- From: Felix Sasaki <fsasaki@w3.org>
- Date: Thu, 4 Oct 2012 19:39:16 +0200
- To: Yves Savourel <ysavourel@enlaso.com>
- Cc: public-multilingualweb-lt@w3.org
- Message-ID: <CAL58czpw51BT7_xL7N+q0_tt0d9ScMq1VTE-csdacw66mjXXoQ@mail.gmail.com>
Thanks, Yves. So how about specifying the processing of domain information against domainMapping: An application implementing domain MUST do the following for processing "domain mappings": - If no "domain mapping" attribute is given, the value coming out of evaluating "domainPointer" is just passed to the application - If a "domain mapping" attribute is given and the evaluation EDP of the XPath "domainPointer" leads to a list of values: 1) map each value from EDP that has a mapping in "domain mapping" to the consumer specific values 2) pass all other values to the application - If a "domain mapping" attribute is given and the evaluation EDP of the XPath "domainPointer" leads to one value that contains 2C KOMMA: 0) split the value of EDP into separate values, using KOMMA as a separator apply steps 1)-2) as before. Felix 2012/10/4 Yves Savourel <ysavourel@enlaso.com> > > A domain pointer could point to the "subjectterm" elements > > <its:domainRule selector="/db:article/db:para" > domainPointer="/db:article/db:info//db:subjectterm"/> > > This would give a list of strings. > > > > With this HTML > > <meta name="keywords" content="news, sports, politics, weather" /> > > We would have one string, which is a comma separated list of strings. > > > > Yves, all, is the above the issue? > > I don't know a solution yet, just want to be sure. > > Yes Felix, it's the issue. > > -yves > > > > -- Felix Sasaki DFKI / W3C Fellow
Received on Thursday, 4 October 2012 17:39:44 UTC