19 Feb
2022
19 Feb
'22
9:43 p.m.
I have been poring through the v7 source code lately, and came across an
oddity I would like to know more about. Specifically, in sh. The code
for sh is c, but it makes *extensive* use of of macros, for example:
/usr/src/cmd/sh/word.c
...
WHILE (c=nextc(0), space(c)) DONE
...
The macros for sh are defined in mac.h:
/usr/src/cmd/sh/mac.h
...
#define BEGIN {
#define END }
#define SWITCH switch(
#define IN ){
#define ENDSW }
#define FOR for(
#define WHILE while(
#define DO ){
#define OD ;}
#define REP do{
#define PER }while(
#define DONE );
...
I can read the resultant code through the lens of my experience coding
c, but I'm curious why the macros and how this came about? In v6, the sh
source is straight up c. Is there a story behind it worth knowing?
Thanks,
Will
19 Feb
19 Feb
10:03 p.m.
On Sat, 19 Feb 2022, Will Senn wrote:
I have been poring through the v7 source code lately,
and came across an
oddity I would like to know more about. Specifically, in sh. The code for sh
is c, but it makes *extensive* use of of macros, for example:
<snip>
I can read the resultant code through the lens of my
experience coding c, but
I'm curious why the macros and how this came about? In v6, the sh source is
straight up c. Is there a story behind it worth knowing?
Apparently Bourne was heavily into ALGOL, and used those macros to make C
into something more familiar.
At least, that's what I concluded by reading her Wikipedia page as well as
the code.
-uso.
10:07 p.m.
On Sat, Feb 19, 2022 at 11:04 AM Steve Nickolas <usotsuki(a)buric.co> wrote:
Apparently Bourne was heavily into ALGOL,
That's sort of an understatement. I believe that when he was at Cambridge,
he was one of the people that helped to take the Algol-X proposal and
turned it into the Algol-68 definition. I also believe he worked on their
famous implementation of same.
11:24 p.m.
This persisted in the shell until S5R2 when someone gratefully undid all those macros.
adb was similarly afflicted.
On Feb 19, 2022, at 11:04, Steve Nickolas
<usotsuki(a)buric.co> wrote:
On Sat, 19 Feb 2022, Will Senn wrote:
I have been poring through the v7 source code
lately, and came across an oddity I would like to know more about. Specifically, in sh.
The code for sh is c, but it makes *extensive* use of of macros, for example:
<snip>
I can read the resultant code through the lens of
my experience coding c, but I'm curious why the macros and how this came about? In
v6, the sh source is straight up c. Is there a story behind it worth knowing?
Apparently Bourne was heavily into ALGOL, and used those macros to make C into something
more familiar.
At least, that's what I concluded by reading her Wikipedia page as well as the
code.
-uso.
10:04 p.m.
Sure - this is lovingly referred to as 'Bourne-gol.' Steve Bourne was part
of the Algol-68 development team at Cambridge before he came to BTL.
Besides his being the 'editor' / 'release manager' of V7 (as Dennis is
said
to exclaimed 'never again' after V6), please also look at the sources to
adb, srb's other major contribution (it is also written using his same
macros). It's also why the syntax of his shell is basically Algol syntax.
Clem
On Sat, Feb 19, 2022 at 10:44 AM Will Senn <will.senn(a)gmail.com> wrote:
I have been poring through the v7 source code lately,
and came across an
oddity I would like to know more about. Specifically, in sh. The code for
sh is c, but it makes *extensive* use of of macros, for example:
/usr/src/cmd/sh/word.c
...
WHILE (c=nextc(0), space(c)) DONE
...
The macros for sh are defined in mac.h:
/usr/src/cmd/sh/mac.h
...
#define BEGIN {
#define END }
#define SWITCH switch(
#define IN ){
#define ENDSW }
#define FOR for(
#define WHILE while(
#define DO ){
#define OD ;}
#define REP do{
#define PER }while(
#define DONE );
...
I can read the resultant code through the lens of my experience coding c,
but I'm curious why the macros and how this came about? In v6, the sh
source is straight up c. Is there a story behind it worth knowing?
Thanks,
Will
10:06 p.m.
https://research.swtch.com/shmacro
On Feb 19, 2022, at 7:44 AM, Will Senn
<will.senn(a)gmail.com> wrote:
I have been poring through the v7 source code lately, and came across an oddity I would
like to know more about. Specifically, in sh. The code for sh is c, but it makes extensive
use of of macros, for example:
/usr/src/cmd/sh/word.c
...
WHILE (c=nextc(0), space(c)) DONE
...
The macros for sh are defined in mac.h:
/usr/src/cmd/sh/mac.h
...
#define BEGIN {
#define END }
#define SWITCH switch(
#define IN ){
#define ENDSW }
#define FOR for(
#define WHILE while(
#define DO ){
#define OD ;}
#define REP do{
#define PER }while(
#define DONE );
...
I can read the resultant code through the lens of my experience coding c, but I'm
curious why the macros and how this came about? In v6, the sh source is straight up c. Is
there a story behind it worth knowing?
Thanks,
Will
10:57 p.m.
Thanks Bakul, that's interesting.
Impetus for the obfuscated c contest?! The same contest where a guy
wrote a pdp11 emulator capable of running v0, v6, and 2.9bsd in 50 lines
of c (Mills, 2018)? I still get a kick out of that code. Particularly
because you can run Mullender's 1984 entry on the 2.9 bsd instance it
hosts (it also works in v7, but it's too fast, even with stty 300).
Thankfully, macros, while adding an additional layer of complexity,
aren't as impenetrable as the ioccc stuff:
mills's prog.c snippet:
struct timeval F,G; struct termios H,U={ T} ; enum{
N=64,a=N<<7,b=a-1,c=a*32,d
=c-1, e=c/ 2,f= a*2, g=a/2,h =g/2,j =h/ 2,Q=V*j*5} ;
char*s=P,K,M;
int* p, l[ a] ,m,n,J,o=A,
O=j,E,R,i,k,t,r,q,u,v,w,x,y,z
,B,C, *D,Z ;int main (){ for(D=mmap...
mullender's mullender.c in it's entirety (requires v7 or better and
pdp11):
short main[] = {
277, 04735, -4129, 25, 0, 477, 1019, 0xbef, 0, 12800,
-113, 21119, 0x52d7, -1006, -7151, 0, 0x4bc, 020004,
14880, 10541, 2056, 04010, 4548, 3044, -6716, 0x9,
4407, 6, 5568, 1, -30460, 0, 0x9, 5570, 512, -30419,
0x7e82, 0760, 6, 0, 4, 02400, 15, 0, 4, 1280, 4, 0,
4, 0, 0, 0, 0x8, 0, 4, 0, ',', 0, 12, 0, 4, 0, '#',
0, 020, 0, 4, 0, 30, 0, 026, 0, 0x6176, 120, 25712,
'p', 072163, 'r', 29303, 29801, 'e'
};
Later,
Will
On 2/19/22 10:06 AM, Bakul Shah wrote:
https://research.swtch.com/shmacro
> On Feb 19, 2022, at 7:44 AM, Will Senn <will.senn(a)gmail.com> wrote:
>
> I have been poring through the v7 source code lately, and came
> across an oddity I would like to know more about. Specifically, in
> sh. The code for sh is c, but it makes *extensive* use of of macros,
> for example:
>
> /usr/src/cmd/sh/word.c
> ...
> WHILE (c=nextc(0), space(c)) DONE
> ...
>
> The macros for sh are defined in mac.h:
>
> /usr/src/cmd/sh/mac.h
> ...
> #define BEGIN {
> #define END }
> #define SWITCH switch(
> #define IN ){
> #define ENDSW }
> #define FOR for(
> #define WHILE while(
> #define DO ){
> #define OD ;}
> #define REP do{
> #define PER }while(
> #define DONE );
> ...
>
> I can read the resultant code through the lens of my experience
> coding c, but I'm curious why the macros and how this came about? In
> v6, the sh source is straight up c. Is there a story behind it worth
> knowing?
>
> Thanks,
>
> Will
820
days inactive
820
days old
7 comments
5 participants
participants (5)
-
Bakul Shah -
Clem Cole -
Ron Natalie -
Steve Nickolas -
Will Senn