| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13
Comments:
<0> the best idea is to refactor the code into a common function say _Init() and call that from each ctor
<0> imho
<1> I'm using pthread_create() to build a thread, and I need to send a (void *)arg - how can I p*** through this arg an array of 4 int elements?
<2> I wouldn't call any code that can fail from a ctor...
<2> You can't throw exceptions from the ctor, correct?
<0> no
<2> No thats not correct, or no you can't throw exceptions from the ctor?
<2> I do flattening from the ctor, then have a second function Initialize() that does basic initialization (that might fail).
<0> eh?
<2> flattening like ***igning all my pointers to (type*)0;
<2> more detailed stuff in another step...
<2> I guess it depends on the complexity of the cl***...
<0> exactly, the original question was about calling another ctor
<0> if the ctor is likely to throw, it'll throw either way
<2> But if you throw from the constructor, the objects destructor isn't called
<0> and?
<2> The stuff that does succeed in the constructor would need to be undone...
<0> fine...but is that relevant to the original question?
<0> this is ctors 101 stuff
<2> I might have missed the original question, I just saw you say to call an Init() function from the ctor
<2> which if performs code that fails, it might be more difficult to undo what the ctor actually accomplished before it threw an exception
<2> where it might be better to call the Init() outside the ctor
<2> for better exception handling...
<3> Kaervos, whay do you mean ?
<3> Kaervos, if an exception is thrown during the construction of an object, all fully-constructed sub-objects are destroyed
<2> The destructor isn't called if an exception is thrown from the constructor
<3> Kaervos, but that's not a problem, since the object was never created
<2> What if you locked a semaphore in the Init() function you call from your constructor, then an exception is thrown before the constructor is complete?
<2> or opened a file
<2> or allocated memory
<3> Kaervos, you must unlock the semaphore, as you would do in any other function
<4> If those resources are wrapped in a cl*** (and they should be, for this reason) then you're fine.
<4> Google for RAII
<2> Tru, Sol.
<3> Kaervos, the fact that the function is a constructor doesn't make a difference in this regard
<4> The pattern is cl*** Resource { ResHandle r; public: Resource() { acquire(r); } ~Resource { release(r); } };
<4> Then if every resource is so acquired (or ownership is thus transferred) exception handling, and indeed resource handling in general, is a breeze.
<2> Its a case for std::auto_ptr
<4> Well, std::auto_ptr is a case of the general pattern of RAII.
<2> I'll go check that out
<3> std::auto_ptr is suited only for heap-allocated objects
<4> Kaervos: More info on the topic: http://www.relisoft.com/resource/resmain.html
<4> Another good article on the same site: http://www.relisoft.com/tools/CppCritic.html
<4> You should read that, Run. There are a couple of niches you could fill with your parser code.
<5> http://www.noidea128.org/sourcefiles/15875.html
<5> can anyone help me?
<4> For future reference, "/join #c main"
<6> wow... it's been a while since I've seen an array of floats :)
<5> :((
<5> im so going to fail this subject
<6> what are all those | in your code?
<5> it's um
<5> like endpoints of where the function should start and end
<5> we are to make them into their own functions
<7> http://www.cplusplus.com/doc/tutorial/functions.html
<5> my question is,
<7> Arrays are p***ed by reference by default
<5> see, on the first part, what i'm thinking is that i'm going to copy the first function to void input()
<7> You don't have to p*** them in and return them
<5> what do you mean
<5> like i can call them from another function?
<7> So p*** the array and its size (10 in this case) to a function that performs the first loop
<7> No, but if you p*** them in, any change you make inside the destination function is automatically communicated back to the calling function
<7> http://www.cplusplus.com/doc/tutorial/functions2.html
<7> Read the first section there, about reference and value parameters
<5> will do
<5> although what you said is still kinda aliento me :S
<8> heh
<7> That's because you didn't read up on functions like you were supposed to
<7> Wait, this is C or C++ you're doing?
<7> I don't know if arrays are p***ed by reference automagically in C, as it has no references of any sort
<7> I think they are, though
<8> Solamente: hello? Any chance you'll at least read a url?
<7> Someone check me on that
<5> c or c++? i'm not sure either
<4> JBlitzen, arrays are treated the same in both.
<7> Well, look, we're not omniscient, you have to at least know what language you're supposed to be learning
<4> anberlin, you're doing C
<6> I think the arrays are p***ed like it's a pointer... but like casted to an array of a specific size
<7> Thank you, sol
<5> is there a way to distinguish?
<7> Yeah, that's what I was thinking, since they aren't really references in C++ either
<7> Well, anberlin, your textbook should make it pretty obvious
<7> Either it's a book on C++ or a book on C
<8> heh
<5> we dont use a textbook
<7> :|
<5> it's all lecture by lecture
<5> prof doesnt believe in books :S
<7> Do your filenames end in .c or .cpp or .cc?
<5> .c
<8> heh
<7> Okay, then it's C
<5> what about for .cpp and .cc?
<7> And, for once, I'm going to agree that your professor is an absolute moron
<7> .cpp and .cc are usually C++ file extensions
<7> .c is usually C
<6> doesn't a .C with a big C mean like C++ too?
<7> (Then there are .h and .hpp and .make extensions, but those don't mean much about which language)
<7> I don't know, but he said .c, not .C
<7> Is it .c or .C, anberlin?
<5> .c
<5> is there a difference?
<8> anberlin: yes
<7> Evidently
<7> Anyway, I wouldn't worry about that, I'm pretty sure it's C
<5> yeah this course is on data structures
<4> ffs
<5> ? anythin wrong
<8> anberlin: he's amazed at your lack of knowledge of what you're doing, I think
<4> The state of software development education is what's wrong.
<4> We see its disastrous results in here every day.
<5> well exactly why im asking for your help:)_
<7> http://www.cprogramming.com/tutorial/c/lesson4.html
<7> anberlin ^
<5> oh alright, i'll go read up on that first.
<7> At some point you'll want to start talking to #C instead of to us
<5> thanks.
<7> yah
<5> same people right?
<8> anberlin: in some cases, probably.
<4> No
<6> are the big internet companies arguing again? my internet is going slow and I have some fat pipes... shouldn't be slow at all
<5> are you guys like, at war
<4> I don't do C unless forced to do so, or paid exorbitant amounts.
<5> "c is better than c++, noooo, c++ is better than c!"
<8> and java's better'n them both!
<8> bwahahaha
<7> <5> are you guys like, at war
<7> Haha
<7> Not really, but they're legitimately different languages
<7> The first advice I gave you out of the block was for C++, which would have screwed you up quite a bit
<5> good to hear
<5> wellwhat is the dif between c and c++?
<7> There are a lot of differences
<5> syntax? etc?
<7> C++ is basically a superset of C
<7> With some exceptions, C++ has the same and many more features than C
<4> Used properly, C++ won't bite you in the butt like C will.
<7> Reference parameters are one of the C++ exclusive features
<7> So, note how long it took to come up with an example of something in C++ that can't be done the same way in C
<4> C is just waiting to drop a buffer-overrun exploit or a memory leak into your application.
<7> About three minutes from when we started talking to you
<4> Used properly, C++ fairly eliminates problems like that.
<5> wow you guys overwhelm me :(
<7> Anyway, read up on that URL and see if you still have questions
<3> what can be done in C and C++ ?
<7> Don't be overwhelmed
<5> heh
Return to
#c++
or
Go to some related
logs:
#linux
how to do backflips
#AllNiteCafe
site:www.quotesdb.info rotards
#AllNiteCafe
#linuxhelp
ellezza
quotes from - Bulibasha
#AllNiteCafe
#chatzone
|
|