[redland-dev] Entities in RDFa

Richard Smith richard at ex-parrot.com
Mon Sep 2 08:53:27 EDT 2013


Is this list dead?  Should I be posting elsewhere?

Richard


Richard Smith wrote:

>
> Last January I submitted a patch to raptor that added a added a new parser 
> option called 'loadDTD'.  The patch hasn't been applied and there's been no 
> subsequent discussion that I've seen.  Perhaps it's my fault for going about 
> submitting the patch in the wrong way or the wrong place, and if so, 
> apologies.
>
> However I still feel that the patch is of definite advantage to raptor, in 
> particular in its handling of entities in RDFa where it's very common to see 
> an example like this:
>
>  <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML+RDFa 1.0//EN"
>      "http://www.w3.org/MarkUp/DTD/xhtml-rdfa-1.dtd">
>  <html xmlns="http://www.w3.org/1999/xhtml"
>        version="XHTML+RDFa 1.0" xml:lang="en">
>    <head>
>      <title>Test</title>
>    </head>
>    <body>
>      <p>This page was written by
>        <span xmlns:dc="http://purl.org/dc/elements/1.1/"
>              property="dc:creator">José</span>.</p>
>    </body>
>  </html>
>
> Note the use of the HTML eacute entity in the name José. (This got mangled in 
> the web archive last time round.) When I submitted the patch, rapper 2.0.6 
> couldn't parse this, and testing today with 2.0.9 it is still the case. If 
> Redland is to be of use with real-world RDFa, without an otherwise 
> unnecessary additional pre-procesing stage, this needs fixing.
>
> The argument that W3 add a 30s delay in serving the DTDs is largely 
> irrelevant.  With the patch, DTD loading only happens only if you 
> specifically request it, and if you have a suitable XML catalog, libxml2 
> won't fetch the DTD from the W3 but from a local copy on your machine. 
> (Under Debian, the w3c-sgml-lib package installs such a catalog for you.) And 
> in any case, if you really need to parse the entity, there will be cases when 
> the 30s delay is quite acceptable.
>
> Is it worth me reworking the patch so that it applies cleanly against the 
> current code base?
>
> Richard


More information about the redland-dev mailing list