case.
2 bugs found.
> recompile ...
65534 bugs found.
|
|
|
|
|
You might want to have a look at "wt" (pronounced 'witty') http://www.webtoolkit.eu/wt]
You still have to rebuild your UI with the wt classes, but its C++ style.
|
|
|
|
|
Thank you for your reply.
I have a question for you. Does "WT" compile the code into "exe"?
Thanksmodified on Friday, February 12, 2010 6:02 PM
|
|
|
|
|
AFAIK it is possible to build an executable for UNIX or Windows with build-in http server or you can build a mod for apache (uses FastCGI). Please see http://www.webtoolkit.eu/wt#/features[^]
|
|
|
|
|
|
Could you introduce a little of the way?
Thanks
|
|
|
|
|
Wt is a C++ library, so you link it in your executable as a DLL. It consists of a convenient set of building blocks (text, buttons, trees, charts, graphics, ...) that you can combine to create a web UI in an object oriented way. You use the building blocks to build a tree which represents your interface (e.g. a Dialog containing a text and two buttons). It works much like traditional C++ GUI toolkits, such as Qt. But instead of rendering on a desktop screen, Wt uses the browser to render. Event handling is transparent: if a user clicks a button in the browser, a function is automatically invoked at the server. The programmer specifies which functions need to be invoked by means of a signal/slot mechanism. From those functions, you can easily interact with your existing application.
If you run a Wt application, it will behave like a web server and listen on a TCP port. When a user surfs to a Wt application, Wt investigates the browser capabilities and renders your widget tree in the most optimal way to your browser (JavaScript, AJAX, browser-specific dialects, ...).
But why don't you take a look at Wt's hello world example?
http://www.webtoolkit.eu/wt/examples/
Or an interactive demonstration of the available widgets:
http://www.webtoolkit.eu/widgets/basics/wpanel
|
|
|
|
|
the standard process is this:
1. form a committee to investigate the migration: is it possible, is it worth the trouble?
2. assign two developers to come up with a prototype. when they deliver an ugly app that does nothing...
3. hire a consultant to tell you that it will cost $X,000,000 and will take 2 years.
4. fire all your developers.
5. hire an outsourcing company who claims they can do it in six months.
6. pout, when they deliver something ugly and broken.
|
|
|
|
|
|
Is it possible to run a macro as a post build step? I've had a good look round but can't see any way to do it...
|
|
|
|
|
| Can you write the macro as a .vbs file? If so you should be able to launch a cmd process to run that. Or are you referring to an IDE macro? Chris Meech
I am Canadian. [heard in a local bar]
In theory there is no difference between theory and practice. In practice there is. [Yogi Berra]
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
