[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: [mizar] Question on Mizar grammar





Hi Michael,

The article funct_2.miz contains the following fragment (lines 933 - 937):

-------------------------------
registration let X;
 let f,g be quasi_total PartFunc of X,X;
 cluster g*f -> quasi_total PartFunc of X,X;
 coherence proof X <> {} or X = {}; hence thesis by Th19; end;
end;
-------------------------------

but the current production rule for functorial registration doesn't allow type
expression:

Functorial-Registration =
	"cluster" Term-Expression "->" Adjective-Cluster ";" Correctness-Conditions .

It seems this rule must be like the following (though I am not sure about it).

Functorial-Registration =
	"cluster" Term-Expression "->" Adjective-Cluster [ Type-Expression] ";"
Correctness-Conditions .

Can anyone confirm it?

Yes, I think you are right - the optional Type was added there quite recently. Sometimes (like in this case) it may help to check the XML specification (the newest comes with the Mizar distro), some version is at http://lipa.ms.mff.cuni.cz/~urban/Mizar.html#FCluster.FCluster . But generally the two specifications do not correspond to each other tightly - they describe different Mizar layers.

Josef