wayne wrote:
* The definition for the unknown-modifier has been changed
slightly to note that unknown modifiers MUST NOT match
"redirect" and "exp".
Ugh.
The differene between <domain-spec> and <macro-string> is
apparently that the former has a trailing <domain-end>.
So for redirect= and exp= you insist on a trailing <toplabel>
or <macro-var> (and syntactically the latter can be nonsense).
In other words better stay from this idea, you can't catch
all bogus redirect= and exp= in syntax. With our new syntax
you'd only catch *( DIGIT / "." ) or similar cases.
I have created a series of ANBF terms in the form of
range$i-$j
Okay, let's first get that ready, without touching the other
problem, I think it's fine.
/ "1" DIGIT ; 10 .. 19
/ "2" DIGIT ; 20 .. 29
How about / (( "1" / "2" ) DIGIT ) ; 10 .. 29
/ "3" %x30-32 ; 30 .. 32
Dito / "30" / "31 " / 32" ; 30 .. 32
/ "10" DIGIT ; 100 .. 109
/ "11" DIGIT ; 110 .. 119
Dito / (( "10" / "11" ) DIGIT ) ; 100 .. 119
/ "12" %x30-38 ; 120 .. 128
Recommended grouping / ( "12" %30-38 ) ; 120 .. 128
/ range1-9 DIGIT ; 10 .. 99
Dito / ( range1-9 DIGIT ) ; 10 .. 99
/ "1" 2DIGIT ; 100 .. 199
Dito / ( "1" 2DIGIT ) ; 100 .. 199
/ "2" %x30-34 DIGIT ; 200 .. 249
Dito / ( "2" %x30-34 DIGIT ) ; 200 .. 249
/ "25" %x30-35 ; 250 .. 255
Dito / ( "25" %x30-35 ) ; 250 .. 255
toplabel = ALPHA / ALPHA *[ alphanum / "-" ] alphanum
toplabel = ALPHA [ *( alphanum / "-" ) alphanum ]
IIRC Bill's validator doesn't not like *[ x ] instead of *( x )
Bye, Frank