[TUHS] What is this 1972 C/NB program?
Dan Cross via TUHS
tuhs at tuhs.org
Fri May 15 00:27:05 AEST 2026
On Thu, May 14, 2026 at 10:24 AM Clem Cole via TUHS <tuhs at tuhs.org> wrote:
> Will. That was the original B/C syntax. Remember that B, like BLISS has
> only one data type: word. So pointers are not special. I suspect this is
> very early C, possibly nB given that it has a char type.
Also, PL/I used `|` and `&` for boolean operators, as shown in the
fragment Thalia initially posted; `II` and `&&` came a bit later.
- Dan C.
> On Thu, May 14, 2026 at 9:22 AM Will Senn via TUHS <tuhs at tuhs.org> wrote:
>
> > it's looking for hyphens as previously explained, but here are a couple
> > of notes:
> >
> > int nread 1;
> >
> > should probably be int nread = 1;
> >
> > static ibuf;
> >
> > should probably be static char *ibuf;
> >
> > or
> >
> > static int *ibuf;
> >
> > These look like bugs, but old c was quirky, so I wouldn't swear by it.
> >
> > Will
> >
> > On 5/13/26 20:57, Thalia Archibald via TUHS wrote:
> > > This mysterious program is in the s1 tape, but has not yet been
> > identified.
> > > Does it look familiar?
> > >
> > https://github.com/DoctorWkt/unix-jun72/blob/master/src/cmd/unknown.c
> > >
> > > It looks like it processes line continuations. It filters files to only
> > runs of
> > > lines continued with hyphen with the adjoining whitespace stripped. Only
> > letters
> > > and hyphens are allowed in such lines.
> > > [a-zA-Z-]+(-\n[\t\n]*[a-zA-Z-]+)*
> > >
> > > But there's bugs, so the grammar is actually the following, where EOF is
> > > included in the negated sets as NUL:
> > > ([a-zA-Z]|-[^\n])+(-\n[ \t\n]*[^ \t\n]([a-zA-Z]|-[^\n])*)*
> > >
> > > Could this be a sort of preprocessor? Perhaps for some sort of a
> > configuration
> > > language?
> > >
> > > It reads the list of named files, printing each filename with "%s:\n \n"
> > before.
> > > The space on an empty line is strange.
> > >
> > > It uses & and | for conditionals and is typed, characteristic of early C
> > and NB.
> > >
> > > It uses this horrid indentation style, which doesn't match ken or dmr.
> > > Do you recognize who? Example:
> > >
> > > while((b[++i] = get(ifile)) != 0)
> > > {if((b[i] >= 'a' & b[i] <= 'z') |
> > > (b[i] >= 'A' & b[i] <= 'Z'))
> > > {c[j++] = b[i];
> > > goto cont;
> > > }
> > >
> > > I've taken some liberties to simplify it below. I changed it to operate
> > on a
> > > single file. And equivalently, I reformatted it, replaced an unnecessary
> > buffer
> > > with a single char, and simplified control flow. See the above link for
> > the
> > > original.
> > >
> > > char c[60];
> > >
> > > main(argc, argv)
> > > int argc;
> > > char *argv[];
> > > {
> > > char b;
> > > int isw, j, k;
> > >
> > > isw = j = 0;
> > > while((b = getchar()) != 0) {
> > > if((b >= 'a' && b <= 'z') || (b >= 'A' && b <= 'Z')) {
> > > c[j++] = b;
> > > continue;
> > > }
> > > if(b == '-') {
> > > c[j++] = b;
> > > if((b = getchar()) != '\n') {
> > > c[j++] = b;
> > > continue;
> > > }
> > > if(j == 1) {
> > > isw = j = 0;
> > > continue;
> > > }
> > > isw = 1;
> > > while(((b = getchar()) == ' ') || (b == '\t')
> > || (b == '\n'))
> > > ;
> > > c[j++] = b;
> > > continue;
> > > }
> > > if(isw == 1) {
> > > k = 0;
> > > c[j++] = '\n';
> > > while(k < j)
> > > putchar(c[k++]);
> > > }
> > > isw = j = 0;
> > > }
> > > }
> > >
> > > Thalia
> > >
> > >
> >
More information about the TUHS
mailing list