25

A 1980 paper called "Register Allocation via Coloring" contains something that looks like source code:

P: PROC(MODE);

 DCL

    MODE     BIT(1),

    (A1,A2,A3,A4,A5,A6,A7,A8,A9,A10,
     B1,B2,B3,B4,B5,B6,B7,B8,B9,B10,
     SUM)         FIXED BIN(15) AUTO,

    (U(10),V(10)) FIXED BIN(15) STATIC EXT;

  IF MODE
     THEN DO;
        A1=U(1); A2=U(2); A3=U(3); A4=U(4); A5=U(5);
        A6=U(6); A7=U(7); A8=U(8); A9=U(9); A10=U(10);
     END;
     ELSE DO;
        B1=V(1); B2=V(2); B3=V(3); B4=V(4); B5=V(5);
        B6=V(6); B7=V(7); B8=V(8); B9=V(9); B10=V(10);
     END;

  LABEL:;

  IF MODE

Original image

I couldn't find what programming language this is. Is it a real historical programming language, or is it some type of pseudo-code?

The beginning of the paper mentions that the research is based on a PL/I Compiler. So maybe it's the programming language that the compiler was built on. I couldn't locate the paper where it talks about the compiler on the Internet.

1
  • 2
    @AndrewT - (edits to title) LABEL in this context (before a colon) is not a keyword, it's a label. It would be a keyword in a LABEL declaration. That's one of the 'interesing' things about PL/I: keywords are not reserved words.
    – dave
    May 13, 2020 at 11:55

2 Answers 2

56

It's PL/I, promoted by IBM as the successor to FORTRAN, Algol 60, and COBOL. That's actual code as far as I recall, not pseudocode.

PL/I had abbreviations for keywords; "DCL" is "DECLARE".

It starts off by defining a procedure (routine) named P with one parameter, named MODE, which is a Boolean variable - BIT(1), then defines a bunch of 16-bit signed integers allocated on the stack, and some statically-allocated arrays.

5
  • 1
    Comments are not for extended discussion; this conversation has been moved to chat.
    – Chenmunka
    May 12, 2020 at 17:06
  • 1
    Also conceivable that it's PL/M, Gary Kildall's microcomputer version of PL/I. But given that they mention PL/I in the paper, it seems likely that it's PL/I.
    – user4766
    May 12, 2020 at 19:17
  • 4
    Given that all the authors worked for IBM at the Watson Research Center, I'd say it's certain it's PL/I.
    – dave
    May 12, 2020 at 23:00
  • The PL/M I used was the Intel variant. It didn't have any abbreviated keywords. Certainly not anything that looked like this.
    – cup
    May 16, 2020 at 15:54
  • PL/1 was weird - it always generated an executable, even when there were compilation errors. You'd wonder why your program was giving the wrong results and then discover that there was a compilation error which caused the logic to go wrong.
    – cup
    May 20, 2020 at 8:43
0

I know that SAS statistical software utilises those three things. I guess you would call it a pseudocode?

PROC is a command for 'ready to use' procedures, like proc tables, proc univariate, etc.

Using a LABEL statement in a DATA step permanently associates labels with variables by affecting the descriptor information of the SAS data set that contains the variables.

The DCL (DECLARE) statement declares a variable of any SCL (SAS Component Language) data type.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .