Struct assignment, historical question

[William J. Leary Jr.]

I've got a history question, and several of you seem familiar with the history
of C, so perhaps someone knows.
The following:

#include<stdio.h>

struct checkit {
int a;
int b;
char c;
} wocka, tocka;

int main(void)
{
wocka.a = 10;
wocka.b = 20;
wocka.c = 'a';

printf("wocka = %d %d %c\n", wocka.a, wocka.b, wocka.c);

tocka = wocka;

printf("tocka = %d %d %c\n", tocka.a, tocka.b, tocka.c);

exit(0);
}

produces this for output:

wocka = 10 20 a
tocka = 10 20 a

That is, tocka = wocka; copied the entire contents of the structure.

But I'd swear that when I was working with C back in the '70's, this did not
work. To perform this copy, I had to do this:

tocka.a = wocka.a;
tocka.b = wocka.b;
tocka.c = wocka.c;

or

memcpy(&tocka, &wocka, sizeof(struct checkit));

Am I misremembering this, or was this actually the case? Or perhaps just a
problem with the Sun compiler I was using? I seem to have a vague recollection
of the original K&R saying that assignments wouldn't work, but likely would
some day.

- Bill

[Jack Klein]

On Mon, 5 Jun 2006 23:18:27 -0400, "William J. Leary Jr."
<Bi********@msn.com> wrote in comp.lang.c:

I've got a history question, and several of you seem familiar with the history
of C, so perhaps someone knows.
The following:

#include<stdio.h>

struct checkit {
int a;
int b;
char c;
} wocka, tocka;

int main(void)
{
wocka.a = 10;
wocka.b = 20;
wocka.c = 'a';

printf("wocka = %d %d %c\n", wocka.a, wocka.b, wocka.c);

tocka = wocka;

printf("tocka = %d %d %c\n", tocka.a, tocka.b, tocka.c);

exit(0);
}

produces this for output:

wocka = 10 20 a
tocka = 10 20 a

That is, tocka = wocka; copied the entire contents of the structure.

But I'd swear that when I was working with C back in the '70's, this did not
work. To perform this copy, I had to do this:

tocka.a = wocka.a;
tocka.b = wocka.b;
tocka.c = wocka.c;

or

memcpy(&tocka, &wocka, sizeof(struct checkit));

Am I misremembering this, or was this actually the case? Or perhaps just a
problem with the Sun compiler I was using? I seem to have a vague recollection
of the original K&R saying that assignments wouldn't work, but likely would
some day.

- Bill

The first 1978 edition of K&R mentioned that structure assignments and
the passing and returning of structures by value were common
extensions that were expected to become universal, but were not yet
allowed on all implementations.

--
Jack Klein
Home: http://JK-Technology.Com
FAQs for
comp.lang.c http://c-faq.com/
comp.lang.c++ http://www.parashift.com/c++-faq-lite/
alt.comp.lang.learn.c-c++
http://www.contrib.andrew.cmu.edu/~a...FAQ-acllc.html

[Suman]

William J. Leary Jr. wrote:

I've got a history question, and several of you seem familiar with the history
of C, so perhaps someone knows.
[ ... cnip code ]
Am I misremembering this, or was this actually the case? Or perhaps just a
problem with the Sun compiler I was using? I seem to have a vague recollection
of the original K&R saying that assignments wouldn't work, but likely would
some day.

No. Struct assignment came in with the ANSI standard.
1. K&R 2: Preface: Paragraph 3: Line 1
2. K&R 2: Chapter 6: Paragraph 3: Line 1

[William J. Leary Jr.]

> On Mon, 5 Jun 2006 23:18:27 -0400, "William J. Leary Jr."

<Bi********@msn.com> wrote in comp.lang.c:

I've got a history question, and several of you seem familiar with the history of C, so perhaps someone knows.
((..omitted..))
Am I misremembering this, or was this actually the case? Or perhaps just a
problem with the Sun compiler I was using? I seem to have a vague recollection of the original K&R saying that assignments wouldn't work, but likely would
some day.

"Jack Klein" <ja*******@spamcop.net> wrote in message
news:dr********************************@4ax.com... The first 1978 edition of K&R mentioned that structure assignments and
the passing and returning of structures by value were common
extensions that were expected to become universal, but were not yet
allowed on all implementations.
.... and ...

"Suman" <sk*****@gmail.com> wrote in message
news:11**********************@u72g2000cwu.googlegr oups.com... No. Struct assignment came in with the ANSI standard.
1. K&R 2: Preface: Paragraph 3: Line 1
2. K&R 2: Chapter 6: Paragraph 3: Line 1

When I remember where my 1978 K&R is, I'll look it up. But you've both just
settled and argument I was having here.

Thank you both.

- Bill

[Suman]

William J. Leary Jr. wrote:
[ snip ]

"Suman" <sk*****@gmail.com> wrote in message
news:11**********************@u72g2000cwu.googlegr oups.com...

No. Struct assignment came in with the ANSI standard.
1. K&R 2: Preface: Paragraph 3: Line 1
2. K&R 2: Chapter 6: Paragraph 3: Line 1

When I remember where my 1978 K&R is, I'll look it up. But you've both just
settled and argument I was having here.

Will you be able to find the lines I mentioned? I doubt. You'd
probably need one a good ten years younger :)

[SuperKoko]

William J. Leary Jr. wrote:

I've got a history question, and several of you seem familiar with the history
of C, so perhaps someone knows.
The following:

#include<stdio.h>

struct checkit {
int a;
int b;
char c;
} wocka, tocka;

int main(void)
{
wocka.a = 10;
wocka.b = 20;
wocka.c = 'a';

printf("wocka = %d %d %c\n", wocka.a, wocka.b, wocka.c);

tocka = wocka;

printf("tocka = %d %d %c\n", tocka.a, tocka.b, tocka.c);

exit(0);
}

produces this for output:

wocka = 10 20 a
tocka = 10 20 a

That is, tocka = wocka; copied the entire contents of the structure.

But I'd swear that when I was working with C back in the '70's, this did not
work. To perform this copy, I had to do this:

tocka.a = wocka.a;
tocka.b = wocka.b;
tocka.c = wocka.c;

or

memcpy(&tocka, &wocka, sizeof(struct checkit));

Am I misremembering this, or was this actually the case? Or perhaps just a
problem with the Sun compiler I was using? I seem to have a vague recollection
of the original K&R saying that assignments wouldn't work, but likely would
some day.

Strictly speaking, K&R C had no structure assignment.
"Classic C" is a term used for pre-standard C, including structure
assignment, enumerations and void (but not void*).
C89 adopted void*, const and prototypes from C++.

I think that the following links are revelant:
http://cm.bell-labs.com/cm/cs/who/dmr/cchanges.pdf
More info on Dennis Ritchie web site : http://cm.bell-labs.com/who/dmr/
Interesting articles by Bjarne Stroustrup :
http://www.research.att.com/~bs/siblings_short.pdf
and http://www.research.att.com/~bs/sibling_rivalry.pdf

And a brief history of the C language by Dennis Ritchie :
http://cm.bell-labs.com/cm/cs/who/dmr/chist.html

Dennis' home page is full of interesting information on the C and B
languages and on UNIX.

[lawrence.jones]

Suman <sk*****@gmail.com> wrote:

No. Struct assignment came in with the ANSI standard.

On the contrary, struct parameters, return values, and assignment were
added to the Unix C compilers just about the time that K&R was actually
published (the "Recent Changes to C" paper noting them is dated Nov. 15,
1978), leading to a great deal of frustration many years later when
MS-DOS compilers appeared that implemented K&R exactly and thus didn't
support them.

-Larry Jones

He doesn't complain, but his self-righteousness sure gets on my nerves.
-- Calvin