Re: BNF for configuration language?

[Larry Dickson <tjoccam@xxxxxxxxxxx>]

(Sigh) It was in my hands all the time, only 6 pages long, so I kept missing it when flicking through the pages. Appendix E of occam 2 Toolset User Manual part 2. Including the important statement: "This syntax should be considered as extending the syntax of occam."

By the way, the occam syntax that I posted last year (in Details-occam-parse of github.com/SpaceSciencesCorp/Hard-Stuff-Language)  proved to have a typo: the definition of dyadic_operator needs to add '=' (the Boolean-valued equals operator). So the old one was consistent but not complete. The fix added only one state to yacc's y.output. I beat it up with a big, complicated occam program and it had no problem. So if anyone is using my OCCAMLIB.Y file, add

| '='

at the end of dyadic_operator BNF definition.

Larry

On Jun 4, 2019, at 7:05 PM, Larry Dickson <tjoccam@xxxxxxxxxxx> wrote:

All,

Here's another question in our quest for rebuilding the circa-1990 occam/Transputer state of affairs as a foundation for other things.

Is there an "official" modified Backus-Naur Form syntax for the occam/Transputer configuration language? This would be analogous to the BNF of section G.2 of the occam 2 Reference Manual (p 86-90), but would refer to the input to occonf, as described in chapter 5 of the Toolset User Manual, part 1.

I could invent one that would fit chapter 5 and the experimental behavior of occonf, but it would be better if it came from an authoritative source. Is it a superset of standard occam, or is #USE file.lku required as the source of the inner code?

Larry Dickson