Yes, we want to make it illegal to have a metric size with "na" and, more
importantly, illegal to have inches with "iso", since mm to inches converts
to a non-ending fraction (which isn't good in names). We also want to make
sure that the entire name string can be treated as a name, namely, it has a
canonical representation for all names, so a Recipient can just compare for
equality with a list of supported/recognized names.
Tom
-----Original Message-----
From: Herriot, Robert
Sent: Tuesday, May 15, 2001 16:31
To: Bergman, Ron; Hastings, Tom N; Ira McDonald (E-mail); Don Wright
(E-mail)
Cc: 'ipp at pwg.org'
Subject: RE: IPP> Revised ABNF per Monday's Phone Conference
I'm concerned about the syntax having required redundancy of class1 with
inches and class2 with mm. Do we want to make it illegal to have a metric
size with "na". If so, we don't need "in".
Also, how does one describe a custom size, such as "7-13in". The grammar
doesn't allow "custom_7-13in". Instead it must be "na-custom_7-13in",
"oe-custom_7-13in" or "asme-custom_7-13in". This doesn't seem right.
Bob Herriot
> -----Original Message-----
> From: Bergman, Ron [mailto:Ron.Bergman at Hitachi-hkis.com]
> Sent: Tuesday, May 15, 2001 9:18 AM
> To: Tom Hastings (E-mail); Ira McDonald (E-mail); Don Wright (E-mail)
> Cc: 'ipp at pwg.org'
> Subject: IPP> Revised ABNF per Monday's Phone Conference
>>> Here is my proposed ABNF to document the agreed restrictions
> in yesterday's
> phone call. I may be missing some of the class names but this should
> correctly linl\k each class to only one set of units.
>>> media-size-self-describing-name =
> ( class1 "_" size-name "_" short-dim "-" long-dim "in" ) |
> ( class2 "_" size-name "_" short-dim "-" long-dim "mm" )
>> class1 = "na" | "asme" | "oe"
>> class2 = "iso" | "jis" | "jpn" | "prc" | "roc" | "om"
>> size-name = ( lowalpha | digit ) *( lowalpha | digit | "-" )
>> short-dim = dim
>> long-dim = dim
>> dim = integer-part [fraction-part] | "0" fraction-part
>> integer-part = non-zero-digit *digit
>> fraction-part = "." *digit non-zero-digit
>> lowalpha = "a" | "b" | "c" | "d" | "e" | "f" | "g" | "h" | "i" |
> "j" | "k" | "l" | "m" | "n" | "o" | "p" | "q" | "r" |
> "s" | "t" | "u" | "v" | "w" | "x" | "y" | "z"
>> non-zero-digit = "1" | "2" | "3" | "4" | "5" | "6" | "7" |
> "8" | "9"
>> digit = "0" | "1" | "2" | "3" | "4" | "5" | "6" | "7" |
> "8" | "9"
>>> Any comments?
>> Ron
>>