> What's the difference between using some like struct node* current as > a temporary variable in the following function... > struct node* current = head; > while (current != NULL) { > current = current->next; > } > while (head != NULL) { > head = head->next; > } > Or wouldn't it matter in this case?
Wouldn't matter in this case. In general, it doesn't matter what you do to arguments, as they are passed by value. (Arrays are sort of an exception, but not really; what is actually passed is a pointer, because the array decays into a pointer, and the pointer is passed by value.)
Often people do this when they expect to need to refer to the original value again later; for instance, iterating over a list with intent to delete members. Sometimes someone will have such a function, then copy and paste it and modify it to serve a new purpose, and the new one no longer needs the stashed copy.
Harmless; I doubt you'll see many optimizers where the code's different these days.
"bartc" <ba...@freeuk.com> writes: > "Chad" <cdal...@gmail.com> wrote in message
<snip other version>
>> int Length(struct node* head) { >> int count = 0; >> while (head != NULL) { >> count++; >> head = head->next; >> } >> return count; >> } <snip> > Probably doesn't matter, but in the second case, 'head' is a misnomer > for the variable.
Is it? At every stage of the loop, count is the number of list items seen and head points to the start of a list whose length is yet to be determined. I don't mind the name at all.
Chad wrote: > On Nov 7, 9:40 am, "bartc" <ba...@freeuk.com> wrote: >> "Chad" <cdal...@gmail.com> wrote in message
>> news:5605a29f-f1a3-4c08-89a9-ecb1c6421fc2@f20g2000prn.googlegroups.com... >>> What's the difference between using some like struct node* current as >>> a temporary variable in the following function... >>> int Length(struct node* head) { >>> struct node* current = head; >>> int count = 0; >>> while (current != NULL) { >>> count++; >>> current = current->next; >>> } >>> return count; >>> } >>> As opposed to not using one in the same function.... >>> int Length(struct node* head) { >>> int count = 0; >>> while (head != NULL) { >>> count++; >>> head = head->next; >>> } >>> return count; >>> } >>> Or wouldn't it matter in this case? >> Probably doesn't matter, but in the second case, 'head' is a misnomer for >> the variable. > Aye, that was sloppy naming on my part.
No, head is a reasonable name for the parameter designating the head of a list and having the parameter properly described for the caller is high on my list. When you reuse it, though, it takes on a different meaning, although I understand Ben's argument that it is the head of the remaining list instead of the provided list. I consider the first version slightly better for documentation purposes.
> What's the difference between using some like struct node* current as > a temporary variable in the following function...
> int Length(struct node* head) { > struct node* current = head; > int count = 0; > while (current != NULL) { > count++; > current = current->next; > } > return count; > }
> As opposed to not using one in the same function....
> int Length(struct node* head) { > int count = 0; > while (head != NULL) { > count++; > head = head->next; > } > return count; > }
> Or wouldn't it matter in this case?
it is mostly a stylistic difference, where using a local variable is a little "cleaner" than using the argument (and is not as "destructive" to the argument in question).
also, of note, is that when using languages which support references (may be called "pass by reference" in many languages), then the semantics will differ as well, since modifying the argument "may" also modify the value in the variable held by the caller.
so, this issue could be an issue in languages such as Perl, VB, or AFAIK some Pascal variants. (apparently Perl and VB do so by default...).
it could also matter in C++, although the syntax is more explicit in this case: int Length(struct node* &head); or, in Pascal and friends: function Length(var head: ^node): integer;
(where the 'var' could be easily missed...)
so, using a local makes ones' intentions a little clearer, even though, technically, in C it is unecessary...
|
