Date:Wed, 24 May 2006 09:37:38 -0400
Reply-To:Encoded Archival Description List <[log in to unmask]>
Sender:Encoded Archival Description List <[log in to unmask]>
From:Stephen Yearl <[log in to unmask]>
Subject:Re: Question about PUBLIC identifier
Comments:To: Encoded Archival Description List <[log in to unmask]>
In-Reply-To:<[log in to unmask]>
Content-Type:multipart/alternative;
Michele Rothenberger <[log in to unmask]> wrote (in part):
>>Your "beebe.sgm" then is actually not part
>>of the public identifier; it maps the public identifier to the instance
>>that resides on your local system and is actually a system id.
Correct. And the W3C reccomendation for XML 1.0 also indicates that
resolution of the referenced resource by the XML parser can be through
either the system id, combination of FPI and system identifier, or
mechanisms external; but that IF the system identifier is present the XML
parser must use this to resolve the resource. This is clearly intended for
resolving entities declared in the DOCTYPE, however-- no XML parser has
knowledge that Susan's <eadid> contains a public/system id coupling.
My reading of ISO/IEEC 9070:1991 suggests that FPIs are composed of two
sections:
object owner and object name, separated by '//'. The three-part format looks
more like the earlier specification ISO 8879:1986. But then again, I don't
have a complete copy of either spec.
St.
Stephen Yearl
Systems Archivist
Yale University Library::Manuscripts and Archives
On 5/24/06, Michele Rothenberger <[log in to unmask]> wrote:
>
> Hi Susan --
>
> Stephen is correct but perhaps you were also asking about the
> legitimacy of your public identifier per se, beyond the confines of EAD?
> ISO 9070:1991 provides the rules for defining ISO assigned public
> identifiers. As I understand it, ISO assigned public identifiers have
> three components, separated from each other by a double slash. The three
> components are:
>
> * an ISO owner identifier (which may be ISO or some other owner -
> in your case Penn State)
> * a public text class (e.g. ENTITY) and an associated public text
> description (e.g. Added Latin 1)
> * a public text language code (e.g. EN for English).
>
> Your public identifier as it stands is this:
>
> -//Pennsylvania State University::University Libraries::Special
> Collections Library
> //TEXT (US::PSt::MGN324::Helen Hulick Beebe Papers, 1927-1998
> //EN
>
> which looks to me like it has all the required elements (though are you
> missing a closing paren?). Your "beebe.sgm" then is actually not part
> of the public identifier; it maps the public identifier to the instance
> that resides on your local system and is actually a system id (or sysid,
> as opposed to pubid). So in addition to being able to drop the
> "beebe.sgm" without violating EAD, you can also drop it without
> violating ISO standards :) Whether or not you might want a sysid
> included, as a way of mapping the public identifier to the appropriate
> local file, is another question.
>
> I second Stephen's suggested way of encoding. The EAD tag library also
> mentions that COUNTRYCODE and MAINAGENCYCODE are required to make the
> <eadid> compliant with ISAD(G) (see
> http://www.loc.gov/ead/tglib/elements/eadid.html) so you might want to
> add those as well.
>
> Michele
>
>
> -=--=--=--=--=--=--=--=--=--=--=--=--=--=-
> Michele Rothenberger
> Special Collections Research Center
> Syracuse University Library
> 222 Waverly Avenue
> Syracuse, NY 13244
> (315) 443-2697
> -=--=--=--=--=--=--=--=--=--=--=--=--=--=-
>
> >>> [log in to unmask] 5/24/2006 8:22 AM >>>
> Susan:
>
> Short answer is no, it doesn't matter. The <eadid> is merely meant to
> provide a globally unique string of characters that identify your
> finding
> aid. Since you are using EAD 2002, I would suggest however that you
> consider
> something like:
>
> <eadid publicid="-//Pennsylvania State University::University
> Libraries::Special Collections Library//TEXT (US::PSt::MGN324::Helen
> Hulick
> Beebe Papers, 1927-1998//EN">beebe.WHATEVER</eadid>
>
> Or even substitute the publicid attribute with 'url' and reference the
> URL
> of the final location of the instance document.
>
> St.
>
> Stephen Yearl
> Systems Archivist
> Yale University Library::Manuscripts and Archives
>
> On 5/23/06, Susan Hamburger <[log in to unmask]> wrote:
> >
> > We are planning to export finding aids from a database into
> EAD-tagged XML
> > documents and ingest them into DLXS for federated searching. The
> techies
> > want to know if the entity declaration in the PUBLIC identifier that
> I
> > currently use for XMetaL-created finding aids is necessary. I
> currently
> > create my finding aids in an XMetaL template as SGML files then
> convert
> > them to XML with NoteTabPro, run the Saxon frameset to create HTML
> > versions
> > for display on the Web, and run Saxon jvm to create PDFs for
> printable
> > copies. All this conversion will be moot when we ingest the XML
> document
> > into DLXS.
> >
> > In the example below, is "beebe.sgm" necessary or could it be
> "beebe.xml"
> > or simply "beebe"? The techie is changing the name to "beebe1"
> without a
> > .xml or .sgm extension before ingesting it into DLXS and it seems to
> work
> > just fine.
> >
> > <?xml version="1.0" encoding="ISO-8859-1"?>
> > <!DOCTYPE ead SYSTEM "file:/c:\pubclass\shared\ead\ead.dtd" [
> > <!ENTITY psulogo PUBLIC "-//Pennsylvania State
> University::University
> > Libraries::Special Collections Department//NONSGML (PSU logo)//EN"
> > "file:/C:\pubclass\documents\psulogo.gif" NDATA gif>
> > ]>
> > <ead
> > ><eadheader
> > langencoding="iso639-2b"
> > scriptencoding="iso15924"
> > dateencoding="iso8601"
> > countryencoding="iso3166-1"
> > repositoryencoding="iso15511"
> > ><eadid
> > >PUBLIC "-//Pennsylvania State University::University
> > Libraries::Special Collections Library//TEXT (US::PSt::MGN
> > 324::Helen Hulick Beebe Papers, 1927-1998//EN" "beebe.sgm"</eadid> .
> . .
> >
> > Am I worrying about retaining legacy data that is unnecessary or is
> there
> > a
> > good reason that the extension should remain? Thanks for your
> insight.
> >
> > Sue
> >
> > Susan Hamburger, Ph.D.
> > Manuscripts Cataloging Librarian
> > Cataloging Services
> > 126 Paterno Library
> > The Pennsylvania State University
> > University Park, PA 16802
> >
> > 814/865-1755
> > FAX 814/863-7293
> >
>
