问个先后的问题 - 未名空间MITBBS历史存档

国际科技财经博客移民网络热点娱乐民生时事公众号

Redian新闻

>未名空间

>Programming - 葵花宝典

问个先后的问题

问个先后的问题# Programming - 葵花宝典

j*k2008-02-17 08:02

1 楼

class A{};
class B{};
class C: public A
{
C(){};
B b;
}
when instantiate C, C ctor first or B b first?

P*e2008-02-17 08:02

2 楼

A -> B -> C

【在 j*****k 的大作中提到】

: class A{};
: class B{};
: class C: public A
: {
: C(){};
: B b;
: }
: when instantiate C, C ctor first or B b first?

p*o2008-02-17 08:02

3 楼

Well, to be more accurate, aren't A's ctor and B's ctor called by C's ctor?

【在 P********e 的大作中提到】

: A -> B -> C

P*e2008-02-17 08:02

4 楼

A's ctor yes,
B's ctor no.

【在 p***o 的大作中提到】

: Well, to be more accurate, aren't A's ctor and B's ctor called by C's ctor?

p*o2008-02-17 08:02

5 楼

Why not? C() {} is C(): b() {} .

【在 P********e 的大作中提到】

: A's ctor yes,
: B's ctor no.

P*e2008-02-17 08:02

6 楼

Sure, you can explicitely write the code this way, then the answer change as
well
Just treat B as build-in type

【在 p***o 的大作中提到】

:
: Why not? C() {} is C(): b() {} .

p*o2008-02-17 08:02

7 楼

The answer should not change no matter how I write the code (well, there
is a difference, if you treat it as a difference, for build-in types and
PODs).
I mean if we are talking about the 'body' (inside {}) of those ctors, then
A->B->C is definitely correct. But I don't know if the interviewer really
want this anwser ...

as

【在 P********e 的大作中提到】

: Sure, you can explicitely write the code this way, then the answer change as
: well
: Just treat B as build-in type

P*e2008-02-17 08:02

8 楼

To anwser it correctly is enough, i checked thinking in c++, that's about it

【在 p***o 的大作中提到】

: The answer should not change no matter how I write the code (well, there
: is a difference, if you treat it as a difference, for build-in types and
: PODs).
: I mean if we are talking about the 'body' (inside {}) of those ctors, then
: A->B->C is definitely correct. But I don't know if the interviewer really
: want this anwser ...
:
: as