QOTD
-
I'm sorry, I forgot they implemented the operator-> too (argh! operator overloading begins to suck ;)
Clickety.I choose a random iterator type, but they all behave the same.
-
Congrats to Christian Graus for thinking up the QOTD. I believe this is the first one I've seen from a CPian. Of course, I can't answer because I haven't a clue when it comes to STL. So I'll just leave the answering of it to the rest of you. :) David Stone It seemed similar to someone saying, "Would you like to meet my knife collection?" Ryan Johnston on Elaine's sig
Ahh, my brain is switched off. For that past few days I've been going though interfaces/abstract classes and coding COM objects in pure C++, to gain a better understanding. Now I think I know this one, as I use them quite a bit, but I've not got the energy to make sure. brain is mush, and I've just finished watching Lord 'o' the Rings. I'm battered, and I think I'm going to have another glass of wine...... :-D(stupid grin)
-
Congrats to Christian Graus for thinking up the QOTD. I believe this is the first one I've seen from a CPian. Of course, I can't answer because I haven't a clue when it comes to STL. So I'll just leave the answering of it to the rest of you. :) David Stone It seemed similar to someone saying, "Would you like to meet my knife collection?" Ryan Johnston on Elaine's sig
What about option E) const_cast<Type&>(*it).Foo() ? Otherwise you're editing a copy of the object. Todd Smith
-
there's probably some deep metaphysical reason why people want to use (*it). instead of it-> . i just do it because that's how i learned it. Ryan Johnston wrote: What kind of programming do you do? a little of everything, but mostly 2d graphics. -c
Conservative: One who admires radicals centuries after they're dead. -- Leo C. Rosten
Chris Losinger wrote: there's probably some deep metaphysical reason why people want to use (*it). instead of it-> . i just do it because that's how i learned it AFAIR it is just because early versions of STL and/or the C++ compiler used those days didn't support the "intuitive" syntax. In our days using
it->Somethingshould work quite fine. -- Daniel Lohmann http://www.losoft.de (Hey, this page is worth looking! You can find some free and handy NT tools there :-D ) -
peterchen wrote: What's wrong with 'it' a You mean, aside from being incorrect code? ;) An iterator is an opaque data type, like say POSITION in MFC. You can't make any assumptions about it actually being a pointer to the underlying data, because iterator is not documented as such. --Mike-- Just released - RightClick-Encrypt v1.3 - Adds fast & easy file encryption to Explorer My really out-of-date homepage Sonork-100.19012 Acid_Helm
Michael Dunn wrote: You can't make any assumptions about it actually being a pointer to the underlying data, because iterator is not documented as such. But you can expect it to support (part of) the syntactic interface of a pointer (operator *(), operator ->() and for some iterators also operator []()) Therefore if you use it in generic (templated) code you can expect it to work like a pointer. We have to be very careful with terms like "pointer", "data" and so on here. What exactly is meant by "pointer". A pointer in the C sense or a "thing" that supports the interface of a pointer and therfore acts like a pointer (which is exactly the idea of a generic iterator used in STL). I am still not sure about the intention of the original question: Given an STL iterator it, what is the correct way to access a pointer to the iterator's data? Does it mean:
- a C-style pointer to the data the iterator refers to? Then the correct answer is c), even if I am not sure it should be written as &(*it) instead of &*it .
- a "pointer interface" to the data the iterator refers to? Then the correct answer is a).
- a "pointer inteface" to the iterator itself? Then the correct answer is b).
- ...
-- Daniel Lohmann http://www.losoft.de (Hey, this page is worth looking! You can find some free and handy NT tools there :-D )
-
Michael Dunn wrote: You can't make any assumptions about it actually being a pointer to the underlying data, because iterator is not documented as such. But you can expect it to support (part of) the syntactic interface of a pointer (operator *(), operator ->() and for some iterators also operator []()) Therefore if you use it in generic (templated) code you can expect it to work like a pointer. We have to be very careful with terms like "pointer", "data" and so on here. What exactly is meant by "pointer". A pointer in the C sense or a "thing" that supports the interface of a pointer and therfore acts like a pointer (which is exactly the idea of a generic iterator used in STL). I am still not sure about the intention of the original question: Given an STL iterator it, what is the correct way to access a pointer to the iterator's data? Does it mean:
- a C-style pointer to the data the iterator refers to? Then the correct answer is c), even if I am not sure it should be written as &(*it) instead of &*it .
- a "pointer interface" to the data the iterator refers to? Then the correct answer is a).
- a "pointer inteface" to the iterator itself? Then the correct answer is b).
- ...
-- Daniel Lohmann http://www.losoft.de (Hey, this page is worth looking! You can find some free and handy NT tools there :-D )
Daniel Lohmann wrote: I am still not sure about the intention of the original question: It's not that complicated. An iterator is not a pointer. "pointer" meaning "C-style pointer". iterator overloads pointer operators so you can treat it as a pointer, but that's a different story. (iterator has a ++ operator, does that make it a number?) Are we next going to debate about what the meaning of the word "is" is? ;) --Mike-- Just released - RightClick-Encrypt v1.3 - Adds fast & easy file encryption to Explorer My really out-of-date homepage Sonork-100.19012 Acid_Helm
-
Congrats to Christian Graus for thinking up the QOTD. I believe this is the first one I've seen from a CPian. Of course, I can't answer because I haven't a clue when it comes to STL. So I'll just leave the answering of it to the rest of you. :) David Stone It seemed similar to someone saying, "Would you like to meet my knife collection?" Ryan Johnston on Elaine's sig
I'm sorry but where was the question, I can't seem to find it? :confused: Nick Parker
-
I'm sorry but where was the question, I can't seem to find it? :confused: Nick Parker
Nick Parker wrote: I'm sorry but where was the question, I can't seem to find it? Scroll up. It's on the top of this page. It's called "Question of the Day". :)
-
Daniel Lohmann wrote: I am still not sure about the intention of the original question: It's not that complicated. An iterator is not a pointer. "pointer" meaning "C-style pointer". iterator overloads pointer operators so you can treat it as a pointer, but that's a different story. (iterator has a ++ operator, does that make it a number?) Are we next going to debate about what the meaning of the word "is" is? ;) --Mike-- Just released - RightClick-Encrypt v1.3 - Adds fast & easy file encryption to Explorer My really out-of-date homepage Sonork-100.19012 Acid_Helm
Michael Dunn wrote: Are we next going to debate about what the meaning of the word "is" is? Mike, you are getting too philosophical for a Friday night. :) Nick Parker
-
Chris Losinger wrote: but, i've honestly never seen that in any live code or example: it's always (*it).Foo(); Are you serious? I have never ever seen someone use (*it).Foo() instead of it->Foo(). Pointers would be so dirty without ->. Ryan Johnston
Absolutely. If you have a pointer to an object, within an iterator, then you need to dereference the iterator in order to get to the object. I forget what the question was ( it's not there now ), but I thought it was to do with passing a pointer, where you need to dereference the iterator, THEN grab the address of the dereferenced object, so you need to do this: &(*it) in order to turn an iterator which points to an object into a pointer to the object. Sadly, VC6 lets you use the iterator as a pointer. This is BAD, and is fixed in VC7. The fact that you can do something nonstandard in VC6 was the point of me asking the question. Having to dereference an iterator has nothing to do with pointers, iterators are not pointers, ( although pointers are iterators ), even if they act like them in some respects. As far as what Chris is talking about, I've never dereferenced an iterator to call a method, I've never had a problem on any STL version using -> to call a method on an iterator. I'd be interested to hear otherwise. Christian We're just observing the seasonal migration from VB to VC. Most of these birds will be killed by predators or will die of hunger. Only the best will survive - Tomasz Sowinski 29-07-2002 ( on the number of newbie posters in the VC forum ) Cats, and most other animals apart from mad cows can write fully functional vb code. - Simon Walton - 6-Aug-2002
-
The answer is C) (*it) is the value, so &(*it) is the address of the value. Or did I miss something? I vote pro drink :beer:
Yes, c is the answer. The reason I asked the question is the VC6 STL is broken and accepts the iterator as a pointer to the object. Christian We're just observing the seasonal migration from VB to VC. Most of these birds will be killed by predators or will die of hunger. Only the best will survive - Tomasz Sowinski 29-07-2002 ( on the number of newbie posters in the VC forum ) Cats, and most other animals apart from mad cows can write fully functional vb code. - Simon Walton - 6-Aug-2002
-
Michael Dunn wrote: You can't make any assumptions about it actually being a pointer to the underlying data, because iterator is not documented as such. But you can expect it to support (part of) the syntactic interface of a pointer (operator *(), operator ->() and for some iterators also operator []()) Therefore if you use it in generic (templated) code you can expect it to work like a pointer. We have to be very careful with terms like "pointer", "data" and so on here. What exactly is meant by "pointer". A pointer in the C sense or a "thing" that supports the interface of a pointer and therfore acts like a pointer (which is exactly the idea of a generic iterator used in STL). I am still not sure about the intention of the original question: Given an STL iterator it, what is the correct way to access a pointer to the iterator's data? Does it mean:
- a C-style pointer to the data the iterator refers to? Then the correct answer is c), even if I am not sure it should be written as &(*it) instead of &*it .
- a "pointer interface" to the data the iterator refers to? Then the correct answer is a).
- a "pointer inteface" to the iterator itself? Then the correct answer is b).
- ...
-- Daniel Lohmann http://www.losoft.de (Hey, this page is worth looking! You can find some free and handy NT tools there :-D )
Daniel Lohmann wrote: But you can expect it to support (part of) the syntactic interface of a pointer (operator *(), operator ->() and for some iterators also operator []()) Therefore if you use it in generic (templated) code you can expect it to work like a pointer. Try it in VC 7 and see what happens :) Daniel Lohmann wrote: a C-style pointer to the data the iterator refers to? Then the correct answer is c), even if I am not sure it should be written as &(*it) instead of &*it . Did I not put a bracket ? My bad. This is what I meant by the question though. Christian We're just observing the seasonal migration from VB to VC. Most of these birds will be killed by predators or will die of hunger. Only the best will survive - Tomasz Sowinski 29-07-2002 ( on the number of newbie posters in the VC forum ) Cats, and most other animals apart from mad cows can write fully functional vb code. - Simon Walton - 6-Aug-2002
-
Daniel Lohmann wrote: I am still not sure about the intention of the original question: It's not that complicated. An iterator is not a pointer. "pointer" meaning "C-style pointer". iterator overloads pointer operators so you can treat it as a pointer, but that's a different story. (iterator has a ++ operator, does that make it a number?) Are we next going to debate about what the meaning of the word "is" is? ;) --Mike-- Just released - RightClick-Encrypt v1.3 - Adds fast & easy file encryption to Explorer My really out-of-date homepage Sonork-100.19012 Acid_Helm
Michael Dunn wrote: (iterator has a ++ operator, does that make it a number?) Operator ++ - hey, that makes it a pointer! ;P ;P :-D -- Daniel Lohmann http://www.losoft.de (Hey, this page is worth looking! You can find some free and handy NT tools there :-D )