Hi all,
I think after our discussion, it is good to have some text to see that we are all on the same page. I will try to describe the use cases as follows:
Assuming we have Example A.1.1
- lexicographicResource (uri: http://example.com/myLexicon, langCode: en) - entry (id: abandon-verb, headword: abandon) - partOfSpeech (tag: verb) - sense (id: abandon-verb-1) - definition (text: to suddenly leave a place or a person) - example (text: I'm sorry I abandoned you like that.) - example (text: Abandon ship!) - label (tag: idiom) - sense (id: abandon-verb-2) - label (tag: mostly-passive) - definition (text: to stop supporting an idea) - example (text: That theory has been abandoned.)
I see the following use cases
Use Case 1: A dynamic web server application supporting editing of all elements.
In this case, the URLs do the following
http://www.example.com/myLexicon: Opens an editing/search interface for the whole lexicon
http://www.example.com/myLexicon/entry/abandon~verb: Opens an editing/display interface for the entry "abandon"
http://www.example.com/myLexicon/entry/abandon~verb/sense/to%20suddenly%20leave%20a%20place%20or%20a%20person: Opens the interface for a sense
Use Case 1b: A dynamic web server application supporting editing but only at the entry level
http://www.example.com/myLexicon/entry/abandon~verb#sense/to%20suddenly%20leave%20a%20place%20or%20a%20person: Opens the entry editor and scrolls to the sense
Use Case 2: Static hosting on GitHub
The uri property in the model should refer to the file as hosted on GitHub
http://user.github.io/repo/lexicon.xml: The XML serialization of the entry
http://user.github.io/repo/lexicon.xml#entry/abandon~verb: The entry within this XML file
Use Case 3: XML (or other) to RDF conversion with URIs
The current conversions in the example use ad-hoc IDs blank nodes in the conversion, but blank nodes are considered harmful. It would be better to assign relative URIs to each element consistently and in a standard way:
e.g.,
<#> a dmlex:LexicographicResource; dmlex:title "Example Dictionary"; dmlex:uri "http://example.com"; dmlex:langCode "en"; dmlex:entry <#entry/abandon~verb> .<#entry/abandon~verb> a dmlex:Entry; dmlex:sense <#entry/abandon~verb/sense/to%20suddenly%20leave%20a%20place%20or%20a%20person>; dmlex:partOfSpeech [ dmlex:tag "verb"; dmlex:listingOrder "1"^^xsd:unsignedInt ]; dmlex:headword "abandon".<#entry/abandon~verb/sense/to%20suddenly%20leave%20a%20place%20or%20a%20person> a dmlex:Sense .
URLs are defined relatively in a single file, meaning the file can be hosted directly (as per Use Case 2).
---
Do we agree that all these use cases are within scope of this committee? (It does not matter how important they are)
Regards,
John
Hi John, just to be sure: is there a mistake in Use Case 2, line (see below) http://user.github.io/repo/lexicon.xml: The XML serialization of the entry, and the "serialization of the entry" is in fact "serialization of the LEXICON", or there is something I don't understand?
Regards, Simon
OASIS Open400 TradeCenter, Suite 5900Woburn, MA 01801USA
Phone+1 781 425 5073
Get Involved
Join an Open Project
Join a Technical Committee
About UsPrivacy