[MarkLogic Dev General] xhtml loading question
James Clippinger
James.Clippinger at marklogic.com
Wed Aug 15 12:43:14 PDT 2007
Well, it depends. I think most people keep attributes in the no
namespace unless there is a convincing reason to use a different
namespace. On the search side, it would depend on your content format
and your business requirements, but doing global XPath on attributes
(e.g., //@size) tends to be rare, so name collisions of attributes are
generally less of a concern than element-name collisions.
Regards,
James
> -----Original Message-----
> From: general-bounces at developer.marklogic.com
> [mailto:general-bounces at developer.marklogic.com] On Behalf Of
> Helen Chen
> Sent: Wednesday, August 15, 2007 3:22 PM
> To: General Mark Logic Developer Discussion
> Subject: RE: [MarkLogic Dev General] xhtml loading question
>
> Thanks for explain it.
>
> Is it good habit to make sure all the attributes for the
> element belong to the same namespace during loading? Or
> otherwise maybe attributes without namespace are easier for searching?
>
> Helen
>
> >>> "James Clippinger" <James.Clippinger at marklogic.com>
> 08/15/07 3:06 PM
> >>> >>>
> Helen, default namespaces never apply to unprefixed
> attributes; attributes without a prefix are always in the no
> namespace. This is an XML specification (see
> http://www.w3.org/TR/2006/REC-xml-names-20060816/#defaulting)
> that has been carried through to XQuery.
>
> I don't know the reasoning behind that design decision by the
> recommendation writers and I personally dislike the
> differential treatment of elements vs. attributes, but it's
> how the language works.
>
> Regards,
> James
>
> > -----Original Message-----
> > From: general-bounces at developer.marklogic.com
> > [mailto:general-bounces at developer.marklogic.com] On Behalf Of Helen
> > Chen
> > Sent: Wednesday, August 15, 2007 2:35 PM
> > To: General Mark Logic Developer Discussion
> > Cc: Helen Chen
> > Subject: RE: [MarkLogic Dev General] xhtml loading question
> >
> > Hi James,
> >
> > I'm interested in this topic and I have an question about the
> > namespace for the
> > attribute:
> >
> > At the data loading time, Dev declared namespace as
> > "http://www.w3.org/1999/xhtml
> > xdmp:document-insert("/content/htmlchk66.html",
> > <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en"
> > lang="en"> <body> <meta name="org" content="abc-de" /> <meta
> > name="product" content="electronics" /> <meta name="education"
> > content="school" /> <meta name="test"
> > content="answer123" /> </body>
> > </html>)
> >
> > My understanding for namespace is: if you declare the namespace,
> > unless you change it or it goes out of the scope, all the following
> > elements and attributes belong to this namespace.
> >
> > But you mentioned that the attributes in Dev's data does
> not have any
> > namespace, I think I misunderstood the concept. Can you explain it
> > more?
> >
> > Thanks, Helen
> >
> >
> >
> > >>> "James Clippinger" <James.Clippinger at marklogic.com>
> > 08/14/07 5:41 PM
> > >>> >>>
> > Dev, the syntax you used would associate the prefix "html"
> > with the XHTML namespace; you could use it like this:
> >
> > declare namespace html="http://www.w3.org/1999/xhtml"
> >
> > cts:search(xdmp:directory(('/'),"infinity"),
> > cts:and-query((cts:element-attribute-value-query(xs:QName("htm
> > l:meta"),
> > xs:QName("name"),
> > "org"),cts:element-attribute-value-query(xs:QName("html:meta"),
> > xs:QName("content"), "abc-de"))))
> >
> > Now, you could use the "default element namespace" syntax to set the
> > default:
> >
> > default element namespace="http://www.w3.org/1999/xhtml"
> >
> > cts:search(xdmp:directory(('/'),"infinity"),
> > cts:and-query((cts:element-attribute-value-query(xs:QName("meta"),
> > xs:QName("name"),
> > "org"),cts:element-attribute-value-query(xs:QName("meta"),
> > xs:QName("content"), "abc-de"))))
> >
> > However, this will return no results in your example, because
> > xs:QName() will, when given an unprefixed string, return
> QNames in the
> > default element namespace. This is the right thing for the QNames
> > being used as elements, but it causes the attribute QNames not to
> > match since they are in the no namespace. This version would work:
> >
> > default element namespace="http://www.w3.org/1999/xhtml"
> > declare namespace no-namespace=""
> >
> > cts:search(xdmp:directory(('/'),"infinity"),
> > cts:and-query((cts:element-attribute-value-query(xs:QName("meta"),
> > xs:QName("no-namespace:name"),
> > "org"),cts:element-attribute-value-query(xs:QName("meta"),
> > xs:QName("no-namespace:content"), "abc-de"))))
> >
> > I think the html-prefix form is more intuitive than this
> one, but it
> > is legal XQuery and will work just fine.
> >
> > Regards,
> > James
> >
> >
> >
> > ________________________________
> >
> > From: general-bounces at developer.marklogic.com
> > [mailto:general-bounces at developer.marklogic.com] On Behalf
> Of Devadoss
> > P
> > Sent: Tuesday, August 14, 2007 5:16 PM
> > To: General Mark Logic Developer Discussion
> > Subject: Re: [MarkLogic Dev General] xhtml loading question
> >
> >
> > Thanks James, it's working.
> >
> > I couldn't understand, why declaring default namespace
> is not working
> > in this case (declare namespace html =
> > "http://www.w3.org/1999/xhtml").
> >
> > Is it because of xhtml namespace?
> >
> > Thanks,
> > Dev
> >
> >
> >
> >
> > On 8/14/07, James Clippinger <James.Clippinger at marklogic.com >
> > wrote:
> >
> > Dev, you need to include the XHTML namespace in
> the element QNames
> > you create for use in the cts:search() if you want the
> query to match
> > those elements. This would work:
> >
> > cts:search(xdmp:directory(('/'),"infinity"),
> > cts:and-query((cts:element-attribute-value-query(fn:QName("
> > http://www.w3.org/1999/xhtml <http://www.w3.org/1999/xhtml> ",
> > "meta"), xs:QName("name"),
> > "org"),cts:element-attribute-value-query(fn:QName("
> > http://www.w3.org/1999/xhtml <http://www.w3.org/1999/xhtml> ",
> > "meta"), xs:QName("content"), "abc-de"))))
> >
> > Note that the attribute QNames don't require
> namespace URIs because,
> > having not been prefixed in the original source, they are in the no
> > namespace. If you wanted to use the fn:QName() function for
> > consistency, this form also works:
> >
> > cts:search(xdmp:directory(('/'),"infinity"),
> > cts:and-query((cts:element-attribute-value-query(fn:QName("
> > http://www.w3.org/1999/xhtml <http://www.w3.org/1999/xhtml> ",
> > "meta"), fn:QName("", "name"),
> > "org"),cts:element-attribute-value-query(fn:QName("
> > http://www.w3.org/1999/xhtml <http://www.w3.org/1999/xhtml> ",
> > "meta"), fn:QName("", "content"), "abc-de"))))
> >
> > Regards,
> > James
> >
> >
> > ________________________________
> >
> > From: general-bounces at developer.marklogic.com
> > [mailto:general-bounces at developer.marklogic.com] On Behalf
> Of Devadoss
> > P
> > Sent: Tuesday, August 14, 2007 4:49 PM
> >
> > To: General Mark Logic Developer Discussion
> > Subject: Re: [MarkLogic Dev General]
> xhtml loading question
> >
> >
> >
> >
> > James,
> >
> > This is the query used to load the document,
> > xdmp:document-insert("/content/htmlchk66.html",
> > <html xmlns="http://www.w3.org/1999/xhtml "
> > xml:lang="en" lang="en">
> > <body>
> > <meta name="org" content="abc-de" />
> > <meta name="product" content="electronics" />
> > <meta name="education" content="school" />
> > <meta name="test" content="answer123" />
> > </body>
> > </html>)
> >
> > and my cts query is
> > cts:search(xdmp:directory(('/'),"infinity"),
> > cts:and-query((cts:element-attribute-value-query(xs:QName("meta"),
> > xs:QName("name"),
> > "org"),cts:element-attribute-value-query(xs:QName("meta"),
> > xs:QName("content"), "abc-de"))))
> >
> > This query returns result, if I load
> > document without xmlns.
> >
> > Thanks,
> > Dev
> >
> >
> > On 8/14/07, James Clippinger
> > <James.Clippinger at marklogic.com > wrote:
> >
> > Dev, can you send the
> > cts:search() call you're using for your search?
> >
> > Thanks,
> > James
> >
> >
> > ________________________________
> >
> > From:
> > general-bounces at developer.marklogic.com [mailto:
> > general-bounces at developer.marklogic.com] On Behalf Of Devadoss P
> > Sent: Tuesday, August 14, 2007 4:33 PM
> > To: General Mark Logic Developer
> > Discussion
> > Subject: Re: [MarkLogic Dev
> > General] xhtml loading question
> >
> >
> >
> > James and Mike,
> >
> > How to resolve
> > (xmlns="http://www.w3.org/1999/xhtml") namespace conflict in
> > xquery; I loaded html document (which has xmlns namespce
> > defined) to MarkLogic and tried to query its element by using
> > cts:search(), but it returns empty row.
> >
> > Can you help me to resolve
> > namespace issue?
> >
> > Thanks,
> > Dev
> >
> >
> > On 8/14/07, Alex Rice
> > <alexr at santafe.edu> wrote:
> >
> > James and Michael:
> >
> > i had not even thought of namespaces.
> > now I am getting nodes back and
> > can use the declare namespace
> > feature etc. Thank you both for
> > responding.
> >
> > Alex Rice
> >
> >
> > On 8/14/07, James Clippinger
> > <James.Clippinger at marklogic.com > wrote:
> > > Alex, I'm not familiar with
> > the XHTML created by Dreamweaver or UpCast,
> > > but I suspect the XHTML
> > elements both tools create are in the XHTML
> > > namespace. Try this (keeping
> > in mind the wildcard-namespace syntax may
> > > be slow):
> > >
> > >
> > doc("/SFI/Update/news/articletest2/test.xml")//*:p
> > >
> > > If it returns items, you can
> > see which namespace each item is in and go
> > > from there.
> > >
> > > Regards,
> > > James
> > >
> >
> > _______________________________________________
> > General mailing list
> > General at developer.marklogic.com
> >
> > http://xqzone.com/mailman/listinfo/general
> >
> >
> >
> >
> >
> > _______________________________________________
> > General mailing list
> > General at developer.marklogic.com
> >
> > http://xqzone.com/mailman/listinfo/general
> >
> >
> >
> >
> >
> > _______________________________________________
> > General mailing list
> > General at developer.marklogic.com
> > http://xqzone.com/mailman/listinfo/general
> >
> >
> >
> >
> > _______________________________________________
> > General mailing list
> > General at developer.marklogic.com
> > http://xqzone.com/mailman/listinfo/general
> >
> _______________________________________________
> General mailing list
> General at developer.marklogic.com
> http://xqzone.com/mailman/listinfo/general
> _______________________________________________
> General mailing list
> General at developer.marklogic.com
> http://xqzone.com/mailman/listinfo/general
>
More information about the General
mailing list