Finally one site with all the C fundas

**Richard Heathfield** · Jun 27 '08, 07:40 PM

Re: Finally one site with all the C fundas

parag_paul@hotm ail.com said:

>

http://concentratedlemonjuice.blogspot.com/2008/06/about-c-puzzles-on-my-blog.html

Hi All, I know you would have come to this page after some searching online. I am currently ranked 23 in the c puzzles section. So if yo...

Don't bother, folks - it's the usual mishmash of myth, error, and
occasional (accidental?) correctness.

--
Richard Heathfield <http://www.cpax.org.uk >
Email: -http://www. +rjh@
Google users: <http://www.cpax.org.uk/prg/writings/googly.php>
"Usenet is a strange place" - dmr 29 July 1999

**Jens Thoms Toerring** · Jun 27 '08, 07:40 PM

Re: Finally one site with all the C fundas

parag_paul@hotm ail.com <parag_paul@hot mail.comwrote:

http://concentratedlemonjuice.blogsp...n-my-blog.html

I actually am searching for more work done on C now, today there are
many sites that will provide you college level understanding of
pointers, buta time comes when in a software project you will have to
go ahead and decide about using some feature of C++ that you never
used

Sorry, but if you want to publish (rather trivial) C puzzles
you first should observe the distinction between C and C++.
Most of the so-called C-puzzles are actually C++ (I can't
judge if the answers are correct in those cases). C and C++
are different languages. Next you should correct some of
the mistakes. E.g.

2: What is wrong with this macro to find out the
square of a number

#define square(a) a*a

ANSWER:
If you pass an expression to this macro, it fails.
Eg: square(2+3) will be expanded to 2+3*2+3 = 11
whereas the actual result should be 25
Inorder to solve this problem you can write the macro as
#define square(a) ((a)*(a))

The correct answer is, of course, that this macro can't be
made safe for everyday use, even with the extra parentheses,
since e.g. 'square(i++)' will still result in an expression
invoking undefined behaviour..

Or

5. Are the statements 1 & 2 right, if we declare str as follows?

char * const str="Hello";

1. *str='W';
2. str="World";

ANSWER:
1-right
2-wrong

Neither 1 nor 2 are valid assignments. The first one isn't since
you're not allowed to change a string literal (even though you
may get away with it on some systems).

BYW, if in #4 you leave out the 'const' in front of

const char *abcd="Test";

you still wouldn't be allowed to do

*abcd='H';

for exactly the same reason.

Or

6. What is the problem with the following piece of code

typedef struct A{
int i;
};

A* RetStruct()
{
A a;
a.i=10;
return &a;
}

won't even compile since there is no type 'A' defined anywhere.
The compiler probably will complain about the ';' on line 3.

These are 3 out of 6 "puzzles" that could be C, the rest is all
C++ as far as I can see.
Regards, Jens
--
\ Jens Thoms Toerring ___ jt@toerring.de
\______________ ____________ http://toerring.de

**Keith Thompson** · Jun 27 '08, 07:40 PM

Re: Finally one site with all the C fundas

Regarding the subject "Finally one site with all the C fundas", a bit
of Googling indicates that "funda" is a contraction of "fundamenta l",
origininating in Indian English. I've never encountered it before. I
would suggest avoiding the term if you want to be understood by
readers outside India.

jt@toerring.de (Jens Thoms Toerring) writes:

parag_paul@hotm ail.com <parag_paul@hot mail.comwrote:

>http://concentratedlemonjuice.blogsp...n-my-blog.html

[...]

2: What is wrong with this macro to find out the
square of a number
>
#define square(a) a*a
>
ANSWER:
If you pass an expression to this macro, it fails.
Eg: square(2+3) will be expanded to 2+3*2+3 = 11
whereas the actual result should be 25
Inorder to solve this problem you can write the macro as
#define square(a) ((a)*(a))
>
The correct answer is, of course, that this macro can't be
made safe for everyday use, even with the extra parentheses,
since e.g. 'square(i++)' will still result in an expression
invoking undefined behaviour..

Also, the statement that "If you pass an expression to this macro, it
fails" is incorrect, or at least incomplete. I can pass the
expression (2+3) to the macro: square((2+3)), and it works just fine.

The answer section explains the inner parentheses around the
arguments, but not the outer parentheses around the whole definition.

The other problem with the macro is its name. By convention, macros
are usually given all-caps names, to warn the reader about the
possibility of, for example, multiple side-effects. This:

#define SQUARE(a) ((a)*(a))

would be unobjectionable . It does have the problem Jens points out,
that any side effects of the argument will occur twice, but using an
all-caps name warns a knowledgeable user to avoid calling it that way.

Or

[snip]

>
Or
>
6. What is the problem with the following piece of code
>
typedef struct A{
int i;
};
>
A* RetStruct()
{
A a;
a.i=10;
return &a;
}
>
won't even compile since there is no type 'A' defined anywhere.
The compiler probably will complain about the ';' on line 3.
>
These are 3 out of 6 "puzzles" that could be C, the rest is all
C++ as far as I can see.

Presumably #6 is C++.

Another problem, whichever language it's intended to be, is that it's
not indented. (And the white text on a black background on the web
site is extremely annoying, at least to me.)

--
Keith Thompson (The_Other_Keit h) kst-u@mib.org <http://www.ghoti.net/~kst>
Nokia
"We must do something. This is something. Therefore, we must do this."
-- Antony Jay and Jonathan Lynn, "Yes Minister"

**pete** · Jun 27 '08, 07:40 PM

Re: Finally one site with all the C fundas

Keith Thompson wrote:

Regarding the subject "Finally one site with all the C fundas", a bit
of Googling indicates that "funda" is a contraction of "fundamenta l",
origininating in Indian English. I've never encountered it before. I
would suggest avoiding the term if you want to be understood by
readers outside India.
>
jt@toerring.de (Jens Thoms Toerring) writes:

>parag_paul@hotm ail.com <parag_paul@hot mail.comwrote:

>>http://concentratedlemonjuice.blogsp...n-my-blog.html

[...]

> 2: What is wrong with this macro to find out the
> square of a number
>>
> #define square(a) a*a
>>
> ANSWER:
> If you pass an expression to this macro, it fails.
> Eg: square(2+3) will be expanded to 2+3*2+3 = 11
> whereas the actual result should be 25
> Inorder to solve this problem you can write the macro as
> #define square(a) ((a)*(a))
>>
>The correct answer is, of course, that this macro can't be
>made safe for everyday use, even with the extra parentheses,
>since e.g. 'square(i++)' will still result in an expression
>invoking undefined behaviour..

>
Also, the statement that "If you pass an expression to this macro, it
fails" is incorrect, or at least incomplete. I can pass the
expression (2+3) to the macro: square((2+3)), and it works just fine.

Also, square(5) works just as well.
What definition are you guys using for "expression "?
Not one of the ones from either of the C standards, I hope.

--
pete

**rahul** · Jun 27 '08, 07:40 PM

Re: Finally one site with all the C fundas

On Jun 9, 6:13 pm, "parag_p...@hot mail.com" <parag_p...@hot mail.com>
wrote:

http://concentratedlemonjuice.blogsp...t-c-puzzles-on...
>
I actually am searching for more work done on C now, today there are
many sites that will provide you college level understanding of
pointers, buta time comes when in a software project you will have to
go ahead and decide about using some feature of C++ that you never
used
>
some funda of C that you never saw ,and it adds to the cleanliness of
code and helps others improve their standards too.
I believe in this policy that when you write good code, you go one
step extra in helping others achieve better coding skill stoo

Probably you should refer the comp.lang.c FAQ first. Its better than
any crap posted
else where in the name of c puzzles. Most of the so called C puzzles
are there with
logical and standard complying answers.

**parag_paul@hotmail.com** · Jun 27 '08, 07:40 PM

Re: Finally one site with all the C fundas

On Jun 10, 10:03 am, rahul <rahulsin...@gm ail.comwrote:

On Jun 9, 6:13 pm, "parag_p...@hot mail.com" <parag_p...@hot mail.com>
wrote:
>

http://concentratedlemonjuice.blogsp...t-c-puzzles-on...

>

I actually am searching for more work done on C now, today there are
many sites that will provide you college level understanding of
pointers, buta time comes when in a software project you will have to
go ahead and decide about using some feature of C++ that you never
used

>

some funda of C that you never saw ,and it adds to the cleanliness of
code and helps others improve their standards too.
I believe in this policy that when you write good code, you go one
step extra in helping others achieve better coding skill stoo

>
Probably you should refer the comp.lang.c FAQ first. Its better than
any crap posted
else where in the name of c puzzles. Most of the so called C puzzles
are there with
logical and standard complying answers.

wow,
I should have posted all the question in the limelight individually.,
I am sorry for using the term Funda

Sorry for putting all the uncorrected question here. See this
acatually improves the content in the long run.
Please take this a beta edition. This is not a write and forget
attempt. I will try to improve the questions and answers give time.

Actually, most people dont dive into the intricacies given the amount
of time they want to spend. Thye want answer to get it done ( with
linux 4.2 , gcc is no more what it used to be )
so it is time that we come together and give the most plausible and
accurate answer to all the questions.

I believe this attempt will take my knowledge further and also it will
increase the fundamentals of the readers in the long run.

I will not discard it, please take my intentions to be of the first
rate.
I know that you will brand it as a myth compilation but lets bring
forward the truth in one place atleast.
The FAQ too has some of them common but trust me it is a collection
over the last few years.

Hope you will help me. I will not put all the question all togetehr. I
will go over them individually, and get it checked

**parag_paul@hotmail.com** · Jun 27 '08, 07:40 PM

Re: Finally one site with all the C fundas

On Jun 10, 10:03 am, rahul <rahulsin...@gm ail.comwrote:

On Jun 9, 6:13 pm, "parag_p...@hot mail.com" <parag_p...@hot mail.com>
wrote:
>

http://concentratedlemonjuice.blogsp...t-c-puzzles-on...

>

I actually am searching for more work done on C now, today there are
many sites that will provide you college level understanding of
pointers, buta time comes when in a software project you will have to
go ahead and decide about using some feature of C++ that you never
used

>

some funda of C that you never saw ,and it adds to the cleanliness of
code and helps others improve their standards too.
I believe in this policy that when you write good code, you go one
step extra in helping others achieve better coding skill stoo

>
Probably you should refer the comp.lang.c FAQ first. Its better than
any crap posted
else where in the name of c puzzles. Most of the so called C puzzles
are there with
logical and standard complying answers.

I will try to get the correct answers here then, But when you search
online you will come across hundreds of answers , will it be better to
ask them here one by one

**rahul** · Jun 27 '08, 07:40 PM

Re: Finally one site with all the C fundas

On Jun 10, 1:25 pm, "parag_p...@hot mail.com" <parag_p...@hot mail.com>
wrote:

On Jun 10, 10:03 am, rahul <rahulsin...@gm ail.comwrote:
>
>
>

On Jun 9, 6:13 pm, "parag_p...@hot mail.com" <parag_p...@hot mail.com>
wrote:

>

>http://concentratedlemonjuice.blogsp...t-c-puzzles-on...

>

I actually am searching for more work done on C now, today there are
many sites that will provide you college level understanding of
pointers, buta time comes when in a software project you will have to
go ahead and decide about using some feature of C++ that you never
used

>

some funda of C that you never saw ,and it adds to the cleanliness of
code and helps others improve their standards too.
I believe in this policy that when you write good code, you go one
step extra in helping others achieve better coding skill stoo

>

Probably you should refer the comp.lang.c FAQ first. Its better than
any crap posted
else where in the name of c puzzles. Most of the so called C puzzles
are there with
logical and standard complying answers.

>
I will try to get the correct answers here then, But when you search
online you will come across hundreds of answers , will it be better to
ask them here one by one

Most of the questions which you will be posting here ( difference
between char a[10], char *a;
#define square(a) ( (a) * (a) ) etc ) have already been discussed
number of times. People generally
don't like going over the same stuff again. That is exactly what faqs
are for. Get a copy of c.l.c
FAQ first and post here if you don't get the answers there.

**Jens Thoms Toerring** · Jun 27 '08, 07:40 PM

Re: Finally one site with all the C fundas

parag_paul@hotm ail.com <parag_paul@hot mail.comwrote:

Actually, most people dont dive into the intricacies given the amount
of time they want to spend. Thye want answer to get it done ( with
linux 4.2 , gcc is no more what it used to be )

What's "linux 4.2"? Do you mean gcc version 4.2? And what's so
bad about it? Also note that people here don't care too much
about specific compilers. An answer isn't taken to be correct
just because a certain compiler does allow something. Instead,
it has to correct in the sense that the C standard does not
contradict the answer.

so it is time that we come together and give the most plausible and
accurate answer to all the questions.

Well, that's exactly what the FAQ was written for. And sometimes,
unfortunately, plausible is not necessary accurate...

Hope you will help me. I will not put all the question all togetehr. I
will go over them individually, and get it checked

You can get it checked here;-) Just post a question and what you
think is the correct answer and have it shredded to pieces;-)
Just don't expect that seemingly simple questions always have
short, simple answers.
Regards, Jens
--
\ Jens Thoms Toerring ___ jt@toerring.de
\______________ ____________ http://toerring.de

**parag_paul@hotmail.com** · Jun 27 '08, 07:40 PM

Re: Finally one site with all the C fundas

On Jun 10, 1:59 pm, j...@toerring.d e (Jens Thoms Toerring) wrote:

parag_p...@hotm ail.com <parag_p...@hot mail.comwrote:

Actually, most people dont dive into the intricacies given the amount
of time they want to spend. Thye want answer to get it done ( with
linux 4.2 , gcc is no more what it used to be )

>
What's "linux 4.2"? Do you mean gcc version 4.2? And what's so
bad about it? Also note that people here don't care too much
about specific compilers. An answer isn't taken to be correct
just because a certain compiler does allow something. Instead,
it has to correct in the sense that the C standard does not
contradict the answer.
>

so it is time that we come together and give the most plausible and
accurate answer to all the questions.

>
Well, that's exactly what the FAQ was written for. And sometimes,
unfortunately, plausible is not necessary accurate...
>

Hope you will help me. I will not put all the question all togetehr. I
will go over them individually, and get it checked

>
You can get it checked here;-) Just post a question and what you
think is the correct answer and have it shredded to pieces;-)
Just don't expect that seemingly simple questions always have
short, simple answers.
Regards, Jens
--
\ Jens Thoms Toerring ___ j...@toerring.d e
\______________ ____________ http://toerring.de

Since in a recent update to QSCB we found out that since the gcc 3.3.6
was pretty lenient and we did not actively pursue compiling the tool
in AIX, we had many bugs in our code that gcc 3.3.6 allowed and we had
to do a thorough code check for all the stuff again

like

1. 3.3.6 allowing a strcuture element to be laid across word
boundaries
2. it allowed no return calls from void functios
3. non casting of the pointer from malloc was allowed.
4. it had relaxed semantics for linking too.

After this exercise, we had to purge many of the existing wrong code.
I accept the fact that yes we were sitting on a code base with
leniency in coding,
Now I really want to get across most of them

Also , I know that getting across puzzles from difference resources
and FAQ s is possible, but I have just tried to compile them in one
single place, without taking the ownership for any.

Please accept it as a museum for old questions that have been
discussed for a long time.

But , at point of time I am sure, I will have a much stronger
database. For C++ questions I am already getting feedback. I will
definitely incorporate them and I expect the same from you .

If the attempt was a flib. I will try to correct myself,.

**Jens Thoms Toerring** · Jun 27 '08, 07:40 PM

Re: Finally one site with all the C fundas

parag_paul@hotm ail.com <parag_paul@hot mail.comwrote:

Since in a recent update to QSCB we found out that since the gcc 3.3.6
was pretty lenient and we did not actively pursue compiling the tool
in AIX, we had many bugs in our code that gcc 3.3.6 allowed and we had
to do a thorough code check for all the stuff again

like

1. 3.3.6 allowing a strcuture element to be laid across word
boundaries

May be due to the machine having a different architecture where
the CPU/memory architecture doesn't allow unaligned accesses.

2. it allowed no return calls from void functios

Also gcc 3.3.6 would have spotted that if invoked with the '-W'
and '-Wall' option. I just can recommend not to use code that
doesn't compile cleanly with these flags set (except for things
that you have checked are safe) - saves a lot of time later
on as you found out the hard way;-)

3. non casting of the pointer from malloc was allowed.

Huh? Casting the return value of malloc() is never necessary
in C and casting only will hide a deeper problem, i.e. forget-
ting to include <stdlib.h>. Or are you talking about C++?

4. it had relaxed semantics for linking too.

But that's nothing related to the C (or C++) part.

Please accept it as a museum for old questions that have been
discussed for a long time.

But a few of them could use a work-over since what help is it
to others if they read (and perhaps believe) something which
is wrong?
Regards, Jens
--
\ Jens Thoms Toerring ___ jt@toerring.de
\______________ ____________ http://toerring.de

**parag_paul@hotmail.com** · Jun 27 '08, 07:40 PM

Re: Finally one site with all the C fundas

On Jun 10, 4:59 pm, j...@toerring.d e (Jens Thoms Toerring) wrote:

parag_p...@hotm ail.com <parag_p...@hot mail.comwrote:

Since in a recent update to QSCB we found out that since the gcc 3.3.6
was pretty lenient and we did not actively pursue compiling the tool
in AIX, we had many bugs in our code that gcc 3.3.6 allowed and we had
to do a thorough code check for all the stuff again
like
1. 3.3.6 allowing a strcuture element to be laid across word
boundaries

>
May be due to the machine having a different architecture where
the CPU/memory architecture doesn't allow unaligned accesses.
>

2. it allowed no return calls from void functios

>
Also gcc 3.3.6 would have spotted that if invoked with the '-W'
and '-Wall' option. I just can recommend not to use code that
doesn't compile cleanly with these flags set (except for things
that you have checked are safe) - saves a lot of time later
on as you found out the hard way;-)
>

3. non casting of the pointer from malloc was allowed.

>
Huh? Casting the return value of malloc() is never necessary
in C and casting only will hide a deeper problem, i.e. forget-
ting to include <stdlib.h>. Or are you talking about C++?
>

4. it had relaxed semantics for linking too.

>
But that's nothing related to the C (or C++) part.
>

Please accept it as a museum for old questions that have been
discussed for a long time.

>
But a few of them could use a work-over since what help is it
to others if they read (and perhaps believe) something which
is wrong?
Regards, Jens
--
\ Jens Thoms Toerring ___ j...@toerring.d e
\______________ ____________ http://toerring.de

hi Jens
I am very happy that you did not straight away bulldozed my idea
flat.
I am willing to make the changes , I will find out the appropriate
solutions and get all of them updated in time.
i am expecitn direct comments and personal search too.
We are not purging every wrong material online , are we?
But a chance to improve will definitely make this valuable with time.
I have been a part of google groups since it took over usenet, or ( in
proper lingo ) it acquired the old mails and discussions.
So I will preserve the sanctity of the whole setup and just ask help
from you.

**Jens Thoms Toerring** · Jun 27 '08, 07:40 PM

Re: Finally one site with all the C fundas

parag_paul@hotm ail.com <parag_paul@hot mail.comwrote:

I am very happy that you did not straight away bulldozed my idea
flat.

Well, I am not yet convinced that making up another C puzzle
site is a good idea. But it's your time you spend on it, so
it's not my place to tell you otherwise.

And you also must consider that every week or so someone posts
here, telling the world that (s)he created this wonderful new
website about C, C puzzles, C tutorials etc. and in at least 95
out of 100 case a quick look shows that a lot of things written
there are plain wrong - which could easily have been avoided by
the author at least reading the C FAQ. So you can't realistically
expect too warm a welcome when you come along here with another
website;-)

I am willing to make the changes , I will find out the appropriate
solutions and get all of them updated in time.
i am expecitn direct comments and personal search too.
We are not purging every wrong material online , are we?

I can't remove anything I didn't put up. I just thought, since
it was your stated intent to help others with your site, that
it would be probably most helpful not to present wrong answers
to start with.

But a chance to improve will definitely make this valuable with time.
I have been a part of google groups since it took over usenet, or ( in
proper lingo ) it acquired the old mails and discussions.

Google didn't "take over usenet" - they just bought the archives
DejaNews had collected over the years. But if Goggle would vanish
today the usenet would continue to exist - I haven't used Google
for sending a single message to a newsgroup, Google groups is
just one of many way you can get access to usenet.

So I will preserve the sanctity of the whole setup and just ask help
from you.

What "sancticity "? And if you just ask me then you do yourself
a disservice. I have been wrong quite a number of times and I
guess that won't change completely, just hopefully will happen
less often the more I learn. By posting here in clc you will
find a lot of keen eyes and if someone makes a mistake it will
be spotted more or less immediately and pointed out. There are
often also different angles from you can look at a problem, so
getting a second (and third and...) opinion often is rather
helpful. If you look back in thus thread I critisized some-
thing about an answer concerning a macro and the Keith came
along and added a number of good points I didn't come up with.

Regards, Jens
--
\ Jens Thoms Toerring ___ jt@toerring.de
\______________ ____________ http://toerring.de

**parag_paul@hotmail.com** · Jun 27 '08, 07:40 PM

Re: Finally one site with all the C fundas

On Jun 10, 8:16 pm, j...@toerring.d e (Jens Thoms Toerring) wrote:

parag_p...@hotm ail.com <parag_p...@hot mail.comwrote:

I am very happy that you did not straight away bulldozed my idea
flat.

>
Well, I am not yet convinced that making up another C puzzle
site is a good idea. But it's your time you spend on it, so
it's not my place to tell you otherwise.
>
And you also must consider that every week or so someone posts
here, telling the world that (s)he created this wonderful new
website about C, C puzzles, C tutorials etc. and in at least 95
out of 100 case a quick look shows that a lot of things written
there are plain wrong - which could easily have been avoided by
the author at least reading the C FAQ. So you can't realistically
expect too warm a welcome when you come along here with another
website;-)
>

I am willing to make the changes , I will find out the appropriate
solutions and get all of them updated in time.
i am expecitn direct comments and personal search too.
We are not purging every wrong material online , are we?

>
I can't remove anything I didn't put up. I just thought, since
it was your stated intent to help others with your site, that
it would be probably most helpful not to present wrong answers
to start with.
>

But a chance to improve will definitely make this valuable with time.
I have been a part of google groups since it took over usenet, or ( in
proper lingo ) it acquired the old mails and discussions.

>
Google didn't "take over usenet" - they just bought the archives
DejaNews had collected over the years. But if Goggle would vanish
today the usenet would continue to exist - I haven't used Google
for sending a single message to a newsgroup, Google groups is
just one of many way you can get access to usenet.
>

So I will preserve the sanctity of the whole setup and just ask help
from you.

>
What "sancticity "? And if you just ask me then you do yourself
a disservice. I have been wrong quite a number of times and I
guess that won't change completely, just hopefully will happen
less often the more I learn. By posting here in clc you will
find a lot of keen eyes and if someone makes a mistake it will
be spotted more or less immediately and pointed out. There are
often also different angles from you can look at a problem, so
getting a second (and third and...) opinion often is rather
helpful. If you look back in thus thread I critisized some-
thing about an answer concerning a macro and the Keith came
along and added a number of good points I didn't come up with.
>
Regards, Jens
--
\ Jens Thoms Toerring ___ j...@toerring.d e
\______________ ____________ http://toerring.de

and that is all the game
may be a discussion for one problem started in the past and it
reached the FAQ but not all reach to a all consenting point. There is
nothing like a consent from all. There is a language ans there is a
reference manual.
If things are vague and ambiguous then we need help otherwise it is
just plain english

Well, to get something done is the major challenge, which I once
forgot and took learning the language as a challenge and hence my
adventures with collection of questions that actually are just vistas
to understand the language better and at times the compiler you are
working on

Finally one site with all the C fundas

Finally one site with all the C fundas

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment