stl list erase

**ES Kim** · Jul 19 '05, 04:31 PM

Re: stl list erase

"Paras" <parsharm@netsc ape.net> wrote in message
news:3F430157.2 0208@netscape.n et...[color=blue]
>
> Hi
>
>
> What is the correct way to delete an element from STL list while
> iterating through the list
>
>
> list<A*> _ls;
> A * a;
> list<A*>::itera tor si1;
> for (si1=_ls.begin( ); si1!=_ls.end(); ++si1) {
>
> if ( (*si1)->check() ) {
> a =(*si);
> _ls.erase(si1);
> /* Now this function will point pint si1 to next element. (no
> invalidate) Now how to still continue with for loop which will be a
> problem now ??? ..
> */
>
> delete a;
>
> }
> }[/color]

list<A*> _ls;
A* a;
list<A*>::itera tor si1 = _ls.begin();
while (si1 != _ls.end())
{
if ((*si1)->check())
{
a = *si1;
// si1 points to the next element after erase()
si1 = _ls.erase(si1);
delete a;

}
else
++si1;
}

--
ES Kim

**Naren** · Jul 19 '05, 04:32 PM

Re: stl list erase

Hello,

erase returns a iterator to first element remaining beyond any elements
removed.so use this

I like *while* so i have changed it to while sorry..;-)

Hope this helps.

//Code

list<A*> _ls;
A * a;
list<A*>::itera tor si1 = _ls.begin();
while(si1 !=_ls.end()) {

if ( (*si1)->check() ) {
a =(*si);
si1 = _ls.erase(si1);
delete a;
}

else
{
si1++;
}
}

"Paras" <parsharm@netsc ape.net> wrote in message
news:3F430157.2 0208@netscape.n et...[color=blue]
>
> Hi
>
>
> What is the correct way to delete an element from STL list while
> iterating through the list
>
>
> list<A*> _ls;
> A * a;
> list<A*>::itera tor si1;
> for (si1=_ls.begin( ); si1!=_ls.end(); ++si1) {
>
> if ( (*si1)->check() ) {
> a =(*si);
> _ls.erase(si1);
> /* Now this function will point pint si1 to next element. (no
> invalidate) Now how to still continue with for loop which will be a
> problem now ??? ..
> */
>
> delete a;
>
> }
> }
>
>[/color]

**John Harrison** · Jul 19 '05, 04:32 PM

Re: stl list erase

"ES Kim" <eskim@svd.co.k r> wrote in message
news:bhuu9g$cat $1@news1.kornet .net...[color=blue]
> "Paras" <parsharm@netsc ape.net> wrote in message
> news:3F430157.2 0208@netscape.n et...[color=green]
> >
> > Hi
> >
> >
> > What is the correct way to delete an element from STL list while
> > iterating through the list
> >[/color][/color]

[snip][color=blue]
>
>
> list<A*> _ls;
> A* a;
> list<A*>::itera tor si1 = _ls.begin();
> while (si1 != _ls.end())
> {
> if ((*si1)->check())
> {
> a = *si1;
> // si1 points to the next element after erase()
> si1 = _ls.erase(si1);
> delete a;
>[/color]

No good reason for the a varaible

delete *si1;
si1 = _ls.erase(si1);
[color=blue]
> }
> else
> ++si1;
> }
>
> --
> ES Kim
>[/color]

john

**Chris Theis** · Jul 19 '05, 04:32 PM

Re: stl list erase

"Naren" <narendranath.t s@in.bosch.com> wrote in message
news:bhv0s7$f85 $1@ns2.fe.inter net.bosch.com.. .[color=blue]
> Hello,
>
> erase returns a iterator to first element remaining beyond any elements
> removed.so use this
>
> I like *while* so i have changed it to while sorry..;-)
>
> Hope this helps.
>
> //Code
>[/color]
[SNIP][color=blue]
>
> else
> {
> si1++;[/color]

In this case I'd recommend to use ++si1; instead because it might be more
efficient. Anyway there's nothing to lose if you use prefix instead of
postfix notation in a case like this.
[color=blue]
> }
> }
>[/color]

Chris

**Gavin Deane** · Jul 19 '05, 04:32 PM

Re: stl list erase

Paras <parsharm@netsc ape.net> wrote in message news:<3F430157. 20208@netscape. net>...[color=blue]
> Hi
>
>
> What is the correct way to delete an element from STL list while
> iterating through the list
>
>
> list<A*> _ls;
> A * a;
> list<A*>::itera tor si1;
> for (si1=_ls.begin( ); si1!=_ls.end(); ++si1) {
>
> if ( (*si1)->check() ) {
> a =(*si);
> _ls.erase(si1);
> /* Now this function will point pint si1 to next element. (no
> invalidate) Now how to still continue with for loop which will be a
> problem now ??? ..
> */
>
> delete a;
>
> }
> }[/color]

Other people appear to have answered your question, but just to point
out:

Names beginning with an underscore are reserved for the implementation
in the global namespace so, depending on context, _ls might not be the
best name for your list.

GJD

**Rolf Magnus** · Jul 19 '05, 04:32 PM

Re: stl list erase

Gavin Deane wrote:
[color=blue]
> Names beginning with an underscore are reserved for the implementation
> in the global namespace[/color]

I thought this only applied if that undersore was followed by an
uppercase letter.

**Ron Natalie** · Jul 19 '05, 04:32 PM

Re: stl list erase

"Rob Williscroft" <rtw@freenet.RE MOVE.co.uk> wrote in message[color=blue]
>
> If I remember/understand correctly, those names and those with two
> underscores are reserved anywhere, i.e. so the implementor can use
> them as macro's.[/color]

And also to set off the decorations from the rest of the identifier (the
double-underscore prohibition isn't just for leading undersdcores but
anywhere in the identifier).[color=blue]
>
> Identifiers with a single leading underscore are reserved only in
> the global namespace.[/color]

NO. Indentifiers with leading underscore and a captial letter are not
permitted anywhere. If you look at a lot of the template library you'll
see many implementations use these symbols as their template variables,
etc...to keep things from colliding with possible user macro definitions.

It is _ followed by non-uppercase letters that is reserved ONLY in the
global namespace. These are typically used for certain implementation
defined global functions etc...

**Rob Williscroft** · Jul 19 '05, 04:32 PM

Re: stl list erase

llewelly wrote in news:86lltoz167 .fsf@Zorthluthi k.local.bar:
[color=blue][color=green]
>> If I remember/understand correctly, those names and those with two
>> underscores are reserved anywhere, i.e. so the implementor can use
>> them as macro's.[/color]
> [snip]
>
> Two *consecutive* underscores; i.e. this_is_okay this__is_not
>
>[/color]

Yes, It's what I meant but not what I wrote, Thanks.

Rob.
--

http://www.victim-prime.dsl.pipex.com/

**Gavin Deane** · Jul 19 '05, 04:33 PM

Re: stl list erase

Rob Williscroft <rtw@freenet.RE MOVE.co.uk> wrote in message news:<Xns93DDBF 15546DFukcoREMO VEfreenetrtw@19 5.129.110.201>. ..[color=blue]
> llewelly wrote in news:86lltoz167 .fsf@Zorthluthi k.local.bar:
>[color=green][color=darkred]
> >> If I remember/understand correctly, those names and those with two
> >> underscores are reserved anywhere, i.e. so the implementor can use
> >> them as macro's.[/color]
> > [snip]
> >
> > Two *consecutive* underscores; i.e. this_is_okay this__is_not
> >
> >[/color]
>
> Yes, It's what I meant but not what I wrote, Thanks.
>
> Rob.[/color]

I guess the level of uncertainty here lends weight to the argument
that it's simpler just to avoid leading underscores altogether (which
is my preference anyway). I checked the standard before my original
post in this thread so I'm fairly sure it was right, but the fact that
I felt I needed to supports that argument.

GJD

**Kevin Goodsell** · Jul 19 '05, 04:33 PM

Re: stl list erase

Gavin Deane wrote:[color=blue]
>
> I guess the level of uncertainty here lends weight to the argument
> that it's simpler just to avoid leading underscores altogether (which
> is my preference anyway).[/color]

I certainly agree with that. I've tried, but never been able to
determine for sure what the exact rules are.

-Kevin
--
My email address is valid, but changes periodically.
To contact me please use the address from a recent posting.

**Gavin Deane** · Jul 19 '05, 04:33 PM

Re: stl list erase

Kevin Goodsell <usenet1.spamfr ee.fusion@never box.com> wrote in message news:<3f441c01@ shknews01>...[color=blue]
> Gavin Deane wrote:[color=green]
> >
> > I guess the level of uncertainty here lends weight to the argument
> > that it's simpler just to avoid leading underscores altogether (which
> > is my preference anyway).[/color]
>
> I certainly agree with that. I've tried, but never been able to
> determine for sure what the exact rules are.[/color]

The rules are in 17.4.3.1 in the standard if you have a copy. Well
they're there whether you have a copy or not, but it will be harder
for you to look them up if you haven't ;-)

And as I said, just because I _can_ look them up doesn't mean I want
to every time I choose a variable name, so I steer clear of leading
underscores altogether.

GJD

**Alan Chen** · Jul 19 '05, 04:34 PM

Re: stl list erase

"Naren" <narendranath.t s@in.bosch.com> wrote in message news:<bhv0s7$f8 5$1@ns2.fe.inte rnet.bosch.com> ...[color=blue]
> Hello,
>
> erase returns a iterator to first element remaining beyond any elements
> removed.so use this
>
> I like *while* so i have changed it to while sorry..;-)
>
> Hope this helps.
>
> //Code
>
> list<A*> _ls;
> A * a;
> list<A*>::itera tor si1 = _ls.begin();
> while(si1 !=_ls.end()) {
>
> if ( (*si1)->check() ) {
> a =(*si);
> si1 = _ls.erase(si1);
> delete a;
> }
> else
> {
> si1++;
> }
> }[/color]

There's a bug in this code. If the second to last element is deleted,
the last element will never be checked for deletion.

**Alan Chen** · Jul 19 '05, 04:34 PM

Re: stl list erase

"Naren" <narendranath.t s@in.bosch.com> wrote in message news:<bhv0s7$f8 5$1@ns2.fe.inte rnet.bosch.com> ...[color=blue]
> Hello,
>
> erase returns a iterator to first element remaining beyond any elements
> removed.so use this
>
> I like *while* so i have changed it to while sorry..;-)
>
> Hope this helps.
>
> //Code
>
> list<A*> _ls;
> A * a;
> list<A*>::itera tor si1 = _ls.begin();
> while(si1 !=_ls.end()) {
>
> if ( (*si1)->check() ) {
> a =(*si);
> si1 = _ls.erase(si1);
> delete a;
> }
>
> else
> {
> si1++;
> }
> }[/color]

Nevermind, there is a bug in my brain ... never post w/o morning coffee...

stl list erase

stl list erase

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment