Jeremy, thanks for the education. If there is a more appropriate forum for this discussion, please direct me. --tk
On Sun, Mar 1, 2009 at 10:36 PM, Jeremy Carroll <[email protected]>wrote: > > > I'll let Scott or Holger answer about how best to use this URI in TBC, but > I'll comment on the URI itself: > > > http://cpe.mitre.org/files/cpe-specification_2.1.pdf > [[ > A CPE Name is a percent-encoded URI with each name starting with the prefix > (the URI > scheme name) “cpe:”. Note that the scheme "cpe:" is not registered as an > official URI scheme > with IANA. > ]] > > Thus cpe: is a scheme name, *not* a namespace prefix. > > & thus when percent-encoding your example, you were over-enthusiastic. > Definitely the first, and second items should not be percent encoded. > > cpe:/a%3Aapache%3Ahttp_server%3A1.3.30 > > In fact, pasting the unencoded URI into composer (current developer build), > as a URI seems OK, and my reasonably well-honed intuitions about what is and > what isn't a legal URI suggest that too. > > cpe:/a:apache:http_server:1.3.30 > > However, and this is feedback that probably should be sent to CPE/NIST, > (and I'm cc-ing Evan Wallace, on the SemWeb side), this particular format > with the version number at the end, is problematic with the semantic web > qname convention, since there is no appropriate rightmost NCName (I haven't > checked about the '.'s, I don't think they are NCStartChar's but not totally > sure). > > [Evan feel free to respond either on some other list, please CC me, or > assuming you don't have post permission, if not I can forward any response > to this list] > > This means that you cannot use this URI as a property name in RDF/XML, and > using it as a class name may hit bugs in RDF libraries such as Jena. > > Thus I strongly recommend that all projects using such URIs exclusively use > N3 format. > Given the use of qnames in the TBC UI, Holger may need to advise as to > whether that would be sufficient for successful use within TBC. > > Further note: while the percent-encoded form does end in an NCName A1.3.30, > the left hand side: > cpe:/a%3Aapache%3Ahttp_server%3 is not a URI (the bad % escape at the end) > and hence cannot be a namespace name. I expect the namespace split point > code in Jena (which I believe TBC is reliant on) gets this example wrong, > (unless it has been improved since I left the Jena team). > > It is fairly hard to predict how several of the layers of software will > react to these URIs. > > One work-around, might be to always use a blank node for these things, and > then a datatype property, with range xsd:anyURI to link to the cpe URI. This > prevents any suggestion that the namespace split point algorithm should or > could be used for these URIs, at the cost of distorting your modeling. > > Jeremy > > > > > > > -----Original Message----- > > From: [email protected] [mailto: > topbraid-composer- > > [email protected]] On Behalf Of TK > > Sent: Sunday, March 01, 2009 7:56 PM > > To: TopBraid Composer Users > > Subject: [tbc-users] NIST Common Platform Enumeration > > > > > > Hello all, > > I'm doing some modeling using an emerging standards from MITRE called > > CPE. (http://cpe.mitre.org/) > > In short, they hope to name all applications with a unique URI. I > > have a question about using a very complex name in TBC. > > > > The names in CPE are based on URI's but when placed into RDF, I > > believe I am going to have to do a lot of %-escaping. Let me start > > with an example. > > > > The CPE name representing the Apache HTTP Server version 1.3.30 is: > > cpe:/a:apache:http_server:1.3.30 > > > > If I were to create a CPE class in TBC and then go to create an > > instance where the name reads 'cpe:/a:apache:http_server:1.3.30' it > > would not fly. The only way I can get it in to TBC is to escape like > > so: > > cpe%3A%2Fa%3Aapache%3Ahttp_server%3A1.3.30 > > > > When having to deal with complex names, especially RDF SUBJECTS which > > must be in the form of a RDF URI Reference, what is the best practice > > with TBC users? I guessing that there is no other choice but to %- > > escape. > > > > Thanks, > > --tk > > > > > > > > > > > > > > > -- "The nervous system organizes itself so as to compute a stable reality" - Maturana & Varela --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "TopBraid Composer Users" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/topbraid-composer-users?hl=en -~----------~----~----~----~------~----~------~--~---
