M4 language not showing up in PrimeView

[rzuckerm]

The original boolean M4 language PR was merged last week, and I noticed that M4 was added in PrimeView. However, I'm not seeing M4 showing up in the Language filter. How long does it usually take before new languages show up?

Could the fact this the "name" field is in lowercase have anything to do with this?

[rbergen]

First off: well spotted!

For the language to show up in the filter list, there has to be at least one entry for the language in question in the benchmark results (i.e. it's not enough for the language to be included in the language map file).

I've run a benchmark of just the M4 directory, with this output:

$ make DIRECTORY=PrimeM4
info: Unconfined mode: false
info: Detected architecture: amd64
info: [PrimeM4][solution_1] Building...
info: [PrimeM4][solution_1] Running...
error: No data was collected!

So, that explains the absence in the results or the language filter. Which makes the next question why the solution's results don't show up in the report. The reason is simple, but also subtle enough for me to also overlook it when I reviewed the solution. It relates to the output format:

$ docker run m4-1
Passes: 2, Time: 6.12, Avg: 0.326, Count: 78498, Valid: true
rzuckerm-m4;2;6.12;algorithm=base;faithful=no

Passes: 1, Time: 6.68, Avg: 0.149, Count: 78498, Valid: true
rzuckerm-m4-bit;1;6.68;algorithm=base;faithful=no,bits=1

The problem is that:

• The thread count is missing.
• There is a semicolon between the algorithm and faithful tags, where it has to be a comma.

The correct output would have been:

Passes: 2, Time: 6.12, Avg: 0.326, Count: 78498, Valid: true
rzuckerm-m4;2;6.12;1;algorithm=base,faithful=no

Passes: 1, Time: 6.68, Avg: 0.149, Count: 78498, Valid: true
rzuckerm-m4-bit;1;6.68;1;algorithm=base,faithful=no,bits=1

If you can put a PR together to fix this, I'll merge it expeditiously.

[rzuckerm]

PR #1013

[rbergen]

Thanks. It'll have to be updated with one more change: the semicolon (;) between "algorithm=base" and what follows needs to be replaced with a comma (,). All tags must be comma-separated.

[rzuckerm]

Sorry about that. I seem to be having trouble getting the format. Hopefully, it's right now?

[rbergen]

I think it is now. I'll confirm that locally.

[rbergen]

$ make DIRECTORY=PrimeM4
info: Unconfined mode: false
info: Detected architecture: amd64
info: [PrimeM4][solution_1] Building...
info: [PrimeM4][solution_1] Running...
                                                          Single-threaded
┌───────┬────────────────┬──────────┬─────────────────┬────────┬──────────┬─────────┬───────────┬──────────┬──────┬───────────────┐
│ Index │ Implementation │ Solution │ Label           │ Passes │ Duration │ Threads │ Algorithm │ Faithful │ Bits │ Passes/Second │
├───────┼────────────────┼──────────┼─────────────────┼────────┼──────────┼─────────┼───────────┼──────────┼──────┼───────────────┤
│   1   │ m4             │ 1        │ rzuckerm-m4     │   2    │ 6.04000  │    1    │   base    │    no    │      │    0.33113    │
│   2   │ m4             │ 1        │ rzuckerm-m4-bit │   1    │ 6.82000  │    1    │   base    │    no    │ 1    │    0.14663    │
└───────┴────────────────┴──────────┴─────────────────┴────────┴──────────┴─────────┴───────────┴──────────┴──────┴───────────────┘

I'd expect some M4 results in PrimeView within the next 24 hours. :)

[rbergen]

There it is:

The two implementations are indeed also in the results list.

It showed up in this report first, which was the first benchmark run after the format changes were merged.

[rzuckerm]

Great! Thanks @rbergen ! However, should that be M4 instead of m4?

[rzuckerm]

Actually, no I guess it shouldn't. I just assumed that the m was capitalized. Both Wikipedia and GNU spell it that way.

[rbergen]

Yeah, I did look at that. :)